THE CHEMISTRY OF THE-NON-BENZENOID HYDROCARBONS- And Their Simple Derivatives BY -tx BENJAMIN T. BROOKS, PH.D. First Edition BOOK DEPARTMENT The CHEMICAL CATALOG COMPANY, Inc. ONE MADISON AVENUE, NEW YORK, U. S. A, 1922 COPYRIGHT, 1922, BY The CHEMICAL CATALOG COMPANY, Inc. All Rights Reserved Press of J. J. Little & Ives Company New York, U. 8. A. PREFACE The beautiful, interesting, and often facile chemistry of the benzene hydrocarbons has somewhat overshadowed the chemistry of the ali- phatic open chain and cyclic non-benzenoid hydrocarbons. Certainly the chemistry of the former series has been much more fully rounded out. Judging from the customary method of treatment accorded them in our textbooks, there is some confusion in the arrangement of subject matter which does not give the student a proper idea of the close re- lationships and similarity of chemical behavior possessed by all the non-benzenoid hydrocarbons. Mr. Wells, in his "Outline of History," says: "There is a natural tendency in the human mind to exaggerate the differences and resemblances upon which classification is based, to suppose that things called by different names are altogether different, and that things called by the same name are practically identical. This tendency to exaggerate classification produces a thousand evils, . . ." This tendency, which Mr. Wells deplores, is well shown by the use of the term "hydro-aromatic" hydrocarbons and the classifi- cation of cyclohexane and its derivatives with benzene. This term is still employed for cyclohexane and its simple derivatives, although its behavior is almost identical with that of normal hexane. The same applies to cyclopentane and its simple derivatives as compared with normal pentane, yet cyclopentane cannot be termed a "hydro-aromatic" hydrocarbon. Cycloheptane, cyclooctane, cyclononane, cyclobutane and cyclopropane should certainly be classed and described together with cyclohexane, as indeed Aschan and a few others have done. The differences in chemical properties between benzenoid ring systems and NON-BENZENOID hydrocarbons are well established, but in spite of the enormous amount of work done, have not yet received adequate explanation. As regards the aliphatic hydrocarbons proper, these fare very badly in most works on organic chemistry, particularly in the briefer text- books. Usually, the entirely erroneous statement, or implication, is made that the so-called type reactions given for the first two or three members of the methane series hold good for the higher members. Be- yond the fact that all of them may be completely burned to carbon 3 .72442 fc *-W 1 ^ rW 4 PREFACE dioxide and water, such statements are hardly in accord with the known facts. We note that the chemistry of the first five members of the methane series and also the ten carbon atom or terpene group, mostly cyclic hydrocarbons, have been much more extensively and carefully studied than the remainder. Some of the reasons for this are fairly apparent. Thus, the essential oils afford a convenient source of substances of the terpene group which may generally be isolated easily in a state of purity. The natural fatty glycerides or other con- venient sources readily yield a limited number of fatty acids, nearly all of them normal, i. e., acids of one, two, three, four, five, six, eight, ten, twelve, fourteen, sixteen, eighteen, and twenty-four carbon atoms. Research in many of these special fields has accordingly been greatly facilitated by the availability of suitable material and has often been much stimulated by an intimate relation to industry. It may also be pointed out that while in the aromatic series a rich variety of raw materials may easily be isolated or prepared, crystal- line derivatives are almost the rule, permitting easy purification, iden- tification, and manipulation in small quantities; that substitution re- actions are usually capable of control to form chiefly one product or a very limited number of products or isomers; but in the aliphatic series this is not the case. Petroleums probably contain all of the nor- mal paraffine hydrocarbons up to C 26 H 54 and perhaps farther in the series, and perhaps hundreds of naphthenes which are for the most part yet unknown. Not only is it at present impossible to isolate pure individual substances from this complex raw material, but few methods of synthesis applicable to the higher members of the aliphatic series or the more complex naphthenes have been developed. The reader seeking only material of industrial interest may object to the inclusion of much subject matter which is solely of theoretical interest and the searcher who scorns industrial processes will find much in the present volume that is unorthodox. The author desires to make no apology for the inclusion of both classes of subject matter; the de- scription of any special subject of science should be systematic if we are to retain our conception of science as classified knowledge, and the author does not feel that descriptions of industrial processes and refer- ences to patent literature detract from the value of the compilation, considered as a scientific monograph. In a treatise of purely industrial purpose the checker-board plan, in which economic value determines exclusion or inclusion of subject matter, may perhaps be justified, but the author believes that the best results will be obtained by broader PREFACE 5 scientific treatment of industrial subjects. The author is well aware that patent literature, in spite of oaths and notaries' seals, is not bound by the same standards of truth that govern the publication of purely scientific papers and has accordingly treated such matter critically and with caution. The mechanical art and engineering of petroleum refining has been perfected to a degree which, measured by profit and general utility, deserves commendation, but it is a development which has been very little dependent upon chemical knowledge. More thorough knowledge of the chemistry of the non-benzenoid hydrocarbons will surely re- sult in better and less wasteful methods of refining and may lead to the conversion of petroleum hydrocarbons into other useful products by chemical methods. In the present state of our knowledge, it would be rash to prophesy what may be accomplished in this direction; but be- fore much work of this kind can be done, a great deal of painstaking, systematic research in the field of the non-benzenoid hydrocarbons must be carried out which may never be utilized directly in an industrial process. The writer does not urge research in this field solely on the ground of the utility of the possible results. Those who attempt to justify scientific research by financial returns do not always have a very strong case, and to attempt to balance any particular industry upon the point of an original scientific discovery is to leave out of ac- count the contributions of a host of other people, which the scientist seldom appreciates. Such arguments convince nobody and often arouse the resentment of engineers and business men and others who know better. The upbuilding of a great mass of information and generaliza- tions, new experimental methods and new substances, in the field of the non-benzenoid hydrocarbons, will enable industry to select certain bits of knowledge suited to further progress and our everyday welfare. Every original investigator making real contributions to the fabric of knowledge is thus a contributor to the common weal. This, while not the sole justification of research, is the correct form of the argu- ment of the utility of scientific investigation. This point of view has a very direct bearing on the question of re- search in the field of the non-benzenoid hydrocarbons. The petroleum, rubber, turpentine and essential oil industries stand in need of further systematic theoretical research in this field of chemistry. Work along broad lines, involving the work of a great many investigators for a great many years, is required. American chemists have heretofore played a singularly insignificant part in this field of research and to 6 PREFACE realize this it is only necessary to mention the names of Wallach, Sir William H. Perkin, Jr., Semmler, Engler, Grignard, Sabatier and the Russian group, Ipatiev, Kishner, Markownikow, Wagner, Konowalow, Zelinsky, Aschan, Bredt, Ostromuislenski, Lebedev, Gustavson, Char- itschkov, and others. All of these men have exercised their influence in universities or technical schools, and the inference may accordingly be drawn that we must look to our American universities, rather than to the petroleum or other industrial interests, to initiate and carry on such research in America. And if the American petroleum industries second their efforts, as the Nobel Brothers have done in Russia, a vast amount of work of permanent scientific and potential industrial value can be done. The present monograph is not a catalog of all the hydrocarbons which might be described. The writer has endeavored to show the close relationships which hold generally throughout the chemistry of the non-benzenoid hydrocarbons and, on the other hand, to point out that the chemical behavior of the more complex hydrocarbons of the paraffine series and the alicyclic hydrocarbons cannot be assumed from the chemical behavior of a few of the simpler hydrocarbons. The chem- istry of the ethylene bond is emphasized because of its great impor- tance and because most of our knowledge of its behavior under dif- ferent circumstances and influences is empirical. Much important work has been done since the appearance about twenty years ago of Aschan's "Alicyclische Verbindungen" and Semm- ler's admirable volumes on the terpenes and this work has been briefly reviewed and the attempt has been made to treat it in such a way that will be helpful in wider fields of organic research. TABLE OF CONTENTS. PAGE CHAPTER I. THE PARAFFINES 13 1. Occurrence of the paraffines in nature. a. Natural gas; composition, behavior under pressure; separation of the con- stituents. b. Petroleum: difficulty of isolating simpler members of the paraffine series: paraffines produced by bio- logical processes; general character and probable mode of origin of petroleums. c. Other natural sources of paraffines. 2. Formation of the paraffines. a. Pyrolysis of organic matter; effects of heat on non-benzenoid hydrocarbons; oil gas ; formation of aromatic hydrocarbons ; the gasoline prob- lem. 6. Synthesis of the paraffines. Alkyl halides and metallic couples; the Grignard reaction; reduction of alco- hols or alkyl iodides by hydriodic acid; catalytic hydro- genation of olefines; miscellaneous special methods. CHAPTER II. CHEMICAL PROPERTIES OF SATURATED HYDRO- CARBONS 52 1. Oxidation; conversion of paraffine wax to fatty acids by air oxidation; hardening of petroleum residuums by blowing with air. Other oxidizing reagents: 2. Behavior to nitric acid; nitration with dilute nitric acid. 3. Alkyl halides. General methods for preparing alkyl chlorides, bromides and iodides; dissociation of the simpler alkyl halides by heat; behavior of alkyl halides to alcoholic alkali; general reac- tions. CHAPTER III. THE PARAFFINE HYDROCARBONS .... 76 1. Methane; oxidation; inflammability; chlorination ; syn- thesis from water gas or carbon monoxide. 2. Ethane, propane, butanes. 3. The pentanes; hexanes; heptanes. 4. Octanes; synthesis of octanes as typical of methods now known; nonanes and decanes. 5. Paraffines C 10 H 22 to C 6 oH 122 ; paraffine wax from petroleum. 6. Table: Physi- cal properties of the paraffines. 7. Notes on the refining of petroleum distillates. CHAPTER IV. THE ETHYLENE BOND Ill 1. Recent conceptions of valence and the ethylene bond; Baeyer's strain theory; stability of the ethylene bond and 7 TABLE OF CONTENTS PAGE carbocyclic structures. 2. Chemical properties of unsatu- rated substances of the ethylene type. a. Modification of the chemical behavior of the ethylene bond by substituents ; influence of the double bond on the chemical behavior of substituents. b. Addition reactions, halogens, halogen acids, hypochlorous acid, aqueous mineral acids and the addition of water. c. Unsaturated hydrocarbons and sul- furic acid; refining of petroleum oils. d. Auto-oxidation e. Reaction with sulfur and sulfur chloride; vulcanization of rubber. /. Ozonides: Use of ozone in determining con- stitution. g. Properties of the HC = CH CO group. h. Addition reactions frequently used for identification of unsaturated hydrocarbons; nitrosyl chloride, nitrous acid; use of nitrosyl chloride for the synthesis of ketones. i. Other substances which combine with the ethylene bond, aniline, urea, hydrogen sulfide, hydrocyanic acid, etc. 3. The preparation of unsaturated hydrocarbons: Decom- position of saturated hydrocarbons, alcohols and organic halides by heat; barium soaps and sodium ethoxide; the Grignard reaction; exhaustive methylation of amines. CHAPTER V. THE ACYCLIC UNSATURATED HYDROCARBONS . .158 1. Ethylene, physical properties, chemical behavior; pro- duction from acetylene, from ethyl alcohol; coal gas, oil gas; catalytic oxidation to formaldehyde; |3p-dichloroethyl sulfide; reaction with sulfuric acid and the industrial syn- thesis of alcohol; Hofmann and Sand's ethanol compounds. 2. Propylene; physical properties and general chemical behavior and rules of addition; industrial propyl alcohols. 3. Butylenes and amylenes; chemical behavior. 4. Ole- fines, six to nine carbon atoms; difficulty of synthesis or separation of pure hydrocarbons. 5. Decene's and ali- phatic terpenes; myrcene, ocimene and allo-ocimene. 6. Derivatives of 2 . 6-dimethyl-octane ; the citral group; gera- niol and citral a nerol and citral 6; linalool; citronellol; a- and (3-ionone; irone. 7. Sesquicitronellene ; spinacene. 8. Cholesterylene and its relation to cholesterol. CHAPTER VI. POLYMERIZATION OF HYDROCARBONS . . . 210 1. Substituted ethylenes and the effect of substituents on polymerization; the conjugated dienes, their chemical be- havior and the synthesis of rubbers. 2. The constitution of rubber, its depolymerization ; review of research on the syn- thesis of rubber; raw materials ' and the question of indus- trial synthesis. 3. Methods of polymerization. TABLE OF CONTENTS 9 PAQB CHAPTER VII. CYCLIC NON-BENZENOID HYDROCARBONS . . 233 1. General methods of synthesis. By polymerization of un- saturated hydrocarbons; decomposition of calcium and barium salts of dicarboxylic acids; condensation of dicar- boxylic acid esters by sodium; by sodium and malonic acid ester; the Grignard reactions; dihalogen derivatives and sodium; disodium derivatives of carboxylic acids and iodine or bromine; ring closing by elimination of water from alde- hydes; diazoacetic ester and the synthesis of cyclopropane derivatives; condensation of nitriles by sodium ethylate to imino compounds and their hydrolysis; Kishner's hydrazine method; hydrogenation of benzenoid hydrocarbons. 2. Cyclopropane and its simple derivatives. 3. Cyclobutane and its simple derivatives.- 4. Cyclopentane and its simple derivatives. a. Syntheses from cyclopentanone. b. Naph- thenic acids, synthetic and from petroleums. c. Substi- tuted cyclopentanes. CHAPTER VIII. CYCLIC NON-BENZENOID HYDROCARBONS: THE CYCLOHEXANE SERIES 278 1. The hydrogenation of benzene; catalytic production of cyclohexanols and cy clohexanone ; cyclohexene and cyclo- hexadienes. 2. Alkyl derivatives of cyclohexane, synthetic and from petroleum; cantharene. 3. Mono-cyclic sesqui- terpenes. CHAPTER IX. CYCLIC NON-BENZENOID HYDROCARBONS: THE PARA-MENTHANE SERIES 315 1. Limonene and dipentene; carvomenthene ; para-men- thane; the constitution of limonene; syntheses of limonene and the terpineols. 2. Terpinolene and the terpinenes; Semmler's carvenene. 3. Crithmene. 4. The oxides; gen- eral behavior of oxides; 1.8-cineol, 1.4-cineol, pinol and ascaridol. 5. Other menthenols. 6. Menthol; stereochem- istry of menthol and menthone; the menthenones, piperi- tone and pulegone; Buchu camphor; carvone. 7. The phel- landrenes. CHAPTER X. CYCLIC NON-BENZENOID HYDROCARBONS: ORTHO- AND META-MENTHANE DERIVATIVES 384 1. Sylvestrene; Its synthesis from carvone; Perkin's syn- thesis. 2. Ortho-menthane derivatives; synthesis by Per- kin. 10 TABLE OF CONTENTS PAGE CHAPTER XI. CYCLIC NON-BENZENOID HYDROCARBONS: BICYC- LIC AND TRICYCLIC HYDROCARBONS 396 1. Santene. 2. Sabinene, thujene and carene. 3. Tetra- hydro and decahydronaphthalene. 4. Hydrogenation of indene, anthracene and phenanthrene. 5. Nomenclature of bicyclic and tricyclic hydrocarbons. 6. Bicyclic and tri- cyclic sesquiterpenes. CHAPTER XII. BICYCLIC NON-BENZENOID HYDROCARBONS: THE PlNENES AND FENCHENES 420 1. Character of commercial turpentines. 2. Constitution of a-pinene; chemical reactions of a-pinene. 3. Beta-pinene; synthesis and constitution. 4. Bornyl chloride and its de- composition products. 5. Pinolene; tricyclene; the fen- chenes. CHAPTER XIII. BICYCLIC NON-BENZENOID HYDROCARBONS: CAM- PHENE, BORNYLENE AND CAMPHOR . . . . . 453 1. Review of research of the constitution of camphene and bornylene. 2. a. Camphor; constitution of camphor and its oxidation products; camphoric and related acids. b. Epicamphor. c. Derivatives of camphor. 3. Synthetic camphor. a. Plantation camphor vs. synthetic camphor. b. The preparation of bornyl chloride; conversion of bornyl chloride to camphene, bornyl acetate and borneol; hydra- tion of camphene to borneol. c. Other processes for the conversion of pinene to borneol; the Thurlow and similar processes. d. Oxidation of the borneols; impurities in syn- thetic borneols and camphor. CHAPTER XIV. CYCLIC NON-BENZENOID HYDROCARBONS : CYCLO- HEPTANE, CYCLO-OCTANE, CYCLONONANE AND POLYNAPH- THENES 511 1. Cy cloheptane ; cycloheptene, cycloheptadiene and cyclo- heptatriene. 2. Cycloheptanone ; eucarvone. 3. Cyclo- octane; cyclo-octotetrene. 4. Cyclononane. 5. Polynaph- thenes; lubricating oils. CHAPTER XV. REARRANGEMENTS .524 Cyclobutane and cyclopentane derivatives; a-pinene and bornyl chloride; cyclobutyl amine to cy clopentanol ; cyclo- pentane and cyclohexane derivatives; Meerwein's researches on pinacones; borneol and camphene. TABLE OF CONTENTS 11 PAOB CHAPTER XVI. PHYSICAL PROPERTIES 538 1. Density and molecular volume; melting-point and boil- ing-point. 2. Optical properties; absorption of light, color and fluorescence; molecular refraction and influence of structure on refractivity ; molecular dispersion ; magnetic rotation; optical activity and methods of synthesis of opti- cally active hydrocarbons; optical activity of petroleum. 3. Thermochemistry of the non-benzenoid hydrocarbons; specific heat; latent heat of vaporization; heat of combus- tion. 4. Dielectric constants; static charges of oils pro- duced by friction; transformer oils. 5. Viscosity; effect of ring closing on viscosity; viscosity of petroleum oils; vis- cosity and lubrication; effect of dissolved paraffine on vis- cosity of oils. 6. Solubility; petroleum fractions in other solvents; paraflme wax in various solvents; terpene hydro- carbons in dilute alcohol; solubility of methane and other gases in oils; sulfur in petroleum oils; dissolved sulfur in rubber; liquid sulfur dioxide as a solvent for unsaturated hydrocarbons and Edeleanu's refining process. 7. Colloids; greases and jellies; emulsions; adsorption and the use of fuller's earth; fractional separation of hydrocarbons by fuller's earth. CHAPTER XVII. PHYSIOLOGICAL AND RELATED PROPERTIES . 591 1. Odor. 2. Physiological effects; narcotic action of the simpler hydrocarbons; terpene alcohols and ketones; nat- ural and synthetic camphor; halogen derivatives of the paraffines. Chapter I. The Paraffines In any systematic treatment of the non-benzenoid hydrocarbons, it is difficult to subdivide the subject matter into divisions or chapters, which do not unduly emphasize minor class differences. Thus cyclo- hexane is not ordinarily considered as a paraffine or saturated hydro- carbon although its chemical behavior might very properly place it in this class. On the other hand, the cyclopropane ring frequently ex- hibits properties of unsaturation which are nearly identical with those characteristic of the ethylene bond. However, since a discussion of the hydrocarbons of the series C n H 2 n+ 2 may rationally serve as a ground work, this series will be considered first. Occurrence of the Paraffines. From the economic standpoint by far the most important natural sources of the paraffine hydrocarbons are natural gas 1 and petroleum. The industrial utilization of natural gas has been practically limited to the United States, although the Chinese may claim priority as re- gards its first industrial use since old Chinese writings describe its collection from shallow dug wells, piping through tubes of bamboo and burning for the evaporation of brine. Since practically all the natural gas produced in the United States is consumed as fuel or burned for the production of carbon black, very little attention has been paid to its chemical composition. In rare in- stances natural gas contains as much as 95 per cent methane but an average gas contains about 85 per cent methane, 1.0 to 3 per cent nitrogen and 12 to 15 per cent ethane and other paraffines. Unusual geological conditions, but little understood, result in gases containing large percentages of nitrogen, hydrogen sulfide or carbon dioxide. Hy- drogen sulfide is normally not a constituent of natural gas but is fre- quently encountered in gases in the Gulf Coast territory. Nitrogen occurs in the gas of the northern Texas fields to the extent of about 38 per cent and it is of interest to note that this gas also contains helium 1 In 1917 the consumption of natural gas in the United States was 795 billion cubic feet. (Northrop in Westcotts' "Handbook of Natural Gas," p. 106.) 13 14 [CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS 'hi amtiunts sufficient for its extraction on a large scale for filling dirig- ible balloons. The composition of natural gas is usually reported in terms of methane and ethane, these percentages being derived by cal- culation from the results of combustion in an explosion pipette. That hydrogen does not occur in natural gas is now generally accepted, Philipps 2 having shown that the early analyses in which hydrogen was reported, were faulty. Typical analyses reported by Burrell and Oberf ell 3 are as follows : TYPICAL ANALYSES OF NATURAL GAS. B.T.U.per cu. ft. CH 4 C 2 H C0 2 N 2 O a (760mm. Source of Gas % % % % % OC.) Texarkana, Ark 96 0.0 0.8 3.2 0.0 1,022 Noblesville, Ind 86.8 6.2 0.8 6.2 0.0 1,040 Leavenworth, Kan 91.3 4.5 0.8 3.4 0.0 1,066 Erie, N. Y 79.9 15.2 0.0 4.9 0.0 1,134 Columbus, 80.4 18.1 0.0 1.5 0.0 1,193 Guthrie, Okla 69.4 20.6 0.1 9.9 0.0 1,062 Muskogee, Okla 92.1 4.1 0.4 3.4 0.0 1,057 Pawhuska, Okla 66.5 20.7 0.3 12.5 0.0 1,093 Fort Worth, Tex. 4 51.3 10.4 0.1 38.2 0.0 740 Bow Island, Canada * ... 87.6 0.9 . . 11.2 The percentages of methane, ethane, propane and higher methane homologues can be determined accurately by fractional distillation at low temperatures. 5 Thus a sample of natural gas supplied to the city of Pittsburgh in 1915 was shown to have the following composition: Methane 84.7 per cent. Ethane 9.4 " " Propane 3.0 ' Butane and other hydrocarbons 1.3 " " Nitrogen 1.6 " " In recent years the practice of removing the light gasoline vapors, mostly butane, pentane and hexane, by absorption and compression methods has become almost universal, at least where large gas supplies are available. High pressure gas from new fields contains relatively very little gasoline vapor, the highest yields being obtained from low pressure gas associated with petroleum. 6 The removal of gasoline va- *Am. Chem. J. 16, 406 (1894). a TJ. S. Bur. Mines. Techn. Paper 109. 4 This gas in northern Texas contains about 0.9% helium which is being separated at the U. S. Government plant at Petrolia, Texas. The Canadian gas contains 0.33% helium. "Burrell, Seibert & Robertson. U. S. Bur. Mines. Techn. Paper 10^ (1915). 6 The yield of gasoline obtained by absorption methods from so-called dry gas is from 0.5 to 0.75 gallons per 1000 cubic feet. When the initial gas pressure is 300 to 500 pounds per square inch the yield of gasoline by the absorption method is about 0.3 gallon per 1000 cubic feet. The compression method alone is not employed when the gas contains less than 0.75 gallons of gasoline per thousand cubic feet of gas. THE PARAFFINES 15 pors slightly lowers the fuel value of the gas, normally one gallon of gasoline per 1000 cubic feet lowering the calorific value of the gas about 5 per cent. 7 The yield of carbon black is considerably dimin- ished by the removal of gasoline vapors from the gas. In common practice the average yield of carbon black was about 1 pound per 750 cubic feet when very rich, low pressure gas was employed for this pur- pose. The behavior of natural gas under pressure is of industrial im- portance from another standpoint, i. e., the measuring or metering of gas under pressure. Although the gas pressure of new wells in new fields may be as high as 1600 pounds per square inch, it is usually necessary to compress the gas from lower pressures to about 650 pounds per square inch for transmission through long pipe lines. Methane deviates considerably under pressure, from the behavior of a perfect gas and Amagat 8 has shown that at 40 atmospheres it is about 9 per cent more compressible and at 100 atmospheres is 17 per cent more compressible than a perfect gas. Burrell and Robertson 9 have shown that the average natural gas is considerably more compressible than pure methane, at 35.5 atmospheres this deviation amounting to about 15 per cent as compared to the compressibility of a perfect gas. The fuel value of natural gas is commonly given as 1000 B.T.U. per cubic foot measured at 0C but owing to the presence of ethane (1719 B.T.U. per cubic foot) and other hydrocarbons, the value 1100 B.T.U. is a better average value. Since in ordinary fuel practice the water formed in the combustion is practically never condensed, the latent heat of evaporation of this water should be deducted to give a net heating value. 10 Ethane, propane and butane may easily be separated from natural gas in conjunction with the removal of gasoline vapors and, as Burrell and Robertson have shown, each of these hydrocarbons may be isolated in a very pure state by fractional distillation at low tempera- tures. In view of the low cost of the separation of oxygen and nitro- gen by liquid air methods, it is certain that pure ethane, propane and butane could be made available in large quantities at very low cost. These hydrocarbons are not now utilized (other than as fuel), but research in the direction of their chemical utilization is in progress. T Dow. TL S. Bur. Mines. Techn. Paper 253 (1920). "Landolt & Bornstein. Physikalische Tabellen, 1905, 65. U. S. Bur. Mines, Techn. Paper 104 (1915). 10 Richards, "Metallurgical Calculations," Ed. 1918, p. 25, gives the net heating value of 970 B. T. U. for methane, the water formed remaining uncondensed. Cf Waidner & Mueller, "Industrial Calorimetry," U. S. Bur. Standards, Techn. Paper S6 16 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The deviation of ethane, propane and butane from the behavior of a perfect gas is, of course, greater than is the case with methane, and when compressing gas mixtures containing all of these hydrocarbons, as in the separation of gasoline from natural gas by compression, the behavior is the resultant of many factors and the most efficient method of operating a compression plant for the production of gasoline can, at the present time, be determined only by experiment. 11 According to well understood principles, when the pressure on a gas, containing condensable vapors is increased, the partial pressure of the vapor in- creases until its saturation pressure is reached at which point conden- sation to liquid begins. Thus if a gas is saturated with pentane vapor at atmospheric pressure, compression to two atmospheres will liquefy one-half of the pentane; if the partial pressure of the pentane is origi- nally one-tenth the saturation pressure, then compression to ten atmos- pheres will be required to reach the saturation point and this pressure must then be doubled, i.e., to twenty atmospheres, to liquefy one- half the pentane. But when other condensable hydrocarbons are pres- ent, these simple relations no longer hold true. The importance of re- moving the heat resulting by compression is indicated by the accom- panying figures showing the vapor pressure curves of the simpler nor- mal paraffine hydrocarbons. See also vapor pressure curves of the simpler paraffines on page 88. Few petroleums consist mainly of hydrocarbons of the paraffine series, but the lighter, low boiling fractions of most petroleums consist of these hydrocarbons almost exclusively. Particularly is this true of light Pennsylvania oil. Since much of the earlier chemical work on petroleum was carried out with distillates of this particular oil, it is often erroneously stated that "American" petroleum consists of paraf- fines and "Russian petroleum" consists of naphthenes and polynaph- thenes of the series C n H 2n ,C n H 2n _ 2 , etc. Generally it may be said that the petroleums of no two producing regions are the same. Although the petroleum typical of the Pennsylvania field probably contains the largest per cent of paraffines, the higher boiling, viscous fractions of this crude contain but a few per cent of C n H 2I1+2 hydrocarbons and these are removed by chilling, thus manufacturing the "paraffine wax" of commerce. Lubricating oil derived from Pennsylvania and other petroleums consists chiefly of hydrocarbons of the class C n H 2n _ 2 , 12 but their structure is unknown and no pure individual hydrocarbons have "Mabery, Am. Chem. J. 1305, 231. " Anderson, J. Ind. d Eng. Chcm. 12, 547 (1920) ; Dykema, U. S. Bur. Mines. Bull. 151 (1918). THE PARAFFINES 17 been isolated from them. Vaseline isolated from Pennsylvania pe- troleum, is a mixture of hydrocarbons of the empirical formulae Degrees Fahrenheit. -13 +32 77 122 167 212 257 302 347 392 437 482 527 572 Critical 34.3 50 At. Critical 102 48.3 At. -50 -25 50 75 100 125 150 175 200 225 250 275 300 Degrees Centigrade. Vapor pressure curves of the simpler paraffine hydrocarbons. (W. O. Snelling in Hamor and Padgett's "Examination of Petroleum.") C n H 2n _ 2 and C n H 2n _ 4 . Petroleums from certain American fields con- tain no parafnnes, for examples, Coates 13 has shown that the lighter distillates of the oil from the Jennings, Louisiana, field consist exclu- sively of cyclic hydrocarbons of the C n H 2n series. 18 J. Am. Chem. 8oc. 28, 384 (1906). 18 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS The paraffine wax of commerce consists of a mixture of hydro- carbons of the paraffine series from about C 22 H 46 to C., 6 H 54 . The natural waxes of the ceresin type are evidently not normal paraffines but isomeric hydrocarbons probably identical with the amorphous wax of petroleum oils (see below). The number of hydrocarbons which have been isolated from petro- leum is very small. The old procedures, which supplied chemical lit- erature with a formidable array of names, empirical formulae and boiling-points, consisted in carefully fractioning a quantity of petro- leum and collecting fractions boiling between narrow limits. Formulae and names were then assigned to these fractions on the basis of com- bustion analyses and molecular weight determinations. The extremely careful work of Young shows how very difficult the separation of only two hydrocarbons may be when the diffierence in boiling-points is as much as 8, as in the case of n.pentane and isopentane. Young and Thomas 14 were able to separate n.pentane and isopentane in fairly pure condition only after thirteen fractional distillations through a very efficient dephlegmating column, and Young states that he was not able to isolate pure heptanes from light petroleum ether by fractional distillation. He regards the presence of n.hexane and isohexane in American and Russian petroleums as established, but the presence of other hexanes is open to question. Markownikow was able to isolate cyclohexane and methyl cyclopentane in fairly pure state from Baku oil by a combination of chemical treatments and fractional dis- tillation. 15 In the course of his work, Young showed that benzene and hexane form a constant boiling mixture boiling at 65-66. Although the distillation of two closely related hydrocarbons, for example, two members of the series C n H 2n+2 , as a constant boiling mixture is very improbable yet it is a possibility. Also owing to the fact that the boiling-points of a series of isomers may extend over a wide range, for example 22 in the case of the hexanes, it is evident that the prob- lem of isolating pure hydrocarbons from petroleum distillates is practi- cally a hopeless one, except in very simple cases as noted above. Paraffine hydrocarbons are produced in a variety of biological proc- esses. The best known example of this method of their production is methane, the name "marsh gas" referring to its formation in bogs where cellulose undergoes anaerobic fermentation. The amylobacteria 14 J. Am. Chem. Soc. 71, 440 (1897). "Aschan, Ber. SI, 1801 (1898). Markownikow, Ann. SOI. 154 (1898); Ber. SO 1532 (1897). THE PARAFFINES 19 of van Tiegham, 16 evolve methane from cellulose and in this fermenta- tion the other major products are carbon dioxide and the simple fatty acids. 17 Whether small proportions of other gaseous hydrocarbons are simultaneously produced has not been determined. As regards the theory of the biological origin of natural gas and petroleum, the for- mation of methane from buried cellulose material is capable of experi- mental duplication but this cannot yet be said of the higher homo- logues. Normal heptane has been obtained from the "petroleum nuts" Pittosporum resiniferum of the Philippines, 18 from the oleoresin of Pinus sabinmna and the wood turpentine of Pinus jeffreyi. 19 The higher paraffines occur in small quantity in many essential oils. Com- mercial rose oil contains sufficient paraffine or "stearoptene" to sepa- rate in large crystals, on chilling. This crude stearoptene has been separated into paraffines melting at 22 and 40 to 41. Heptacosane C 27 H 56 and hentriacontane C 31 H 64 occur in bees' wax 20 and the latter hydrocarbon also occurs in the resin of tobacco and the leaves of Gym- nema sylvestre, Olea europcea or the European olive, an African vine Morinda longiflora and Lippia scaberrina 21 According to Meyer and Soyka [Monatshefte, 84, 1159 (1913)], candelilla wax, used in making phonograph records, contains about 74 to 76 per cent of do-triacontane, C 32 H 66 . Small quantities of crystalline paraffine wax also occur in certain eucalyptus oils, e. g., Eucalyptus paludosa and Eucalyptus smithii 22 Pentatriacontane C 35 H 72 melting at 74.5-75 occurs in the leaves of Eridictyon calif or nicum 23 Pentacontane, C 50 H 102 , has been found in Lancashire coal. Altogether several tons of dark colored wax were found which after purification and decolorizing melted at 92.7- 93 and boiled at 420-422 under 15 mm. pressure. This hydro- carbon is the highest homologue of the paraffine series which has been found occuring naturally. 24 The Character and Probable Mode of Origin of Petroleums. The development of the petroleum industry had its beginnings almost coincident with the very rapid development of organic chem- "Compt. rend. 88, 205 (1879). "Lafar: Tech. Mykologie. Vol. III. 260 (1906). 18 Bacon, Philip J. Set. 4, 115 (1909). 19 Schorger, J. Ind. d Eng. Chem. 7, 24 (1915). "Schwalb, Ann. 235, 110 (1886). 21 Power & Tutin, J. Chem. Soc. 91, 1916 (1907) ; 93, 874 (1908). "Smith, J. Chem. Abs. 106, 399 (1914). "Power & Tutin, J. Chem. Soc. Abs. 90, 885 (1906). M Sinnatt & Barash, Inst. Min. Eng. 1919, Nov. 11. 20 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS istry. The petroleum industry was largely an American development but extensive research in organic chemistry was for long carried out almost exclusively in Europe, which is one reason for the comparative neglect of petroleum research. Also during these earlier years few American chemists had the facilities and time at their disposal neces- sary for research. Most American chemists of that period were an- alytical chemists, with the result that the earlier investigations of pe- troleum consisted in laboriously fractioning petroleum distillates and christening the various fractions normal undecane, normal dodecane, etc., etc. With the exception of the notable pioneer work of Mabery very little work of permanent value was done on petroleum in America during this long period. Young demonstrated 26 the presence of n . pentane and isopentane in petroleum, also the presence of n.hexane and isohexane and n. hep- tane and isoheptane, but considered the presence of isomeric hex- anes and heptanes as probable but not proven. The presence of cyclo-hexane, methyl-cyclopentane 27 and a limited number of homo- logues has been proven in the case of light naphtha from Baku oil. The isolation of a fraction having a constant boiling-point is not necessarily indicative of a pure single hydrocarbon. Two isomers, or two totally different hydrocarbons may have practically identical boiling-points. 28 Five of the known octanes boil within the range 114-118. Con- stant boiling mixtures are also known, the separate constituents of which may have quite different boiling-points. For example, pure n.hexane boils at 68.95 and benzene at 80.2, but a mixture of the two containing 10 per cent benzene boils at 69 and a mixture contain- ing 27.3 per cent benzene at 69.5. This behavior of benzene and hexane explains the fact that on nitrating petroleum fractions contain- ing benzene, the fraction yielding the most dinitrobenzene is that boiling at 60-65, not that boiling at 75 to 85. For a similar rea- son the fraction 90-100 contains more toluene, when this is a minor constituent, than the fraction distilling at 105-115. All petroleums which contain paraffine hydrocarbons as the chief constituents of their lighter fractions, as the Pennsylvania, Mid-Conti- nent, and light Texas crudes, show a rapidly increasing per cent of naphthenes as the boiling-point rises with successive fractions. In the light lubricating fractions the paraffine hydrocarbons, series CnH 911 2 , seldom exceeds three per cent and after their removal by chilling, re- 28 J. Chem. 8oc. 7S, 907 (1898). 27 Young, loc. cit.; Markownikow, Ber. SO, 1222 (1897). 28 Jackson & Young, J. Chem. Soc. 73, 926 (1898). THE PARAFFINES 21 suiting in the paraffine wax of commerce, the lubricating oil remaining is practically free from hydrocarbons of this class. Paraffine wax of commerce, melting ordinarily from 48 to 62 C, consists chiefly of a mixture of hydrocarbons of 23 to 28 carbon atoms. The melting- points and boiling-points of some of the definitely known paraffine hydrocarbons are given in the following table: BOILING-POINTS OF HYDROCARBONS OF THE PARAFFINE SERIES. Formula Name Boiling-Point "C. CHio n. butane 0.1 isobutane 10.5 n. pentane -f 36.3 isopentane 27.95 tetramethyl-methane 9.5 n. hexane 68.95 2 methyl pentane 62. 3 methyl pentane 64. 22 dimethyl butane 49.6 -49.7 2.3 " " 58.08 C T Hi4 n. heptane 98.2 -98.5 " 2 methyl hexane 89.9 -90.4 " 3 " " 90. -92. " trimethyl methane 95. -98. " 22 dimethyl pentane 78. " 2.4 " " 83. -84. 3.3 " " 86. -87. CsHis n. octane 125.8 " 2 methyl heptane 116. " 3 " " 117.6 MELTING-POINTS AND BOILING-POINTS OF HYDROCARBONS OF THE PARAFFINE SERIES. Formula Name Boiling-Point C Melting-Point C CsHw 4 methyl heptane 118. 2.4 dimethyl hexane 109.8-110. 2.5 " 109.2 3.4 " " 116. -116.2 diethyl-isopropyl methane 114. 2.2.3.3.tetramethyl butane 106. -107. +103. n.nonane 149.5 51. 3 methyl octane 142.4-143.4 4 ethyl heptane 138. -139. 2.5 dimethyl heptane 133. -137. 2.6 " " 132. ioHa n.decane 173. 32. 2.6 dimethyl octane 156.5-158. 2.7 " " 159.6 3.6 " 159.8-160.8 n.undecane 194.5 26.5 n.dodecane 214.5 12. 2.4.5.7 tetramethyl octane 208. -210. n.tridecane 234 6.2 n.tetradecane 252.5 +5.5 n.pentadecane 270.5 +10. n.hexadecane 287.5 + 19. -20. 7.8 dimethyl tetradecane 263. -265. below 30 22 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS MELTING-POINTS AND BOILING-POINTS OF HYDROCARBONS OF THE PARAFFINE SERIES. Formula Name Boiling-Point C Melting-Point C Ci 7 Hss n.heptadecane 303. +22.5 n.octadecane 317. 28. n.nonadecane 330. 32. n.eikosane 205. (15mm.) 36.7 n.heneikosane 215. (15mm.) 40.4 n.dokosane 224.5 (15mm.) 44.4 n.trikosane 234. (15mm.) 47.7 n.tetrakosane 240. (15mm.) 51.5 n.hexakosane ...... 56.6 Owing to the fact that paraffine wax does not crystallize readily in well formed crystals, even from crude petroleums which are free from asphaltic matter, until after distillation, it has been supposed that the crystalline paraffine is at least partly derived from a parent substance, "proto-paraffine," which breaks up during distillation and thereby yields the freely crystallizing paraffine wax. 29 Rakuzin 30 has shown that crude petroleums contain soft, medium and hard paraffines of crystalline structure. Marcusson 31 slowly distilled ceresine thereby decomposing it to a mixture of well crystallized paraffines and liquid hydrocarbons. The substance known to the refiners as amorphous wax and which gives much trouble to the wax manufacturer, may possibly be ordinary paraffine, whose crystallization is interfered with by col- loids, substances capable of gelatinizing on chilling or may in fact con- sist of paraffine derivatives, "proto-paraffines," for example, naph- thenes having very long paraffine side chains which on pyrolysis yield crystalline paraffine wax and an unsaturated naphthene or its polymers. A better method of separating or destroying amorphous wax is a problem of first importance to the refiners, but the real nature of amorphous wax has not been determined. The most definite informa- tion on this point is contained in a recent paper by Marcusson 32 who showed that amorphous wax is probably identical with ceresine and there is considerable evidence that ceresine consists of a mixture of branched chain or isoparaffines, a hypothesis first put forward by Za- loziecki. 33 Heretofore ceresine has generally been regarded as a mix- ture of the higher normal paraffine homologues. Marcusson com- pared the physical and chemical properties of a crystalline paraffine and a refined natural ceresine of practically identical melting points. Zaloziecki, Z. f. angew. Chem. 1888, 126. 30 J. Russ. Phys.-Chem. Soc. 1914, 1544; J. Chem. Soc. Abs. 106, 489 (1914). 81 Chem,. Ztg. 1915, 581, 613. "Ohem. Ztg. 1915, 613. **Z. angew. Chem. 1888, 126. THE PARAFFINES 23 Paraffine Ceresine Melting-point 56.5 -60.5 57.5 -60.1 Solidifying-point 59.2 59 Sp. Gr. at 15 0.885 0.917 Sp. Gr. at 60 0.781 0.798 Mol. Wt 330. 420. Paraffine is harder than ceresine in penetration tests, is markedly more soluble and is less viscous than ceresine at 70. Paraffine is only slightly attacked by fuming sulfuric acid, 33% S0 3 , at ordinary tem- peratures, but ceresine is energetically attacked. The action of nitric acid is also more energetic on ceresine. On dissolving paraffine in hot mineral oil and then cooling, the paraffine crystallizes out but with ceresine, under the same conditions, a vaseline-like deposit is obtained. Marcusson has examined the distillation products of ceresine and the oily product consists of a mixture of saturated hydrocarbons and de- fines of low molecular weight. No evidence of the presence of naph- thenes was obtained. The formation of branched chain hydrocarbons or so-called iso- paraffines may possibly be explained by the decomposition of montan wax, which as shown by Meyer and Brod 34 consists chiefly of an acid, C 28 H 56 2 , and a solid alcoholic wax. This acid of montan wax is not a normal chain fatty acid but a branched chain compound. Paraffine is formed during the distillation of asphalt base oils by the decomposition of the asphaltic matter. This is in accord with the observation that large amounts of crystalline paraffine are contained in shale oil, the wax not being present as such in the original shale but formed by the decomposition of the complex kerogen of the shale ; also the distillate obtained by the low temperature distillation of coals rich in volatile matter contains crystalline paraffine, which is not pres- ent as such in the original coal. In addition to the problem of separating simple mixtures of hydro- carbons by fractional distillation and the separation of paraffine wax by chilling and crystallizing, it should be noted that other special meth- ods must be resorted to, to isolate substances of a particular class from a particular petroleum. Petroleums contain varying proportions of the following classes of substances, all of which are very imperfectly known chemically: (1) Paraffine hydrocarbons, liquid and solid, series C n H 2n+ a. (2) Saturated monocyclic or napththene hydrocarbons, empirical formula cya*,. (3) Saturated polycyclic hydrocarbons, empirical formulae. v^n-H^n 2, OnHzn 4, OnHzn 6} CtC. 14 Monatshefte, 1913, 1153. 24 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS (4) Benzenoid hydrocarbons and derivatives. (5) Unsaturated hydrocarbons. (Present in distillates but probably not pres- ent in most crude petroleums.) (6) Asphaltic matter. (7) Sulfur derivatives. (8) Nitrogenous substances. (9) Organic acids ("naphthenic" acids, not of the fatty acid series). (10) Coloring matter and fluorescent substances (these substances may belong to other classes enumerated above). The majority of American petroleums yield gasolenes and kero- senes consisting chiefly of paraffine hydrocarbons. All American pe- troleums which contain large proportions of these lighter distillates, such as the light Appalachian, mid-Continent and northern Texas crudes, yield gasolene and kerosene of this character. Low boiling distillates consisting of cyclic hydrocarbons or naphthenes are usually derived from heavier crudes yielding very little of the lighter dis- tillates for example, the heavy California and the Jennings, Louis- iana crude from which Coates 35 has isolated dicyclic hydrocarbons C 10 H 18 to C 13 H 24 and the Russian and Galician oils from which cyclo- pentane, cyclohexane, and a series of their derivatives has been iso- lated. 36 The determination of the structure of these naphthenes, coupled with the difficulty of their isolation in a state of purity, is a task as difficult as any in organic chemistry, and it is doubtful if very much light will be thrown on their constitution until it is shown that chemical methods of utilization may lead to the extraction of greater profits, than are now obtained, though it is easily conceivable that the latter result cannot be arrived at without the former. Beilstein and Kurbatow 37 showed that the more volatile hydro- carbons of Russian petroleum possessed the empirical formula C n H 2n , exhibited none of the reactions of olefines and in their general chemical behavior resembled the hydrocarbons of the methane series. Two hy- drocarbons of the formula C 6 H 12 , one 38 boiling at 72 and the other 39 at 80 were isolated. Cyclohexane, prepared by Baeyer, proved iden- tical with the latter hydrocarbon from Russian petroleum and it was then shown that the isomeric hydrocarbon was methyl cyclopentane. Markownikow obtained evidence of the presence of cycloheptane in the fraction boiling at 115-120 of a Caucasian oil. With the exception of the bicyclic decahydronaphthalene isolated by Ross and Leather 40 88 J. Am. Chem. Soc. 28, 384 (1906). *Ann. 301, 154 (1898) ; 302, 37 (1898) ; 87 Ber. IS, 1818, 2028 (1880). 307, 342 (1899). "Kishner, J. Ritss. Phys.-Chem. Soc. 20, 118 (1890). "Markownikow, Ann. 802, 1 (1898). * Analyst 31, 284 (1906) ; This hydrocarbon is now made commercially by the catalytic hydrogenation of naphthalene. THE PARAFF1NES 25 from Borneo petroleum the structure of the higher boiling naphthenes is largely a matter of conjecture. PHYSICAL PROPERTIES OF SOME SATURATED CYCLIC HYDROCARBONS. Empirical Formula C^Hs 04x17.0x13 C^.CH, r.C 2 H5 C 3 H 3 .(CH3) S Name Cyclopropane tt Cyclobutane ** Methyl cyclopropane 41 *Cyclopentane ** Methyl cyclobutane ** 1.1 dimethyl cyclopropane 4 * 'Cyclohexane 4T *Methyl cyclopentane ** Ethyl cyclobutane *" 1.2.3. trimethyl cyclopropane 1.1.2. " Cycloheptane (suberane) M *Methyl cyclohexane 5 * 1.1 dimethyl cyclopentane ra 1.2 " " i-1.3 " Cyclo-octane M Ethyl cyclohexane M M 1 . 1 dimethyl cyclohexane M 1.2 dimethyl cyclohexane n "Ladenburg & Kriigel, Per. S2, 1821 (1899). Willstatter & Bruce, Ber. 40, 3979 (1907). Demjanow, Ber. 28, 21 (1895). ** Markownikow, Ann. 327, 59 (1903). * 5 Perkin & Colman, J. CJiem. Soc. 53, 201 (1888). "Gustavson & Popper, J. pr. Chem. (2), 58, 458 (1898). "Perkin & Freer, J. Chem. Soc. 53, 203 (1895). "Zelinsky & Gutt, Ber. 41, 2431 (1908). Zelinsky & Zelikow, Ber. S),, 2857 (1901). M Willstatter & Kametaka, Ber. 41, 1480 (1908). 61 Sabatier & Senderens, Compt. rend. 132, 566 (1901). 62 Kishner, CJiem. Cent. 1908, II, 1860. "Zelinsky & Rudsky, Ber. 2.9, 405 (1896). M Willstatter & Veraguth, Ber. 40, 968 (1907). 65 Kursanoff, Ber. 32, 2973 (1899). M Crossley & Renouf, J. Chem. Soc. 87, 1498 (1905). "Sabatier & Mailhe, Compt. rend. 141, 20 (1905). C 8 H 1 Boeing-Point C Sp. Gr. 35. n 11. - 12. 0.7038^0 4. - 5. 49. 0.7635^ 39. - 42. 21. 81. 0.7934 ^p- 70. - 71. **10 72.2- 72.5 0.7540^ 65. - 67. 0.6946^ 57. - 59. 0.6832^- 118. 0.8275^1 18 100. -101. 0.7662^- 4 88. 0.7547$! 92. - 93. 100 0.7581^ 94 93. 0.7410|L 145.3-146.3 M.-Pt. 11.5 0-850 51 132. -133. 0.7913^ 120. n 124. 0.8002^ 26 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS PHYSICAL PROPERTIES OF SOME SATURATED CYCLIC HYDROCARBONS. Name Empirical Formula Boiling-Point C Sp. Gr. 1.3 dimethyl cyclohexane " " 118. 0.7869 ^o 1.4 w 119. 0.7861 5! 1 methyl-3-ethyl cyclopentane B8 C B H 8 R CH 3 + CH 2 = CH.R 86 The destructive distillation of bituminous shales, lignites and peat yields distillates containing paraffine hydrocarbons; and paraffine wax has for many years been manufactured from the distillates of shale in Scotland. By the distillation of ordinary bituminous coking coal at low temperatures a distillate rich in paraffine wax is obtained. These solid paraffines like those from petroleum are normal hydrocarbons of 24 to 29 carbon atoms. 87 Effects of Heat on Non-Benzenoid Hydrocarbons. The changes brought about by heating non-benzenoid hydrocarbons have long been of industrial interest and importance, particularly in 'the manufacture of oil gas, carburetted water gas, the pyrolysis of pe- troleum oils for the manufacture of kerosene and more recently gaso- lene or motor fuel from heavier hydrocarbons. These processes are problems of technology or engineering, rather than chemistry, but more recently a desire to know more concerning the chemical reactions in- volved and their relationships has been indicated by the character of many of the published researches. The two fundamental reactions which take place when hydrocar- bons are heated to the decomposition point are, first, the rupture of the carbon-to-carbon structure and second, the dissociation of hydrogen from carbon. These two reactions probably occur simultaneously at- tended by a sequence of other reactions, but special catalysts may greatly accelerate one or the other type of reaction, for example, nickel, palladium or platinum may cause dissociation of hydrogen without alteration of the carbon structure, as in the conversion of cyclohexane to benzene in the presence of nickel at 250, or the complete rearrange- ment and splitting of hydrocarbons by gentle heating in the presence of anhydrous aluminum chloride, in which case methane but not hy- drogen is evolved. The earlier technical investigations of the pyrolysis of hydrocar- defines of this type are unstable and rearrange, cf. pp. (150, 151). "Glund, Ber. 59, 1039 (1919). THE PARAFFINES 35 bons centered upon coal tar, benzene, naphthalene and their deriva- tives. In 1866-7, Berthelot published a series of important researches, 88 and stated that at a "dull red heat" equilibrium was established be- tween ethylene, hydrogen and ethane. He discovered a series of con- densations of acetylene; that in the presence of coke, acetylene, at the "temperature at which glass softens" is decomposed almost wholly to hydrogen and carbon; acetylene and ethylene yield a condensation product isomeric, or identical with crotonylene, and acetylene and ben- zene gave naphthalene. Benzene passed through a porcelain tube gave diphenyl, chrysene and a resinous substance, but no anthracene or naphthalene. Toluene gave benzene, unchanged toluene, and large pro- portions of naphthalene. Xylene gave toluene as the principal product. Berthelot's view that acetylene was the parent substance of the ben- zenoid hydrocarbons was vigorously disputed by Thorpe and Young, 89 Armstrong and Miller 90 and Haber 91 who considered that hydrogen or methane were first formed, the residues then condensing or undergoing still further decomposition : 2 G 6 H 6 > C 12 H 10 (diphenyl) + H 2 C 6 H 14 (hexane) C 5 H 10 (amylene) + CH 4 They pointed out that usually acetylene cannot be detected among the products of pyrolysis. Bone and Coward 92 have made a careful study of the thermal decomposition of methane, ethane, ethylene and acetylene and concluded that Berthelot's theory of the attainment of equilibrium between dissociation and recombination of these hydro- carbons is not borne out by the experimental evidence. Their results show: (1) Methane is exceedingly stable. It decomposes almost exclusively into hydrogen and carbon and this decomposition, though rever- sible, is mainly a surface phenomenon, at least at moderate tem- peratures. (2) Acetylene polymerizes at comparatively low temperatures, the optimum temperature range for this polymerization being 600- 700. Acetylene being formed from ethylene, condensation prod- ucts of acetylene will be found among the products whenever ethylene is a primary product of the pyrolysis of hydrocarbons. **Compt. rend. 62, 905, 947 (1866) ; 63, 788, 834 (1866) ; Bull. Soc. Chim, (2) 7, 217 (1867). M Proc. Roy. Soc. 19, 370; 20, 488; 21, 184 (1873). Chem. News 49, 285; Soc. J,9, 74 (1886). 91 J. Gasbel, 39, 377, 395. 435, 452, 799; Ber. 29, 2691 (1896). 82 Soc. 93, 1197 (1908). 36 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS (3) Ethylene and acetylene combine with hydrogen at moderate tem- peratures to form ethane. Whitaker and Leslie 93 obtained evi- dence of hydrogenation at 620 when hydrogen was introduced with oil in an experimental apparatus for making oil gas. These authors also call attention to the fact that in decomposing oil to gaseous products, equilibrium, or rather the ultimate composition which a given temperature tends to produce, is seldom attained, even in apparatus of industrial size, owing to the short period of time, during which the hydrocarbons are subjected to the par- ticular temperature of the operation. One reason for this un- doubtedly lies in the fact that some of the reactions taking place in such systems are strongly endothermic, for example, C 2 H 6 > C 2 H 4 + H 2 = 31,270 calories and such reactions can be maintained only by the absorption of a large supply of energy. 9 * It is well known that the velocities of chemical changes are greatly affected by relatively small changes in temperature. It is, therefore, readily understood that small differences of operating temperature may cause very great differences in the character of the pyrolytic products, a fact apparently first appreciated in industrial operations by Hall. Generally speaking, the temperatures employed for obtaining mo- tor fuel are within the range 410-500 and Rittman, Button & Dean 95 consider that the maximum yield of aromatic hydrocarbons (from pe- troleum oils) is obtained within the range 650-700. Ipatiev &6 states that at 600-700 hexane and cyclohexane yield defines and other hy- drocarbons, but no benzenoid hydrocarbons. Methyl cyclopentane was found among the products. Norton and Andrews 97 found that at 550 hexane was not decomposed and was very slightly affected at 600 but at 700 decomposition with formation of gas, methane and ethyl- ene, propylene, butylene, amylene, hexylene and butadiene but no benzene. Iso-hexane and n . pentane show approximately the same sta- bility and at 700 yield gas and a series of defines. Benzene appears among the products of reaction only at higher temperatures. Thus Haber obtained benzene from hexane by heating to 800 98 and Wor- stall and Burwell obtained it from heptane and octane at 900." J. Ind. & Eng. Chem. 8, 593, 684 (1916). M Lomax, Dunstan & Thole, J. Inst. Petr. Tectin. 3, 76 (1916). 95 U. S. Bur. Mines Bull. 114, Washington (1916). M Bcr. It',, 1984, 2978 (1911). "Am. Chem. J. 8, 1 (1886). 98 Loc. cit. 99 Am. Chem. J. 19, 815 (1897). THE PARAFFINES 37 Benzene and its simple homologues had been found in the liquid ccndensate obtained by compressing oil gas. 100 Armstrong and Miller made a careful study of this liquid condensate from oil gas and iden- tified propylene, amylene, hexylene, heptylene, crotonylene, isoal- lylethylene, benzene, toluene, xylenes, mesitylene, pseudo-cumene and naphthalene. In 1878, a number of processes were described 101 which sought to manufacture benzene hydrocarbons from Russian petroleum by passing the oil through red-hot tubes packed with various materials (the function of which was not evident) . None of these processes were industrially successful. Nikiforoffs' process was apparently a develop- ment from the well known Pintsch gas process, the oil being first de- composed or vaporized at 525-550 and then passed through retorts, similar to the older type of Pintsch gas retort, at 700-1200 under a pressure of about two atmospheres. No further important work on the manufacture of benzene hydrocarbons from petroleum oils by the action of heat, in the absence of catalysts, was made until the recent war period when Hall, working in England, and Rittman and his co- workers in the United States, developed processes, which were operated industrially. Hall decomposes oil at 550-600 and under a pressure of about 70 pounds per square inch when motor fuel is the desired prod- uct and for benzene and toluene the operating temperature is 750 and the pressures 100 to 110 pounds per square inch. 102 A noteworthy me- chanical feature of the Hall process is very rapid passage of the oil and vapors through the heated tubes, which minimizes the deposition of carbon. Rittman employed a temperature of 700 and a pressure of 150 pounds per square inch. In connection with this work, which probably should be regarded as a war time industry, at least so far as the manufacture of benzene and toluene from petroleum is concerned, much valuable experimental work was done. Commercial gas oil, specific gravity 0.817 at 15.5 and boiling at 200-350, in the Rittman apparatus gave a maximum yield of toluene, 3.1 per cent by volume, at 650. The maximum yield of benzene, 4.4 per cent by volume was obtained at 800. The maximum yield of xylene was 1.9 per cent at 750. 103 Other conditions being equal, higher yields of aromatic hydro- carbons are obtained from petroleum containing relatively large pro- i Armstrong and Miller, J. Chem. Soc. 49, 74 (1886) ; Williams, Chem. News 49, 197 (1884). 101 Letny, Bcr. 11, 1210 (1878) ; Liebermann & Burg, Ber. 11, 723 (1878) ; Salzmann & Wichelhaus, Ber. 11, 1431 (1878). 592 TT. S. Pat. 1.175.909; Brit. Pat. 24,491 (1913); 437 (1914); 2948 (1914); 7282 (1914); 12,962 (1914); 1594 (1915); U. S. Pat. 1,194,289; 1,175,910. 103 Egloff, Met. & Chem Eng. 16, 492 (1917). 38 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS portions of benzene and naphthalene derivatives and cyclohexane de- rivatives, such as Borneo petroleum. The oil from the deeper strata of the Kotei field in Borneo was used in England during the war for the manufacture of benzene and toluene. Jones and Wooton first called attention to the unusual character of this oil and more recently Chavanne and Simon 104 have examined the gasoline fraction of this oil and state that they have identified cyclopentane, methylcyclopen- tane, cyclohexane, a dimethylcyclopentane, methylcyclohexane and dimethylcyclohexane. During the war gasoline from this petroleum was sent to England where benzene and toluene were made from it in a fair degree of purity. This gasoline contained about 40 per cent of aromatic hydrocarbons of which about 7 per cent was benzene, 14 per cent toluene, 15 per cent xylenes and 4 per 'cent higher homologues. 105 Brooks and Humphrey 106 found small quantities of benzene and tolu- ene in gasoline made by distilling heavy Oklahoma oil at the relatively low temperature of 420 and a pressure of 100 pounds per square inch. Small yields of aromatic hydrocarbons were also obtained from heavy high-boiling petroleums by heating with anhydrous aluminum chloride and since the temperature employed in the first method is considerably below that at which benzene has been observed to be formed from paraffines or na'phthenes, they conclude that high boiling benzene de- rivatives are present in the original oil, benzene being obtained by their splitting or "cracking." It was apparent from much of the early work on pyrolysis that the character of the products obtained was not solely a function of the tem- perature employed but also of the time or duration of the heating and also the presence or absence of various substances acting catalytically upon the decomposition, either hydrogen dissociation or splitting of the carbon structure, or affecting one or more of the secondary re- actions, for example polymerization of the olefines which are formed. Before discussing the effect of catlysts the results of pyrolysis at mod- erate temperatures will be noted. One of the most conspicuous differences in the results of low tem- perature decomposition is the greatly decreased yield of gas. Exact comparisons are difficult to make on account of variable time factors, different distribution and character of the heated surfaces and the like. Hall states that in the industrial tube type of apparatus developed by him a change of operating temperature from 540 to 580 results in an 104 Compt. rend. 1919, 285. 106 Kewley, Chem. Tr. J. 1921, 380. 109 J. Am. Chem. Soc. 88, 393 (1916). THE PARAFFINES 39 increase of 50 per cent in the quantity of gas obtained. In a small ex- perimental pressure still Brooks, Padgett and Humphrey 107 found, when distilling 85 per cent of the oil used [heavy Oklahoma gas oil] , under pressure, that at 50 pounds pressure and a mean temperature of 410, 24.8 liters of gas were formed per liter of distillate; at 150 pounds pressure and a mean temperature of 425, 58 liters of gas per liter of distillate were produced. The relative area of heated surface (iron) in this small apparatus was quite large, as compared with oil distilling apparatus of industrial dimensions of the Burton type, but the results are indicative of the large difference in gas yield resulting from a comparatively slight temperature change. The effect of in- creased pressure should, per se, decrease the gas yield by polymerizing the olefines. That higher temperatures give large proportions of ole- fines in the^gas is indicated by the following table: COMPOSITION OF OIL GAS MADE IN TUBES MAINTAINED AT DEFINITE TEM PERAT URES , 108 Temperature, deg. C 600. 650. 700. 730. Pressure, Ib 57. 72. 83. 95. Ethylene, per cent 19.3 19.0 17.7 17.5 Propylene, per cent 28.0 28.4 23.9 20.0 Higher olefines, per cent 3.2 4.2 3.5 3.1 Total olefines, per cent 50.5 51.6 45.1 40.6 GASES FROM CRACKING DISTILLATIONS UNDER lOO-Ls. PRESSURE. From Jennings Crude 1 2 3 340 415 422 Temperature in still Per Cent Per Cent Per Cent CO 2 12 0.5 0.0 CO 1.2 0.5 1.3 Illuminants 15.4 15.3 13.0 Hydrogen 0.0 4.0 4.4 Saturated Hydrocarbons 81.5 79.7 81.3 From Paraffine 417 432 437 Temperature in still Per Cent Per Cent Per Cent CO 2 0.0 0.0 0.0 CO 0.0 0.0 0.0 Illuminants 25.4 37.0 33.5 Hydrogen 0.3 0.9 3.0 Saturated Hydrocarbons 74.3 62.1 63.5 Analyses of oil gas are usually reported in terms of total olefines, or illuminants, hydrogen and methane. Accurate analyses, with respect to methane, ethane, propane and other hydrocarbons, made by the method of fractional distillation at low temperatures have not been re- 107 J. Frankl. Itist. 180, 653 (1915). 108 Hall type of apparatus, industrial size. 40 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS ported. The relative proportions of ethylene, propylene and other defines in oil gas made at different temperatures in a commercial size Pintsch gas apparatus is given in the following table: PER CENT ETHYLENE AND PROPYLENE IN OIL GAS. Higher Temp. Olefines Per Deg. C Per Cent Cent 805-650 ............................... 1.4 18.6 660-535 .............................. 1.6 19.0 635-535 ............................... 2.4 22.4 625-535 ............................... 2.6 22.6 615-425 ............................... 3.8 25.7 C 2 H 4 Per Cent 16.3 18.3 12.5 13.7 12.0 Total Olefmes Per Cent 36.3 38.9 37.3 38.5 41.5 The composition with respect to olefines of gas made at definite temperatures in a large industrial size apparatus of the Hall type is as follows: 109 PER CENT OIL GASIFIED IN HALL TYPE APPARATUS AT DIFFERENT TEMPERATURES. Per Cent Temperature Per Cent Ethylene and Deg. C 605 625 645 665 Gas 17.7 26.6 37.6 40.0 685 , 40.8(?) 705 48.7 725 . 66.6 Propylene 47.9 46.1 44.9 43.7 42.6 39.5 38.5 Typical analyses of commercial gases are of interest particularly as regards the relative proportions of methane, ethane, hydrogen and illuminants. 110 AVERAGE COMPOSITION OF COMMERCIAL GASES. Coal gas Carburetted water gas Pintsch gas Blau gas All oil water gas 7.0 Oil gas 31.3 Blue water gas Producer gas (coal) . . . Producer gas (coke) . . . Blast furnace gas Wood gas (pine) 10.6 v v ' ( ^00 Oil gas, Dayton process" 1 14.7 5.6 1.7 7.8 6.1 .. 63.2 ... | 390 "Brooks, Chem. & Met. Eng. 2%, April 7, 1920. 110 Rogers' Industrial Chemistry Ed. 2. Fulweiler, p. 474. 111 Binnall, Gas Age 47, 47 (1921). This process depends upon the partial com- bustion of the oil sufficient to raise sufficient heat to gasify the remainder. About 4 gallons of oil are required to make 1000 cubic feet of gas of 450 B. T. U. The per cent of nitrogen is naturally high. Ilium. CO H 2 CH 4 C 2 H 6 C0 2 O 2 N 2 Cdl. % % %%%%%% Pr. B.T.U. 4.0 8.5 49.8 29.5 3.2 1.6 .4 3.2 16.1 622 13.3 30.4 37.7 10.0 3.2 3.0 .4 2.1 22.1 643 30.0 .1 13.2 45.0 9.0 .2 .0 1.6 43.0 1276 51.9 .1 2.7 44.1 .0 .0 .0 1.2 48.2 1704 7.0 9.2 39.8 34.6 2.6 .2 6.6 19.7 680 31.3 2.4 13.5 46.5 3.0 .3 .0 1.1 38.0 1320 .0 40.9 50.8 .2 .0 3.4 .9 3.5 299 .2 17.6 10.4 6.3 .0 7.3 .7 58.1 ... 161 .0 25.3 13.2 .4 .0 5.4 .6 55.2 137 .0 26.5 3.5 .2 12.8 .1 56.9 100 10.6 27.1 32.7 21.5 4.9 .4 2.6 607 THE PARAFFINES 41 Although data obtained in making oil gas on a small experimental scale have no close industrial parallel, the experimental results of Whitaker and Rittman 112 are of interest as indicating the very marked effect of variations of temperature and pressure. OIL GAS EXPERIMENTS OF WHITAKER & RITTMAN.* Pressure Temp. Ib. per Gas Carbon Tar CH* CA H 2 Ilium. C sq. in. Liters Grams c.c. Liters Liters Liters Liters 650 15. 135 3 163 45.5 13.8 12.1 58.8 650 45. 145 8 133 65.2 16.7 13.1 44.3 750 0.75 146 1 153 .... 18.3 82.0 750 15. 206 18 80 84.5 Vo.is 39.6 63.0 750 45. 194 26 87 110.0 11.8 33.9 30.1 900 0.75 235 12 58 63.4 trace 48.8 110.0 900 15.0 382 115 11 178.1 trace 148.2 50.0 900 45.0 310 165 9 128.9 none 155.0 A 15.5 *40<* Oil used. In a later paper Whitaker and Alexander 113 showed that under the same experimental conditions the composition of the gas produced varies with the rate of oil feed, within rather wide limits, and that even at comparatively slow rates of oil feed equilibrium is not reached. Thus it has been shown that at 1200 hydrogen is in equilibrium with carbon and about 0.3 per cent methane, but Whitaker and Alexander find 6 to 10 per cent methane in their most slowly conducted experi- ments 114 and they emphasize the fact that equilibrium compositions are not obtained in gas making practice and that it would be im- practical to run an oil gas generator at such rates of oil feed as would even approximate equilibrium conditions. Zanetti 115 obtained typical oil gas by decomposing the propane fraction of natural gas gasoline at 750, obtaining ethylene, propylene, butylene and small quantities of liquid hydrocarbons and tars. In view of the fact that the coking of coal at low temperatures yields a distillate containing paraffine wax, naphthenes and olefines and resembling crude shale oil in its general character the coking of coal at higher temperatures with the formation of coal gas and typical coal tars should be regarded as essentially paralleling the high tem- perature pyrolysis of mineral oils, in contact with coke or carbon. 112 J. Ind. d Eng. Chem. 6, 479 (1914). 113 J. Ind. & Eng. Chem. 7, 484 (1915). 14 The commercial manufacture of hydrogen by heating methane or other hydro- carbons to 1200-1300 has been proposed, with various modifications, for example see Uhlinger, U. S. Pat. 1,363,488. 116 J. Ind. & Eng. Chem. 8, 674 (1916). 42 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS As regards the liquid products of pyrolysis of high boiling hydro- carbons at moderate temperatures, low temperature, pressure and slow operation favor the formation of saturated low boiling hydrocarbons. Hall and others have called attention to the relatively large yields of defines and diolefines obtained at 550-600. Such distillates are said to be "highly cracked," absorb oxygen readily, a very general property of diolefines (see pages 212, 216) forming a resinous oxidation product which is often noticed as a sticky film when such oil is permitted to evaporate. Such distillates, containing diolefines, react energetically with sulphuric acid forming tars. Slow distillation under pressure evidently polymerizes the defines since hydrogenation of hydrocarbons at 400-450 in the absence of catalysts and under moderate pressures has not been observed. Although Bergius has hydrogenated fatty oils in the ar>sence of finely divided metallic catalysts by heating with hy- drogen under 30 atmospheres, 116 no hydrogenation of unsaturated pe- troleum hydrocarbons could be detected by Brooks 117 on heating at 196 for 30 hours under a hydrogen pressure of 3000 pounds per square inch. However, Ipatiev 118 noted evidence of hydrogenation at higher temperatures and under pressures up to 340 atmospheres. The low boiling hydrocarbons produced at moderate temperatures are mainly normal saturated paraffines as has been shown by Hum- phrey in the case of a distillate made from the heavy residues of Oklahoma petroleum by distilling at 400-420 and 100 pounds pres- sure. 119 The presence of small quantities of benzene and its homo- logues in such distillates has been noted. Among the diolefines, which have been identified in the low boiling fractions butadiene and isoprene have been repeatedly noted. The yield or relative proportions of these hydrocarbons obtainable in this way is quite small. Engler and Staudinger, 120 however, have patented the manufacture of these conjugated diolefines by the thermal decom- position of mineral oils. Pyrolysis under reduced pressure increases the proportion of unsaturated hydrocarbons, at least among the gaseous products. 121 The polymerization of defines by heating under pressure has been frequently observed. Ethylene, the most stable known olefme, is polymerized in the presence of iron at 380-400 and 70 atmospheres 16 Z. angew. Chem. 191^, 522. 7 J. Frank. Inst. 1915, 658. 8 Ber. 37, 2961 (1904). J. Ind. d Eng. Chem, 7, 180 (1915). 20 German Pat. 265,172 (1912). 121 Whitaker & Rittman loc. cit. THE PARAFFINES 43 pressure, a complex mixture of hydrocarbons being formed. 122 The polymerization of conjugated diolefines at moderate temperature and pressures has been applied to the synthesis of rubber (see page 000) and Semmler 123 has condensed isoprene with limonene and other ter- penes at 275 to form new sesquiterpenes of the empirical formula C 15 H 24 . Lebedev 124 has polymerized allene by heating in glass at 140 obtaining 5% dimeride, 15% trimeride and 80% of more highly polymerized material. Diallyl is very slowly polymerized at 250 to a dimeride and a gummy residue. At 150 2 : 4 hexadine yields chiefly the dimeride. The polymerization of defines is markedly catalyzed by many substances. Gurwitsch 125 polymerized amylene by fuller's earth at ordinary temperature and Hall polymerized the resin-forming constituents (diolefines) contained in light pyrolytic gasoline distillate, by passing the hot vapors through a column of fuller's earth. Fuller's earth, kaolin and alumina are said to slightly increase the yield of low- boiling hydrocarbons. The effect of nickel in a finely divided condition in bringing about equilibrium conditions between unsaturated hydrocarbons, hydrogen and saturated hydrocarbons, has led to quite changed conceptions re- garding the stability of hydrocarbons. The earlier work had to do almost exclusively with the formation of saturated hydrocarbons, with yields which were practically quantitative. The reversible nature of the reaction was not clearly recognized until Sabatier and Senderens showed that cyclohexane was converted into benzene in the presence of finely divided nickel at 270-280. Zelinsky 126 showed that cyclo- hexane and methylcyclohexane are reduced to benzene and toluene respectively, together with free hydrogen, by heating in the presence of finely divided palladium. The reaction is appreciable at about 190, and within the range 200-300, the equilibrium mixture contains very large proportions of benzene. 127 No dihydro or tetrahydro derivatives were found among the reaction products. Hexane, cyclopentane and methylcyclopentane are more stable, and do not yield free hydrogen appreciably below 300. The extraordinary stability of methylcyclo- pentane as compared with cyclohexane is shown by later experiments of Zelinsky, in which a mixture of methylcyclopentane and cyclohexane *Ipatiev, J. Rusa. S8, I, 63 (1906). 113 Ber. 47, 2068, 2252 (1914). m J. S. C. I. 1914, 1224. 118 J. Rus8. 47, 827 (1915). 1M J. Rusa. Phy*.-Clim. Soc. J,S, 1220 (1911) ; Ber. 44, 3121 (1911). 127 Tausz & Putnoky, Ber. 52, 1573 (1919), state that in the presence of palladium black the formation of benzene from cyclohexane is practically quantitative at 270- 300. They confirm the absence of cyclohexane in Pennsylvania gasoline by testing for the formation of benzene under these conditions. 44 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS were passed over palladium black at 300, until no further hydrogen was evolved. The cyclohexane was converted to benzene, thus offer- ing a convenient and easy method of separating these two hydrocar- bons. 128 That these results are brought about by the catalysts is indicated by many observations, for example, Ipatiev 129 had shown previously, that benzene is not formed from cyclohexane or hexane on passing the vapors through an iron tube heated to 650-700 C. At considerably lower temperatures, cyclohexane is readily formed from benzene, the reaction being very rapid in the presence of finely divided nickel and hydrogen at 160. The catalytic effect of iron, copper and aluminum on the dissociation or addition of hydrogen is very slight. Whether or not the iron surface of pressure stills and similar apparatus have any catalytic effect on the pyrolytic changes effected by heating pe- troleum oils is not certain, but since very finely divided iron has only a very slight effect, the catalytic effect of the iron or steel surfaces of industrial apparatus is probably negligible. It has frequently been proposed to insert catalysts into pressure stills and similar apparatus with the object of hydrogenating the de- fines which distillates made in this way normally contain, but these methods have had no technical success. Nickel, the most active cat- alyst of this type, is very quickly covered with coke and thereby ren- dered inactive. 130 Sabatier and Mailhe proposed to remove the carbon from the metal catalyst by heating in a current of steam. 131 The lower temperatures at which the reaction of steam and carbon becomes appreciable have not been determined and this doubtless varies considerably with different forms of carbon. Bergius has converted carbon and water to hydrogen and C0 2 by heating at 300 and 150 atmospheres pressure for 20 days. Although water gas has been manu- factured for many years, high temperatures are always employed since it has long been known that low temperatures favor the formation of CO 2 in the gas equilibrium CO + H 2 + C0 2 + H 2 . 132 A number of patents have described the decomposition of heavy oils in the presence of steam and one patentee claims that iron acts as a catalyst in this steam-hydrocarbon mixture. 133 This process has been carried out on a ls *Ber. 45, 678 (1912). 128 Ber. 44, 2987 (1911). 130 The manufacture of hydrogen from methane in the presence of nickel at 700 as proposed in the Badische process, French Pat. 463,114 (1913), is undoubtedly sub- ject to this difficulty. l U. S. Pat. 1,152,765 (1915); TJ. S. Pat. 1,124,333 (1915). 182 Taylor & Rideal : Catalysis, p. 158. "Noad & Townsend, Brit. Pat. 113,675 (1908). THE PARAFFINES 45 fairly large scale, the tubes or retorts being packed with .iron turnings and a temperature of about 600 maintained. Greenstreet claims that the presence of steam in the zone of decomposition prevents the depo- sition of carbon or reacts with the carbon to form carbon monoxide and hydrogen, the hydrogen being supposed to be taken up by the un- saturated hydrocarbons. In the presence of nickel, Sabatier observed the reaction of steam and carbon to C0 2 and hydrogen at 500. The only catalytic process which has shown great industrial prom- ise is of an altogether different type from the catalysts discussed in the foregoing paragraphs. Abel and also Friedel and Crafts described the decomposition of petroleum hydrocarbons by heating with anhydrous aluminum chloride. 134 Gustavson noted a similar behavior with alu- minum bromide. Heusler noted that unsaturated hydrocarbons are polymerized by aluminum chloride and also that sulfur derivatives are decomposed and the sulfur removed. 135 Aschan also noted the polymerization of olefines in the presence of this reagent and Engler observed that amylene, heated with anhydrous aluminum chloride, 136 yielded a mixture of polymers resembling natural lubricating oil. A number of patents have been recently issued to McAfee, 137 who has determined the technical refinements necessary in the utilization of this catalyst. In addition to a little gas and a mixture of volatile saturated hydrocarbons including an excellent grade of gasoline, a heavy viscous residue is formed, which contains the greater part of the aluminum chloride. This material is very readily carbonized when heated, and the recovery of aluminum chloride from these residues is the really difficult part of the problem, at least from a technical and economical standpoint. The effect of anhydrous zinc chloride and anhydrous ferric chloride is similar but much less effective. Synthesis of the Paraffines. The reduction of alky! halides (chlorides, bromides or iodides) by nascent hydrogen has been accomplished in a number of ways. The method of Gladstone and Tribe 138 of reducing alkyl iodides in alcohol solution by 'the copper-zinc couple has been most fruitful. Many of these e'arlier methods were discovered in the attempt to isolate the so- called radicals; for example, Frankland showed that heating the sim- pler alkyl iodides with water and zinc gave the corresponding hydro- "* Friedel & Crafts, Compt. rend. 100, 692; Gustavson, J. prakt. cJiem. 31, 161; Egloff & Moore, Met. & Chem. Eng. 15, 67, 340 (1916). 135 Brit. Pat. 4769 (1877). 138 Z. angew. Chem. 9, 288, 318 (1893) ; Ber. J,2, 4613 (1909). 137 U. S. Patent. 1,099,096; 1,127,465 and 1,144,304. 188 Ber. 6, 202, 454, 1136, 1873; J, Chem. Soc. 45, 154 (1884). 46 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS carbons, possibly through the intermediate formation of zinc dialkyls. When pure zinc dialkyls are treated with water very energetic decom- position occurs with the formation of a hydrocarbon and zinc hydrox- ide. Zn(C 2 H 5 ) 2 + 2H 2 Zn(OH) 2 + 2C 2 H 6 This method has been displaced by the well-known Grignard reaction, the simpler alkyl halides readily yielding alkyl magnesium halides which are quantitatively decomposed by water to give hydrocarbons. C 2 H 5 Br + Mg + (C 2 H 5 ) 2 C 2 H 5 MgBr. (C 2 H 5 ) 2 OH C 2 H 5 MgBr . (C 2 H 5 ) 2 + H 2 -> Mg + C 2 H 6 + (C 2 H 5 ) 2 \, Ammonia or an amine may be employed instead of water to decompose the magnesium complex. It should be pointed out, however, that an- other reaction takes place with magnesium and alkyl halides which, though a very subordinate reaction in the case of the simpler alkyls, be- comes the principal result with halogen derivatives containing six or more carbon atoms. 139 Thus, like the condensation of propyl bromide by metallic sodium to form n.hexane, propyl bromide and magnesium, in ether, yields a small amount of n.hexane as expressed by the re- action, C 3 H 7 Br+Mg >C 3 H 7 MgBr C 3 H 7 MgBr + C 3 H 7 Br > MgBr 2 -f C 6 H 14 This reaction is an admirable method of synthesis within certain limits. 140 Thus in the terpene series halides such as bornyl chloride react so slowly with magnesium that the Grignard reactions are of practically no value for halogen derivatives of this class. Hydrocar- bons of an odd number of carbon atoms may be synthesized by a slight modification of the method, for example, C 3 H 7 MgBr + C 4 H 9 Br > MgBr 2 + C 7 H 16 C 4 H 9 MgBr + C.H^Br > MgBr 2 + C H 20 A modification of the above method has proven most satisfactory for the preparation of tetramethyl methane, the magnesium complex (CH 3 ) 3 C.MgI being treated with methyl sulfate. 141 189 Grignard & Tissier, Compt. rend. 132, 835 (1901). 140 Alkyl groups may be introduced in the benzene ring by treating magnesium phenyl bromide with propyl or allyl bromide. Tiffeneau, Compt. rend. 1^5, 437 (1907) ; Kling, Compt. rend. 1-37, 756 (1903) ; Brit. Pat. 122, 630 (1919). iFerrario & Fogetti, GO&Z. CMm. Ital. 38, II, 630 (1908). THE PARAFFINES 47 The Grignard reaction has a wide range of usefulness in building up substances having the carbon atom structures of the hydrocarbons desired, the hydrocarbons themselves then being obtained by other methods, for example, C 2 H 5 RCHO + C 2 H 5 MgX- ~>RCH OH RCOCH 3 + C 2 H 5 MgX RCOOCH 3 + 2C 2 H 5 MgX CH 2 RMgX+ ]__ >0 >RCH 2 CH 2 OH CH, the alcohols thus obtained being converted to hydrocarbons by means of the corresponding iodide and reduction, or by decomposing the alco- hols or corresponding halides to olefines and hydrogenating the latter. Hydriodic acid has been widely employed for the purpose of ener- getic reduction. Berthelot 142 heated alcohols or alkyl halides with concentrated hydriodic acid in sealed tubes and discovered that reduc- tion occurs as follows, C 2 H 5 I + HI- _*C 2 H 6 + I 2 Fatty acids may be reduced to paraffines of the same number of carbon atoms by this method and Krafft 143 prepared the normal paraffines from nonane to tetracosane, C 24 H 50 , by converting the ketones, made through the lime salts of the fatty acids, into the corresponding chlo- rides and reducing the latter with hydriodic acid (in the presence of red phosphorus). (C 10 H 21 ) 2 CO > (C 10 H 21 ) 2 CC1 2 > (C 10 H 21 ) 2 CH 2 or C 21 H 44 14 *J. prakt. ctiem. (1), 101,, 103 (1868). ia Ber. 15, 1687, 1711 (1882) ; IS, 2218 (1886). 48 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Magnesium amalgam has been employed for the reduction of alkyl halides by Meunier 144 and Wislicenus showed that the aluminum-mer- cury couple is of wide applicability. 145 Thus isobutyl, n. butyl and n.propyl iodides, treated with the aluminum-mercury couple, give the corresponding hydrocarbons in nearly quantitative yields in a few hours at ordinary temperatures, as compared with heating for 80 to 90 hours as is necessary with the copper zinc couple. Clemmensen has recently shown that ketones and aldehydes are readily reduced to hydrocarbons by the zinc-mercury couple and hydrochloric acid. 146 Zelinsky 147 has employed the zinc-palladium couple and alcoholic hy- drochloric acid with particularly good results in the case of iodine de- rivatives of cyclohexane and cyclopentane. Ordinarily, alkyl iodides give fair yields of the paraffines by reducing with zinc dust in acetic acid. 148 The classical researches of Sabatier and Senderens have shown that ethylene and its homologues may be converted into the corre- sponding hydrocarbons by hydrogen in the presence of nickel or nickel oxide. With ethylene, copper appears to give the best results. 149 Ipa- tiev also employed copper at 300 for the catalytic hydrogenation of trimethylethylene to pure isopentane. 150 Brochet and Cabaret 151 showed that alpha-octene is readily hydrogenated in the presence of active nickel and at atmospheric pressure, at temperatures as low as 65. At 160 p-hex#ne and (3-octene are rapidly hydrogenated but above 200 decomposition occurs with rupture of the carbon chain. The methene group > C = CH 2 , as in substances containing the allyl group, are more readily hydrogenated than other ethylene types. 152 Limonene, in the presence of copper, is hydrogenated to dihydroli- monene, only the A 8.9 group becoming saturated. Platinum black is generally not as effective in catalyzing hydrogenation as nickel and copper, but, with this catalyst also, the methene group is more easily reduced than other types. For example at 260 propylene is quickly reduced to propane and alpha-octene is rapidly hydrogenated at 215 but trimethylethylene and beta-hexene are not affected under these conditions. 153 The relative ease with which the methene group and Compt. rend. IS}, 473 (1902). " 8 J. prakt. Chem. (2), 54, 18 (1896). wChem. Zent. 1913, II, 255. " T Ber. 31, 3205 (1898). " 8 Wislicenus, Ann. 219, 312 (1883). " Sabatier and Senderens, Compt. rend. 130, 1559 (1900) ; 134, 1127 (1902). **Ber. 42, 2089 (1909); 43, 3387 (1910). 181 Compt. rend. 159, 326 (1914). 2 Albright, J. Am. Chem. Soc. 36, 2188 (1914). Sabatier and Senderens; Compt. rend. 124, 1358 (1897); 130, 1761 (1900); 131, 40 (1900) ; 134, 1127 (1902). THE PARAFFINES 49 other olefine types are hydrogenated by the action of sodium and alco- hol is just the reverse of the results noted above. Thus isoeugenol, isosafrol and isoapiol are very readily hydrogenated by sodium and alcohol but their isomers, containing the methene or allyl group, are not. 154 Although the catalytic hydrogenation or "hardening" of fatty oils 155 has become of great industrial importance, unsaturated or "cracked" petroleum distillates have not been successfully treated in this manner, at least not industrially. 156 It is very difficult to remove all of the sulfur from petroleum distillates and very small traces of this element are sufficient to poison the ordinary nickel catalyst. Rub- ber, prior to vulcanization and free from sulfur, does not appear to have been hydrogenated; oily saturated hydrocarbons might result. Unstable cyclic hydrocarbons or naphthenes might be hydrogenated with rupture of the ring, after the manner of the formation of isopen- tane from methylcyclobutane by hydrogen and nickel at 200. 157 By employing relatively high pressures, about 30 atmospheres, Ber- gius has hydrogenated fatty oils at 300 without a catalyst. 158 Whether or not unsaturated hydrocarbons derived from petroleum would also be hydrogenated under these conditions has not been determined but they are evidently not affected at 196 and 3000 pounds hydrogen pressure per square inch. 159 Whitaker and Rittman 16 in the produc- tion of oil gas at temperatures within the range 750 to 800 obtained distinct evidence of hydrogenation of the gaseous defines when hydro- gen was introduced into the mixture, particularly when operating at increased pressures. The platinum metals, when in a colloidal state of subdivision, are particularly useful in hydrogenating defines on a small scale or in the laboratory. Since the reaction is quantitative, they have been fre- quently employed to determine the number of olefine bonds in a sub- stance. The development of this method is due chiefly to Paal, Skita and Willstatter. Colloidal palladium, prepared according to Paal and Skita, 1M Ciamician and Silber ; Ber. 23, 1162. 2285 (1890) : Klages, Ber. 32, 1436 (1899). 158 Cf. Ellis, "The Hydrogenation of Oils," 1919 ; Erdmann, J. prakt. Chem. (2), 91, 469 (1915); Paal, Ber. 1,1, 2273 (1908); Skita, "Katalytische Reduktion," 1912; Sabatier, "La Catalyse," 1913. 1M Cf. Ubbelhode, Petroleum 7, 9, 334 (1912) ; Brooks, Bacon, Padgett and Hum- phrey; J. Ind. & Eng. Chem. 7, ISO (1915). ls7 Zelinsky; J. Soc. Chem. Ind. 32, 216 (1913); Philipow, J. prakt. Chem. (2), 95. 162 (1916). 168 Z. 1. angew. Chem. 1914, 522. 159 Brooks, Bacon, Padgett and Humphrey; J. Ind. Na 2 C0 3 + CH 4 . This reaction does not take place to any extent with the higher fatty acids but fairly good yields of the paraffines are obtained by heating the alkali salts, or soaps, with sodium methylate in vacuo. 164 Kolbe's electrolytic synthesis has often been cited but has been of very little preparative value. Thus, on electrolysing an aqueous solu- tion of sodium acetate the chief products are ethane, C0 2 and hydro- gen. 165 In general the electrolysis of a fatty acid salt yields, in addi- tion to the saturated hydrocarbon, an ester and an olefine. For ex- ample, sodium propionate gives butane, ethylene, ethyl propionate, car- bon dioxide and hydrogen. The higher fatty acids salts yield a mix- ture of reaction products of the same character. 166 Aldehydes have been converted into the corresponding hydrocarbons by electrolytic re- duction but the yields are very poor. 167 The well-known method of Wurtz, consisting in heating alkyl bro- mides or iodides with sodium, has had wide application in laboratory syntheses and it should be particularly pointed out that the reaction proceeds easily and with good yields with alkyl halides of high molecular weight. Alkyl chlorides have seldom been employed for this synthesis although Nef and others have called attention to the fact that alkyl bromides and particularly iodides have a much greater tendency to decompose to olefines, as in the ether reaction. C 4 H 8 + Nal + C 2 H 5 OH (main reaction) C 4 H 9 I + C 2 H 8 ONa V) 4 H 9 .O.C a H 5 + NaI The Wlirtz synthesis has also been useful in ring closing and in the syn- thesis of numerous derivatives of cyclic hydrocarbons of both the ben- 181 Ber. 1,5, 1471 (1912). j. prakt. Ghem. (2) 92, 40 (1915). > M Kelber, Ber. .45, 1946 (1912). IB* Mai: Ber. 22, 2133 (1889). '"Kolbe, Ann. 69, 257 (1849). Peterson, Z. f. Elektrochemie, 12, 141 (1906). '"Scheps, Ber. 1,6, 2565 (1913). THE PARAFFINES 51 zenoid and non-benzenoid type. Normal hexacontane, C 60 H 122 , the longest normal carbon chain compound known, was made by means of this reaction. 168 Optically active hydrocarbons have been prepared by employing the iodides of optically active alcohols. 169 1M Hell and HSgle, Ber. 22, 502 (1889). 188 It should be pointed out that the only satisfactory methods of preparing pure alkyl mono halides are those which utilize the corresponding alcohols. To obtain the primary halides, or alcohols, recourse is often had to the reduction of fatty acid esters by sodium and absolute alcohol, according to Bouveault and Blanc (German Pat. 164,294 (1903). The addition of halogen acid to alpha-olefines gives mainly secondary halides (R.CHX.CH 2 .) Chapter II. Chemical Properties of Saturated Hydrocarbons (1). Oxidation. The oxidation of saturated hydrocarbons by oxygen, or air, and other oxidizing agents is important in several respects, for example, the oxidation of lubricating oil in air compressors, the oxidation and carbonization of lubricating oils in automobile or other types of inter- nal combustion engines, the oxidation and resinification of trans- former oils, the bleaching of oils by air and sunlight and finally the oxidation of paraffine and other hydrocarbon mixtures to fatty or soap forming acids. Unfortunately very little research has been carried out with, pure specimens of different types of hydrocarbons with the re- sult that we know very little regarding their relative ease of oxida- tion. However some of the work recorded, having had to do with com- mercial products, is of industrial, if not scientific interest. As long ago as 1868 Bolley noted that paraffine wax absorbs oxygen at 150 but he made no particular study of the matter. 1 Others noted that when air is passed through hot mineral oils small quantities of acetic and other simple fatty 'acids are formed. 2 Holde noted the oxi- dation and thickening of mineral lubricating oils when heated in thin layers for 10 hours at 100 3 and in 1896 Byerly and Mabery described their now well-known process of manufacturing "artificial asphalt" by blowing air through heavy high boiling petroleum residues for four to five days at about 230. The reaction is strongly exothermic and the temperature may rise to 300-400 at the end of the operation. Water is formed during the process and very little oxygen remains in the final product, typical specimens showing 1.90 to 2.20 per cent oxygen. The bromine absorption values of the product are also low, ordinarily amounting to 14.0 to 19.0. The hardness and other physical properties of this asphalt would seem to indicate that considerable polymerization or condensation takes place during the process. Intermediate products 1 Z. f. Chemie, 1868, 500. 2 Zaloziecki, Z. anyeio. Chem. 1891, 416; Engler & Bock, Chem. Ztg. 16, 592 (1892). J. Soc. Chem. Ind. 13, G68 (1894) ; Ik, 174 (1895). 52 CHEMICAL PROPERTIES OF SATURATED HYDROCARBONS 53 containing oxygen are undoubtedly formed, which may condense with the elimination of the water which is always observed. At lower tem- peratures oxidation by air has a markedly different result: oxygen is absorbed forming fatty or naphthenic acids and some resinous matter. It would appear that at the higher temperatures employed by Byerly and Mabery the oxidation products first formed, conceivably alcohols, aldehydes and ketones, condense with the elimination of water, but at lower temperatures, these primary oxidation products are subjected to further oxidation to fatty or naphthenic acids. According to Worrall and Southcombe 4 lubricating oil may be heated to 750 F. in the presence of steam without causing resinifica- tion or other chemical change (although it may be noted that this is approximately the temperature employed by Burton for cracking heavy oils to gasoline). The resinous oxidation product which is slowly formed on heating mineral oils to 100-150 in contact with air, may partially be pre- cipitated by petroleum ether. The resin behaves as an acid and may be removed by shaking out with alcoholic alkali. Kissling 5 associates this resin with carbonization and for testing purposes has proposed the determination of "tar numbers" and "coke numbers" of lubricating oils, after heating to 150 for 50 hours under standardized conditions. 6 Transformer oils deteriorate by air oxidation particularly when the oil becomes heated as is usually the case when in service. As is indi- cated above, water, carbon dioxide, acid resinous material and simple fatty acids are formed. The latter are sometimes found in much used transformer oils in the form of iron or copper soaps, small quantities of which remain dissolved in the oil, and also in the form of insoluble basic salts or "sludge." Digby 7 states that these metallic soaps prob- ably act catalytically in promoting the oxidation. Waters states that "These substances (resinous) are oxidation products, and are most effi- cient oxygen carriers." . . . "By heat they become polymerized and changed into asphaltic matter." "If they are not removed (as by fil- tration through fuller's earth or bone black) heating the oil in the air produces more asphalt than would otherwise be the case." A particu- lar specimen of a typical resinous deposit showed 76.0 per cent carbon and 7.1 per cent hydrogen. The practical importance of the matter is apparent from the fact that 0.06 per cent of water in a transformer oil */. Soc. Chem. Ind. 2J, 315 (1905). Chem. Ztg. 30, 932 (1906) ; SI, 328 (1907) ; 52, 938 (1908) ; SS, 521, (1909). Compare Waters, U. S. Bur. Standards Bull. 7, 365 (1911) ; Circular 99 (1920). 7 J. Inst. Elec. Eng. 53, 146 (1915). 54 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS reduces its dielectric resistance to about 50 per cent of the value for the same oil when dry. 8 Pure paraffine oil can hold this amount of water in solution and commercial transformer oils are able to hold in solution three to four times this proportion of water. Waters 9 noted the forma- tion of 0.89 per cent of water in a lubricating oil exposed to air and light for 22 days. Light, however, accelerates oxidation by air. The above observations were carried out with refined commercial oils, but small percentages of olefine hydrocarbons were undoubtedly present in all the specimens investigated, since refining as ordinarily carried out with concentrated sulfuric acid does not remove all the olefines, the polymers thereby formed remaining in the oil. Generally olefines are more rapidly oxidized by air than saturated hydrocarbons, but Waters found that, of several oils examined by him, the one having the largest per cent of unsaturated hydrocarbons, as indicated by the iodine number and Maumene test, showed the least oxidation. Wa- ters suggests that these differences may have been due to greater amounts of catalysts or oxygen carriers in the oxidized oils. The con- clusion which may be drawn, however, is that factors other than the presence of olefines are of primary importance. On account of its de- composing action on resins and similar oxidized material, and its ener- getic action on unsaturated hydrocarbons, it is possible that oils re- fined by anhydrous aluminum chloride would be more stable and more resistant to oxidation in service as transformer oils than those oils which have been refined in the usual way with sulfuric acid. Although the accelerating effect of sunlight on oxidation by air is taken advantage of in the industrial sun bleaching of mineral oils, no study of individual hydrocarbons appears to have been made. Cia- mician and Silber 10 succeeded in oxidizing the methyl groups of toluene and xylene to the corresponding acids by air under the influence of sunlight. In the case of non-benzenoid hydrocarbons the group R >CH 2 and R 3 CH, would probably be oxidized rather than methyl R groups. It has been shown by the well-known work of Engler and Weiss- berg " that organic substances, which alone are not appreciably af- fected by air or oxygen, may readily be oxidized in the presence of a 8 The method advocated by C. E. Skinner, of purifying old transformer oils by quick-lime, removes both water and fatty acids. "Loc. cit. Ber. .45, 38 (1912). "Vorgange der Autoxydation, Brunswick, 1904. CHEMICAL PROPERTIES OF SATURATED HYDROCARBONS 55 second substance which is capable of direct oxidation. They have shown that the latter class of substances form peroxides and their hypothesis is that these peroxides may then effect the oxidation of substances which by themselves are inert to oxygen. Thus paraffine wax is only very slowly affected by air or oxygen at 150 but the oxidation is very much accelerated if a small quantity of previously oxidized material is introduced. Unsaturated hydrocarbons which are capable of forming peroxides, according to Engler's theory RCH = CHRi + 2 > RCH CHR v may in this way bring about the oxidation of saturated hydrocarbons. Based upon this theory the oxidation of paraffine has been brought about by first chlorinating at 160 followed by decomposition of these chlorides by heating to 300 and then oxidizing to fatty acids. 12 Organic peroxides are decomposed by moisture which explains the finding of Charitschkoff mentioned above. Thus linseed oil shows greater increase in weight on "drying" in dry air than in moist air, at least during the first few days' exposure. The oxidation of paraffine wax by air at 120 and 150 was noted as long ago as 1868, 13 but under the stress of the conditions prevailing in Central Europe during the war intensive research on the synthesis of fatty acids was carried out by a special commission of the German government, presided over by C. Engler. Numerous researches of the same character were undertaken by private concerns and a number of patents and published papers have recently appeared dealing with this subject. The statements of different investigators regarding the ef- fect of metallic oxides and other substances introduced as catalysts is very contradictory but the most complete results published up to the present time indicate that the best yields are obtained without the addition of any catalytic material other than a small amount of pre- viously oxidized material added to initiate the reaction. 14 The use of air under pressure accelerates the oxidation 15 but the substitution of oxygen for air causes the reaction to proceed too rapidly and per- oxides are formed and accumulate to such an extent that violent ex-, plosions are apt to occur. When the oxidation is slowly and carefully carried out waxy esters of the fatty acids and higher alcohols, formed 12 Schaarschmidt & Thiele, Ber. 53B. 2128 (1920). ls Bolly & Tuchschmidt. Z. f. Chemie. 1868, 500; Jazukowitsch, Ber. 8, 768 (1875). "Griin. Ulbrich & Wirth, Ber. 5SB. 987 (1920). 18 Loffl, Chem. Ztg. kk, 561 (1920). Schneider, J. Soc. OJtem. Ind. 40, 141A. (1921), uses tubular retorts and air under 70 atmospheres pressure. 56 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS evidently as intermediate products, may be isolated from the oxidized mixture. 16 The oxidation to fatty acids, with yields amounting to approximately 86 per cent of the theory, takes place in a remarkably short time, this yield being obtainable in 12 hours at 160 in the ab- sence of catalysts. In a normal oxidation the peroxides noted above are decomposed, probably assisting in the oxidation in the manner indi- cated by Engler and Weissberg. Carbon dioxide, formic, acetic and other simple, volatile fatty acids are formed and the yield of these ap- pears to vary within wide limits, one of the "tricks" of the process being so to conduct the oxidation that only small proportions of these malodorous acids are formed. Presumably these volatile acids are removed by blowing with live steam, the residue having an acid num- ber of 180 to 200, being then neutralized by alkali and the unsaponifi- able portion returned for further oxidation. According to Loffl 17 acids satisfactory for soap manufacture have not yet been obtained, the addi- tion of 10 to 20 per cent of cocoanut or palm oil being necessary to produce a soap of the desired detergent qualities. According to Loffl 120 is the best working temperature with air under about 45 pounds pressure. The presence of water, continually introduced with the air in the form of steam, favors the production of the higher fatty acids and in the absence of water or its removal as fast as formed the product is highly colored and partially resinified. As is usual in such cases a large number of special patents have appeared 18 claiming special ad- vantages for various catalysts and other minor details of operation although the general process seems to have been broadly covered by previous publications, particularly the patent of Schaal. 19 Most of the published work on this subject has had to do with the oxidation of paraffine wax, probably with the idea of manufacturing fatty acids identical with fatty acids occurring in natural fats and oils, but in view of the much larger quantities of liquid naphthenic hydrocarbons of fifteen to twenty carbon atoms (present in kerosene and the inter- mediate or fuel oil distillates) and the lower cost of such material, it would seem highly desirable to study the oxidation of such oils under similar conditions. Although the carboxylic acid derivatives of the naphthenes as exemplified by the Russian naphthenic acids, have ob- jectionable and very persistent odors, it is probable that these cyclic 18 Griin. Ulbrich & Wirth. Ber. 53B. 987 (1920). "Loc. cit. 18 Pardubitzer Fabr. Akt. Ges. f. Mineralolindustrie, Brit. Pat. 131,301 ; 131,302 ; 131,303; Schmidt, Brit. Pat. 109,386 (1907) ; Cf. also Fischer & Scheider, Ber. 53, 923 (1920) ; Kelber, Ber. 53, 66 (1920) ; Bergman, Z. f. angew. Chem. 31, I, 69 (1918) ; Holde, Chem. Ztg. 78, 447 (1920) ; Plauson, Brit. Pat. 156,141 (1919). 19 Schaal, German Pat. 32, 705. CHEMICAL PROPERTIES OF SATURATED HYDROCARBONS 57 hydrocarbons would be decomposed by oxidation to open chain acids. Montan wax is more resistant to air oxidation than paraffine wax. 20 Harries has applied his well-known method of ozonization to highly unsaturated oils such as the oily distillates obtained by the low tem- perature carbonization of coal and lignite. When saturated hydrocarbons are burned with insufficient air for complete combustion, a little formaldehyde is formed. From a hexane fraction and isopentane Stepski 21 obtained water, carbon dioxide, for- maldehyde, ethylene and small quantities of propylene, butylene and amylenes. The yields of formaldehyde and ethylene by known meth- ods are too small for the process to be of industrial value. The action of chemical oxidizing agents on saturated hydrocarbons shows that certain structures are more easily oxidized than others. Zelinsky and Zelikow 22 have noted that hydrocarbons of the type R >CHR R for example (C,H 5 ) 2 CH.CH 3 are readily oxidized by one per cent potassium permanganate solution. As contrasted with this, methane and ethane are only very slowly oxidized by five per cent perman- ganate solutions. 23 The hydrocarbon 2 . 6-dimethyloctane is fairly sta- ble to permanganate at 100 but in the presence of unsaturated hydro- carbons (menthene) the dimethyloctane is oxidized rather rapidly even at 50. - 4 (3-Butylhexane is rapidly oxidized by alkaline permanganate solution at 80 to 90, but the only oxida.tion products which can be detected are carbon dioxide and formic acid: by oxidizing it at 25 a very small amount of butyric acid can be recognized. 25 Hydrocarbons of the type R 1 R 2 R 3 CH are also very easily oxidized by concentrated nitric acid, Sp. Gr. 1.53 but normal hydrocarbons, at ordinary tempera- tures are only very slowly acted upon. Less concentrated acid, Sp. Gr. 1.42 26 gives a mixture of nitro derivatives and oxidation products of the normal hydrocarbons, and the least oxidation and maximum yields of nitro derivatives are obtained by heating, preferably in sealed tubes, with dilute nitric acid of 1.075 specific gravity. 27 Paraffine wax is slowly oxidized by nitrogen peroxide 28 at temperatures within the 20 Schneider, J. Soc. Chem. Ind. W, 140A. (1921). ^Monatsh. 23, 773 (1902). M Ber. 34, 2865 (1901). 23 V. Meyer & Saam, Ber. SO, 1438 (1897). "Kishner, J. Russ. 1,5, 1788 (1913). "Levene & Cretcher, J. Biol. Chem. 33, 505 (1918). "Worstall, Am. Chem. J. 20, 209 (1898) ; 21, 213 (1899). 27 Konowalow, /. Russ. Phys.-Chem. Soc. 27, 418 (1895) ; Chem. Zentr. 1900, I, 975. 28 Granacher, Helv. Chim, Acta. 3, 721 (1921). 58 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS range 110-150. A mixture of fatty acids, from acetic upwards in the series, is produced. Alkaline solutions of these fatty acids are red in color due to the presence of nitro compounds. As might be expected the results of oxidizing with nitric acid and by permanganate are quite different. The fatty acids, with the exception of acetic acid, are almost invariably more readily oxidized than the hydrocarbons and large yields of the former could, therefore, hardly be expected among the reaction products. Prshevalski 29 has shown that the higher normal fatty acids are oxidized by permanganate at two points, i. e., at the carbon atom adjacent to the end methyl group and also at the carbon atom adjacent to the carboxyl group. Isobutyric acid is oxi- dized to the oxy acid (CH 3 ) 2 = C C0 2 H but with hydrocarbons the molecules are completely broken up. Nitric acid, however, forms a series of fatty acids and dicarboxylic acids. In addition to carbon dioxide and the simpler fatty acids, oxalic, succinic and adipic acids have been observed among the oxidation products. 30 Oxidation by nitric acid may become violent at 100. 31 Chromyl chloride, Cr0 2 Cl 2 has been employed by Etard 32 and by Miller and Rohde 33 to oxidize the aliphatic side chains of benzene de- rivatives. Toluene yields benzaldehyde and ethyl benzene is oxidized mainly at the CH 2 group to form acetophenone. This interesting re- action, however, has not been applied to the study of the paraffine hy- drocarbons, although Etard oxidized hexane to a chloroketone and Schulz 34 treated a number of light fractions from Boryslaw petro- leum with chromyl chloride, obtaining ketone mixtures which were not further studied or identified. Sulfur. Sulfur reacts with paraffines and naphthenes on heating, hydrogen sulfide being evolved, but little is known regarding the other products formed. Galletly 35 first noted that hydrogen sulfide could conven- iently be prepared by heating sulfur and paraffine wax. Somewhat 29 J. Chem. Soc. Abs. 1913, I, 1151. 80 Markownikow, Chem. Zentr. 1899, I. 1064; II, 472. 473; Ber. 82, 144 (1899); J. prakt. Chem. (2), 59, 556 (1899). 31 Young & Francis, J. Cliem. Soc. 73, 928 (1S98). 32 Compt. rend. 90, 534 (1880). 33 Ber. 23, 1070 (1890). 3 *Petr. 6, 189. 36 Chem. News 2! lt 107 (1871). CHEMICAL PROPERTIES OF SATURATED HYDROCARBONS 59 later Lidoff 36 made hydrogen sulfide by passing naphtha vapors into sulfur at 350 to 400. Friedmann 37 isolated thiocresol as one of the reaction products of sulfur and methylcyclohexane but was unable to isolate benzene from the reaction product of sulfur and hexane. He isolated dinitrobenzene after nitrating the product, which Friedman states is possibly due to the initial formation of cyclohexadiene which on nitration is converted into dinitrobenzene. Guiselin 38 noted that "benzine" dissolves about 0.5% sulfur at 20, the higher boiling distil- lates dissolving somewhat more than this. Raffo and Rossi 39 state that pyridine catalyzes the evolution of hydrogen sulfide when hydro- carbons and sulfur are heated. Markownikow 40 obtained xylene from an octonaphthene. Normal hexane 41 is practically inert to sulfur at 210 but parafrme wax or heavy greases react vigorously at this tem- perature. Prothiere 42 obtained 48 liters of hydrogen sulfide from 70 grams of sulfur and 30 grams of vaseline. It has occasionally been suggested that sulfur might be employed to remove hydrogen from paraffines or petroleum oils to form highly unsaturated oils having several double bonds and such products presumably would have the general character of drying oils. However, since sulfur reacts much more readily with the ethylene bond than it does upon saturated hydro- carbons, the result is a certain amount of carbonized material and un- changed oil or paraffine, a hydrocarbon molecule once being reacted upon then rapidly reacting with more sulfur to form a series of prod- ucts of unknown character, the final product resembling asphalt, or when strongly heated, petroleum coke. When added to heavy residuum and blown with air, sulfur has the effect of giving a markedly harder so-called asphalt. 43 Sulfur derivatives frequently exhibit a much greater tendency to polymerize than their oxygen analogues and this fact may account for the greater hardness, i. e., greater degree of poly- merization, of asphalts made from residuum high in sulfur; for example, that from Mexican petroleum. Under these conditions a large part of the sulfur contained in or added to the original residuum remains in the final product. The following results obtained by blowing a residuum, 12 Be, from Texas Gulf Coast petroleum, with air are rep- resentative. M Chem. Zentr. 1882, 22. "J. Chem. Soc. A 6*. 1917. I, 13. 88 Petroleum, 1913, 1309. * g Gcusz. Chim. Ital. 44, 104 (1914). *>Ber. 1887, 1850. 41 Spanier, Dissertation, Karlsruhe, 1910 42 Ohem. Zentr. 1903, I. 492. 43 Brooks & Humphrey, J. Ind. d Eng. Chem. 8, 746 (1917). 60 CHEMISTRY OF THE NOtf-BENZENOJD HYDROCARBONS EFFECT OF SULFUR ON HARDNESS OF BLOWN ARTIFICIAL ASPHALT. Sulfur Temp, Hours Penetration added % C Blown mm.* Flowing-point C (1) None 210 14 Too soft for measurement (2) 4.0 210 10 61 73 (3) 6.0 210 10 28 109 (4) 8.0 210 10 17 148 (5) 8.0 215 10 13 167 *Penetration of No. 2 needle, 100 gram weight for 5 seconds at 25C. Nitration of Non-Benzenoid Hydrocarbons. Probably on account of the great industrial importance of nitro derivatives in the aromatic series, the nitration of non-benzenoid hy- drocarbons of open chain and cyclic structure has been relatively little investigated. Oxidation by nitric acid generally takes place to a much greater extent in the case of saturated non-benzenoid hydro- carbons than with those of the aromatic series and the relative yields of oxidation and nitration products depend upon many factors, chief of which are the concentration of the nitric acid used and the tempera- ture. The constitution of the hydrocarbon is also of importance. The use of dilute nitric acid, Specific Gravity 1.025 to 1.075, at 115 to 125, constitutes a method whereby fairly good yields of nitro-deriva- tives may be obtained. The reaction is usually carried out in sealed tubes in the case of very volatile hydrocarbons, but easily nitrated hydrocarbons are preferably heated with the dilute acid under a re- flux condenser. These methods are due chiefly to Konowalow, 44 and to Markownikow. 45 Concentrated or fuming nitric acid or nitric-sulfuric acid nitrating mixture gives mostly oxidation products. Hydrocarbons containing a tertiary hydrogen atom, R 3 CH, are most easily nitrated; for example, 2, 5-dimethylhexane yields a dinitro derivative, 46 which is insoluble in alkali solution and which exhibits the exceptional property of being crystalline, melting at 124-125. CH 3 - C - CH 2 CH 2 - C - CH 3 -- > CH 3 - C - CH 2 CH 2 - C - CH 3 H H N0 2 N0 2 The hydrocarbon 2, 6-dimethylheptane similarly gives the tertiary nitro derivative, which is easily separated from the relatively small amount of primary and secondary nitro-compounds by the solubility "Ber. 25, 1244 (1892) ; 28, 1852 (1895) ; 29, 2199 (1896). * 8 J. prakt. Cliem. (2) 59, 564 (1899). 46 Konowalow, J. Russ. Phys.-Chem. Boc. 38, I, 109 (1906). CHEMICAL PROPERTIES OF SATURATED HYDROCARBONS 61 of the latter two classes in aqueous alkali. This solubility in alkali of nitro derivatives of the types CH.N0 2 and CH 2 N0 2 is a general property common to all nitro derivatives of these types, 47 the alkali salts probably having the constitution represented by the formulae CH^N and ONa ONa The hydrocarbon 2,7-dimethyloctane yields, with concentrated nitric acid, the primary nitro derivatives, but with dilute acid gives 2,7- dinitro-, 2,7-dimethyloctane, melting point 101.5-102. As contrasted with the above hydrocarbons, containing tertiary hydrogen, the hydrocarbons (CH 3 ) 3 C.CH 2 CH 3 and (CH 3 ) 3 C.CH 2 CH 2 CH 3 are nitrated only with difficulty; in fact, the former can be purified from isomeric hydrocarbons by repeatedly nitrating the fraction boil- ing at 48-51. 48 When these hydrocarbons are nitrated, the nitro- group is attached to the carbon atom next to the (CH 3 ) 3 .C group. The normal paraffines are also nitrated much less readily than their branch chain isomers. Di-isopropyl (CH 3 ) 2 CH.CH(CH 3 ) 2 reacts very ener- getically with nitric acid at 20, but not with the nitric-sulfuric acid nitrating mixture commonly employed to nitrate benzene. That saturated non-benzenoid hydrocarbons are more easily ni- trated by dilute nitric acid than the benzene ring is shown by a number of examples. Phenylcyclohexane is nitrated in the cyclohexane, not in the benzene ring. 49 Here also nitration takes place at the tertiary hydrogen atom yielding 1 -nitro- 1-pheny Icy clohexane CH 2 CH 2 N0 2 H 2 C C CH 2 CH 2 C 6 H 8 O/io-xylene with dilute nitric acid, Sp. Gr. 1.075 at 110 gives o-tolylnitromethane, 50 which like all primary nitro compounds easily forms alkali salts. Dilution of nitric acid with acetic acid has practi- Cf. Nef. "Constitution of the Nitroparaffines." Ann. 270, 331 (1892) ; 280, 263 (1894). Markownikow. Chem. Zentr. 1S99, II, 472 : Ber. 82, 1446, (1899) : Ber. S3, 1908 (1900). "Kursanoff. J. Chem. Soc. Abs. J907, I, 599. "Konowalow. J. Chem. Soc. Aba. 1905, I, 762. 62 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS cally the same effect as dilution with water; dilution and heating di- rects the nitration chiefly to the side chain, forming nitro derivatives and also acids by oxidation. In accord with the above observations 1 . 2-dipheny Ipropane gives the primary nitro derivative rather than substitution in the benzene ring, C 6 H 5 CH 2 CH(C 6 H 5 ) .CH 2 NO 2 . Ben- zoyl nitrate is a reagent, which with benzene, toluene, phenol, anisole, naphthol, coumarine and thiophene gives nitro derivatives very smoothly, but when several methyl groups are present, nitration of a methyl group takes place, as in durene. 51 C 6 H 2 (CH 3 ) 4 - ~>C 6 H 2 (CH 3 ) 3 .CH 2 N0 2 In nitrating p-cymene the isopropyl group is attacked at the ter- tiary hydrogen atom forming p-methylacetophenone, by oxidation, un- less special precautions are taken, 52 advantage being taken of the fact that the paraffines are but little affected by nitric-sulfuric acid nitrat- ing mixture. The aliphatic ketones are much more reactive to nitric acid than the hydrocarbons. Nitric acid, specific gravity 1.38, yields a mixture of products of which dinitro ketones and dinitro hydrocarbon deriva- tives are conspicuous, the formation of these products being accom- panied by splitting of the carbon structure, probably as indicated by the reaction, (a) RCO . CH 2 R' > RCO . C (N0 2 ) 2 R' (b) R . CO . C (N0 2 ) 2 R' + H 2 > RCOOH + R'CH (N0 2 ) 2 Menthone is readily nitrated by dilute nitric acid to the mononitro derivative 53 = =0 81 WillstKtter & Kubli. Ber. W, 4152 (1909). 82 Andrews, J. Ind. d Eng. Ctiem. 10, 453 (1918). M Konowalow, Ber. 26, Ref. 878 (1893) ; 28, Ref. 1054 (1895) CHEMICAL PROPERTIES OF SATURATED HYDROCARBONS 63 Suflonation The difference in the ease with which benzenoid hydrocarbons on the one hand and paraffine or non-benzenoid hydrocarbons on the other are sulfonated is not as great as is commonly supposed. Although data with respect to hydrocarbons of known character and purity are extremely meager, Worstall 54 showed that n . hexane, n . heptane, and n. octane are readily sulfonated by fuming sulfuric acid at the tem- perature of a water bath and Markownikow 55 states that naphthenes also are reacted upon by fuming sulfuric acid, both sulfonation and oxi- dation taking place. Paraffine wax is attacked by warm fuming sul- furic acid but oxidation rather than sulfonation is the result. 56 Oxida- tion occurs with fuming acid and saturated hydrocarbons to a much greater extent than in the case of benzene and its derivatives. Hy- drocarbons containing a tertiary hydrogen group as in di-isopropyl (CH 3 ) 2 CH.CH(CH 3 ) 2 are much more readily sulfonated and oxi- dized than normal paraffine hydrocarbons and it is possible that the large losses experienced in the refining of lubricating oils by concen- trated sulfuric acid are in part due to the sulfonation and oxidation of branched chain hydrocarbons. Halides. Preparation and Properties. In the following pages the methods of preparation and more par- ticularly the properties of the simpler alkyl halides will be discussed. Very little work has been done with fluorine derivatives, and such information as we have does not indicate that fluorine derivatives possess particularly interesting or valuable properties. When writing of the halogen derivatives, it will, therefore, be understood that gen- erally chlorides, bromides or iodides only are meant. Our knowledge of the simpler alkyl halides, especially chlorides, has recently been much extended by the development of synthetic rubber, and this, it may be noted, is coincident with the production of enormous quantities of electrolytic chlorine at very low cost. Cheap chlorine makes many processes industrially possible, which heretofore have been only of theoretical interest. The conditions for the chlorination of methane have been noted (page 79). Chlorine reacts readily with butane and pentane and the higher paraffines in the cold and in diffused daylight. It has repeatedly been observed that in chlorinating petroleum ether a sluggish so-called induction period is first noted. The chlorine dissolves in the hydro- "Am. Chem. J. 23, 654 (1898). 85 J. Russ. Phys.-Ctiem. Soc. 1892, 141. Michailescu & Istrati. Bull. Soc. Sci. Bucharest. 13, 143. 64 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS carbon apparently without reacting, but, after a few minutes, the color is suddenly discharged and the reaction thereafter proceeds very rapidly. An excellent example of this peculiar induction period 57 has been observed in the case of brominating cyclohexanone and 1, 1-di- methylhexanone(3). Catalysts are not necessary, although the pres- ence of moisture is distinctly advantageous in the case of the more volatile hydrocarbons. 58 The chlorination of petroleum pentane has become of industrial importance in connection with the manufacture of synthetic amyl acetate (see page 89) . In order to produce mainly monochlorides, it is necessary to stop the chlorination after the concen- tration of monochlorides in the reaction mixture has reached about 20 per cent, and separate the unchanged pentane by fractional distilla- tion. The relative proportions of the isomeric monochlorpentanes formed are not known. Cyclohexane is more reactive to halogens than n.hexane. When n.hexane is chlorinated, the CH 2 groups, not the CH 3 groups, are attacked 59 The chlorination of paraffine wax is carried out industrially, the product being used as a solvent for dichloramine T. 60 Boiling the product with aniline readily removes most of the chlorine. The higher boiling petroleum fractions, are readily chlorinated in diffused day- light at ordinary temperature, but the products are very unstable. Bromine reacts more slowly with the paraffines. Pentanes may be brominated readily under the influence of intense illumination, and the higher paraffines react readily with bromine when gently warmed and illuminated. In the presence of metallic iron or ferric bromide, bro- mine readily forms a series of substitution products in which one bro- mine atom is attached to each carbon atom, thus CH 3 CH 2 CH 2 CH 2 CH 3 > CH 2 Br . CHBr . CHBr . CHBr . CH 2 Br . Normal heptane and an excess of bromine in the presence of iron yields 1, 2, 3, 4, 5, 6, 7-heptabromoheptane. Ethyl bromide may be brominated under these conditions to ethylene bromide and propyl bromide to 1 . 2-dibromopropane. 61 Bromides have been made by treat- ing chlorine derivatives, such as CC1 4 , C 2 C1 4 and C 2 C1 6 , with anhydrous aluminum bromide. 62 Sodium iodide reacts with many alkyl chlorides and bromides to "Crossley & Renouf. J. Chem. &oc. 91, 81 (1907). ""Aschan, Chem. A~bs. 1919, 2868. 59 Ber. 89, 2138 (1906); Strauss claims to be able to prepare primary mono- chlorides by chlorinating at reduced pressures and temperatures above the boiling- point of the hydrocarbons. (German Pat. 267, 204.) Dakin & Dunham. Brit. Med. J. 1918, I, 51. "V. Meyer & Mtiller, J. prakt. Chem. (2) J,6, 171 (1892). 2 Gustavson, Chem. Zentr. 18S1, 131,642. CHEMICAL PROPERTIES OF SATURATED HYDROCARBONS 65 give the corresponding alkyl iodide and sodium chloride or bromide. Sodium iodide dissolves readily in acetone and this solvent yields the best results in carrying out the reaction, which usually takes place at once at ordinary temperatures, with the separation of sodium chloride or bromide. 63 Unsaturated substances may be brominated without affecting the double bond, substitution taking place, by employing N-bromo-aceta- mide. Thus (CH 3 ) 2 C = C(CH 3 ) 2 yields (CH 3 ) 2 C = C(CH 3 ) CH 2 Br. Sodium hypobromite, although a very energetic oxidizing agent, converts acetone into carbon tetrabromide (and acetic acid). Bromo- form also yields carbon tetrabromide with this reagent. 64 In the great majority of cases, it is much preferable to prepare alkyl halides from an alcohol or olefine than by treating the hydro- carbons themselves with chlorine or bromine, the latter method giving mixtures of isomeric derivatives. Since it is usually possible to ob- tain the simpler aliphatic alcohols in a state of purity, they constitute a valuable raw material for the preparation of pure mono-halides. The methods employed will only be mentioned and reference made to original articles or works on preparative methods for further data. 65 (1) Hydrochloric acid gas, and methyl or ethyl alcohol in the presence of zinc chloride gives good yields of the corresponding chlo- rides, but this method is practically valueless with the higher alcohols on account of the instability of the higher alkyl chlorides in the pres- ence of zinc chloride. Tars or heavy polymers are formed with the higher alcohols. However, Norris 66 has obtained good yields from a large number of alcohols by using a large excess of concentrated hydro- chloric acid, without zinc chloride. (2) Hydrobromic acid and hydriodic acid give very much better yields, than hydrochloric acid. With the simpler alcohols the well- known sulfuric acid and sodium bromide method gives excellent re- sults, but not with the higher alcohols. In the case of the higher alco- hols much better results are obtained by the method of Norris, 66 in which the alcohol is gently heated with the concentrated aqueous acid Finkelstein, Ber. 43, 1528 (1910). "Dehn, J. Am. CJhem. Soc. SI, 1220 (1909). Weyl, Methoden d. Org. Chem. II. 1077. Norris, Watt & Thomas, J. Am. Chem. Soc. 38, 1071 (1916). Norris & Mulliken : J. Am. Chem. Soc. 48, 2093 (1920). According to the author's experience, the halides prepared according: to this method are purer and much preferable to similar products made by other methods. Particularly is this true when the halides are to be employed in a Grignard reaction. According to German Patent 280,740 the addition of calcium chloride is advantageous. 66 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS and the alkyl halide removed as formed by distillation from the mix- ture. This method gives especially good yields with tertiary amyl alcohol, tertiary butyl alcohol, octyl alcohol, cetyl alcohol, etc. By the Norris method good yields are obtainable in many cases with con- centrated hydrochloric acid. In this connection it is of interest to note that the glycol S(CH 2 CH 2 OH) 2 gives the dichloride quantitatively with concentrated aqueous hydrochloric acid at 60. 67 Tertiary alco- hols of the type ROCH 2 C(OH)R 2 give excellent yields of the chloride on warming with 38% hydrochloric acid. 68 (3) The use of PC1 3 , PC1 5 and PBr 3 in preparing alkyl chlorides and bromides is well known, but the yields are greatly reduced by the formation of esters of phosphorus or phosphoric acid. Iodides are com- monly made by introducing iodine into a mixture of the alcohol and red phosphorus which method has been recently improved for the simpler alkyl iodides by Adams and Voorhees. 69 Dehn and Davis 70 state that yields of 85 and 88 per cent of isobutyl and iso-amyl chlo- rides respectively can be obtained from the corresponding alcohols by adding PC1 3 to a mixture of the alcohol with concentrated aqueous zinc chloride. In the case of tertiary alcohols, acetyl chloride fre- quently reacts abnormally, giving the corresponding chlorides instead of the acetates, for example, dime thy Ibutylcarbinol thus yields the cor- responding chloride. (4) It is well known that olefines combine with halogen acids to form alkyl halides. In the case of hydrocarbons, generally the halo- gen will combine with that carbon atom of the olefine group which has combined with it the least number of hydrogen atoms, which generali- zation is known as Markownikow's rule: CH 3 CHI . CH 3 CH 3 >C CH 3 CH 3 | Br However, small quantities of the isomeric halides are sometimes formed. Thus propylene yields very small quantities of n.propyl OT H. T. Clarke, J. Chem. Soc. 101, 1583 (1913). Gomberg, /. Am. Chem. Soc. 41, 1415 (1919). This chloride is the now well known "Mustard Gas." M Paloma, Ghent. Abs. 1919, 2862. /. Am. Chem. Soc. 41, 789 (1919). 70 J. Am. Ohem. Soc. 29, 1328 (1907) ; When PC1 ? reacts with an alcohol, succes- sive formation and decomposition of the whole series of possible alkyl phosphites results, and the series of reactions may be arrested by choosing the experimental conditions to get very large yields of P(OR) 3 , P(OR) 2 .OH, P(OR).(OH) 2 or P(OH) and 3 RC1. [Milobeudzki and Sachnowski, J. Chem. Soc. Alia. 1918, I. 477.] CHEMICAL PROPERTIES OF SATURATED HYDROCARBON** 67 iodide, 71 and isobutylene yields, with a solution of HBr in acetic acid, about 93 per cent tertiary butyl bromide and 7 per cent isobutyl bro- mide. 72 Acetic acid solutions of hydrogen chloride and bromide have given particularly good results in the terpene and sesquiterpene series, where crystalline hydrochlorides are often difficult to obtain. 73 The ability of an olefine to combine with halogen acid depends somewhat upon its structure. Thus trimethylethylene CH, >C = CH.CH 3 combines readily with hydrogen chloride, but the CH 3 isomeric amylenes do not. Advantage of this fact is taken in one of the synthetic rubber processes, 74 in which normal pentane is chlorinated to a mixture of monochlorides. The monochlorides are converted into amylenes by passing over quicklime at 385 to 400 and the resulting amylene vapors are passed over alumina at 450. The amylene frac- tion boiling from 34 to 38 contains trimethylethylene, which is re- moved by combining with hydrogen chloride and the tertiary amyl chloride, boiling point 84 to 86, isolated by fractional distillation. Unstable carbon ring structures are often ruptured by halogen acids. Pinene in acetic acid solution gives mainly dipentene dihydro- chloride, with rupture of the bridged or cyclobutane ring. Bromocyclopropane 75 and bromocyclobutane 76 and cyclopropyl carboxylic acid are converted into open chain compounds by concen- trated hydrobromic acid. Thus CH 2 >CHBr + HBr > CH 3 CHBr.CH 3 Br. CH 2 CH 2 > CH . C0 2 H + HBr > CH 2 Br . CH 2 CH,CO 2 H H 2 Br:H vs j, CH 2 :-CH 2 CH 2 Br CH 3 CH 2 -CHBr * CH 9 CHBr L 2 "Michael & Leighton, J. prakt. Chem. 60, 348, 446 (1899). 72 Ipatiev & Ogonowsky, Ber. 86, 1988 (1903). 73 For good results, the reaction mixture saturated with HC1 or HBr should be allowed to stand two or three days in a cool, dark place. 7 *Badische A. & S. Fab. Brit. Pat. 18, 356 (1911). " Willstatter & Bruce, Ber. 40, 4457 (1907). "Perkin, J. Chem. Soc. 65, 950 (1894). 68 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS It will be noted in the last case that the bromine atom of the HBr molecule combines in such a way as to place it in the position farthest removed from the bromine atom already present in the ring. This il- lustrates the positive-negative rule of Michael, 77 which is not as em- pirical as Markownikow's rule. According to Michael's principle the combination of two molecules, for example, halogen acid and an olefine, tends to occur with such structural results as will give the maximum degree of entropy, that is the neutralization of the chemical energies or affinities of the reacting atoms. This generalization is, therefore, a special case of Ostwald's hypothesis that "every system tends towards that state whereby the maximum entropy is reached. 78 Michael formulated his principle after a comprehensive study of addition reactions. The marked influence of a methyl group is shown in the formation of CH 3 CHBr.CH 3 from propylene, CH 3 CBr 2 CH 3 from CH 3 CBr=:CH 2 , CH 2 Br.CH 2 .C0 2 H the chief product of HBr and CH 2 =<^H.C0 2 H, etc. The rule is not without many exceptions, however. Faworsky 79 considers the matter from the standpoint of relative reaction velocities. By heating isopropyl bromide to 250 sev- eral times, removing the fraction boiling at 69-70 each time he was able to effect 20 per cent conversion of isopropyl bromide to normal propyl bromide. The conversion of normal propyl bromide to iso- propyl bromide is therefore reversible and the addition of HBr to propylene takes place in part contrary to Markownikow's rule and Michael's principle, the result being dependent upon the relative ve- locities of the two reactions. (1) CH 3 CH = CH 2 + HBr CH 3 CHBr.CH 3 (main result) . (2) CH 3 CH = CH 2 + HBr > CH 3 CH 2 CH 2 Br Similar reversible relations were found in the bromopentane series. Faworsky confirmed the earlier observation of Eltekow that isobutyl and tertiary butyl bromides are in equilibrium at about 210, as noted in the following: CH 3 CH 3 (CH 3 ) 3 CBr^HBr+ >C = CH 2 ? >CHCH 2 Br. CH 3 CH 3 In a similar manner it was shown that ethylidene bromide is present in the mixture resulting from heating ethylene bromide: "J. prakt. Chem. 1,6, 205 (1892). 78 J. prakt. Chem. 60, 286, 292 (1899); Bcr. 39, 2138 (1906). "Ann. 354, 325 (1907). CHEMICAL PROPERTIES OF 1 SATURATED HYDROCARBONS 69 CH 2 Br CHBr 5 1 1 + HBr CHCH 2 Br > > C CH 2 CH 3 CH 8 | | CH 3 CH 3 (b) > C CH 2 + C 2 H 5 OH > > C CH 3 CH 3 || CH 3 | OC 2 H 5 In a similar manner isobutyl iodide and silver cyanate yield a mixture of about two parts tertiary butyl isocyanate and one part of the iso- butyl derivative; silver acetate in acetic acid yields about two parts tertiary butyl acetate to one part isobutyl acetate. 89 Wischnegradsky showed that secondary iso-amyl alcohol with halogen acids yields chiefly the tertiary halide. 90 CH 3 >CHCH.CH 3 CH 3 CH 3 I > >C.CH 2 CH 9 6H CH 3 | Also the secondary halide, when heated with lead hydroxide, yields the tertiary alcohol. But the facts are somewhat more involved than is indicated above. Thus isobutyl alcohol, when decomposed by heat, and the primary isobutyl halides with alcoholic alkali gives a mixture of butylenes 88 J. prakt. chem. (2), 12, 26 (1875). 87 The nature of the decomposition products of alkyl halides in the presence of anhydrous aluminum chloride, either alone or in the presence of saturated or un- saturated non-benzenoid hydrocarbons, has never been carefully investigated. Cf. Meyer, Ber. 27, 2766 (1894). 88 Ber. 24, 2832 (1891) ; 31, 1344 (1898) ; 32, 3647 (1899). Nitrated tertiary butyl- toluene and xylene are known commercially under the name of artificial musk. 88 Butlerow, Ann. 168, 143 (1873); Nef, Ann. 309,150 (1899). 90 Ann. 190, 342 (1878). CHEMICAL PROPERTIES OF SATURATED HYDROCARBONS 71 which have been shown to contain isobutylene and a and (3-normal butylenes. CH 3 CH 3 > CH . CH 2 X - > > C = CH 2 and CH 3 CH 3 and CH 3 CH 2 CH = CH 2 Nef believed that such facts could best be explained by the inter- mediate formation of a cyclopropane ring, which structure is known to be ruptured easily. Thus CH 3 - CH - CH 2 X CH 3 - CH - CH < CH 2 H CH 2 H CH 3 - CH - CH 2 CH 3 - CH - CH 2 - CH 2 CH 3 CH = CHCH 3 CH 2 ^CH 3 CH 2 CH = CH 2 CH 3 CH 3 (a) >CHCH 2 Br - >C CH 2 CH 3 CH 3 | | CH 3 CH 3 (b) >C CH 2 + C 2 H 5 OH -- > >C CH 3 The reaction of the solvent is frequently important in such reactions. Isobutyl iodide and silver acetate give a small yield of about equal parts of isobutyl acetate and tertiary butyl acetate. Tertiary butyl iodide, however, does not give the acetate except when acetic acid is employed as a solvent. It is significant also that tertiary butyl iodide gives only isobutylene when treated with silver cyanide, oxide or cyanate, nothing resembling a so-called double decomposition reaction taking place. Tertiary butyl iodide and silver nitrate in alcohol solution gives no trace of tertiary butyl nitrate; nitric acid is not known to react with a double bond to give an alkyl nitrate, in the same manner that sulfuric acid yields alkyl sulfuric esters. 92 92 Nef, loc. cit. 72 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS In the amylene series several cases of rearrangement have been well established which are capable of a similar explanation. Thus primary iso-amyl alcohol and the corresponding halides yield chiefly trimethyl ethylene; CH 3 CH 3 / >CHCH 2 CH/ CH 3 >CH.CH 2 CH 2 X CH 3 CH 3 >C CH 2 CH 8 \/ CH 2 CH 3 CH >C CH 2 CH CH 3 CH 3 = CH.CH 3 The formation of this olefine from bromotetramethylmethane, ob- served by Tissier 93 may be similarly explained without resorting to the vague idea of the "wandering" of the methyl group. ->(CH 3 ) 2 C-CH 2 (CH 3 ) 2 C CH 2 Br - (CH 3 ) 2 C CH 9 -H CH, H > (CH 3 ) 2 C CH 2 CH 2 -* (CH,) 2 C = CH . CH 3 Another rearrangement involving a change in the position of a methyl group is that noted by Coutourier. 94 CH 3 > C CHBr CH 3 CH 3 CH 3 1 >C_-C-CH 3 2 H CH 3 CH 3 >C CH.CH 3 CH 3 |/ CH, CH 3 CH 3 CH 3 CH 3 Some support for Nef's theory of such changes is found in the properties of the cyclopropane ring (see page 77) . The mechanism of dissociation of the alkyl halides and their so- *Ann. chim. phys. (6), 29, 361 (1893). "Ann. chim. phya. (6), 26, 464 (1892). CHEMICAL PROPERTIES OF SATURATED HYDROCARBONS 73 called double decomposition reactions is of fundamental importance. Nef 95 has advanced the theory, which he has developed from his previous studies of bivalent carbon, that alkylidene dissociation first occurs. (a) RCH 2 CH 2 X > RCH 2 CH< + HX (b) RCH 2 CH< -^RCH = CH 2 Whether or not olefmes are found in the reaction products de- pends upon the presence or absence of substances capable of reacting with the very reactive alkylidene, the rate of this reaction as com- pared with the rate of the rearrangement to the olefine, and other secondary factors. Thus Nef explains the apparently contradictory results obtained by previous investigators by showing that when ethyl ; chloride is decomposed by heating to 550 and the gases subsequently passed over soda lime to remove the hydrogen chloride, a nearly quan- titative yield of ethylene was obtained. If, however, ethyl chloride is passed directly into hot soda lime at 550 ethyl alcohol or rather I the decomposition products of ethyl alcohol under these conditions, I acetate, carbonates, methane and hydrogen, are obtained. Hydrogen chloride acting upon the soda-lime liberates water, which may then react with the labile, reactive alkylidene as follows: (a) CH 3 CH< + HOH (b) CH 3 CH< + H 2 The behavior of the simpler alkyl halides to alcoholic alkali has been thoroughly investigated by Nef, with the results summarized be- low: Halide Olefine % Ether % Temp. C. C.HsBr 11. 60. -70. 70 5 C 2 H 5 I 14. 60. 40- 90 CH 3 CH 2 CH 2 Br 20. 60. 80-100 CH 3 CH 2 CH 2 I 36.4 40. 80-100 CH 3 CHBr.CH 3 75.0 17. 80-100 CH 3 CHI.C 3 H 93.6 0.? 80-100 (CH 3 ) 2 CHCH 2 C1 .... 37. 120 (Sodium " 38 5 170 I isobutylate used (CH 3 ) 2 CHCH 2 Br 64. 23. 90-100 (CHa) 2 CHCHJ 98. 0. 90-100 CH, (CH 3 ) 2 C< 97. 0. 90-100 a "Ann. 309, 128 (1899) ; SIS, 3 (1901). 74 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Halide Olefine % Ether % Temp.C. CH 3 (CH 3 ) 2 C< 97. 0. 90-100 I (CH 8 ) 2 CH.CH 2 CH 2 Br .... 70.5 90-100 (CH 3 ) 2 CH.CH 2 CH 2 I .... 51. 90-100 C 2 H 8 (CH 3 ) 2 C < 80. ? 50- 60 Br CHJ3r.CHj.Br > vinyl bromide, quantitative. Vaubel 96 has shown that allyl halides give chiefly allyl ether with alcoholic alkali under a wide variety of conditions. (For the influ- ence of double bonds upon the reactivity of adjacent halogen atoms see page 000.) Nef 97 has also shown that the alkyl sulfates, ethyl, n . propyl, isobutyl and iso-amyl, react with alcoholic caustic potash to give mainly the ethers ROC 2 H 5 . On heating the alkyl halides with water, alcohols and olefines are formed. The employment of high pressures during the hydrolysis greatly increases the yield of alcohols from chloropentanes. 98 Acetates are formed when the alkyl halides are heated with an ace- tate of sodium, potassium, silver or lead and the best results appear to be obtained in glacial acetic acid under pressure. As with ether for-; mation noted by Nef the best yields of alkyl acetates are obtained from the alkyl chlorides, iodides giving the poorest yields. This well- known method is of general application. It is applied industrially in the manufacture of artificial amyl acetate and also in the terpene series in the conversion of bornyl chloride into the acetates of borneol and isoborneol. 99 The alkyl halides and metallic nitrates give very small yields of alkyl nitrates. Thus Bertrand, 100 with methyl, ethyl and propyl iodides and silver nitrate obtained free nitric acid, and small quantities of ethers and alkyl nitrates. Tertiary butyl iodide and alcoholic silver nitrate yield isobutylene and tertiary butyl ethyl ether in about equal amounts. 101 Ethylene bromide and alcoholic silver nitrate gives a trace only of the dinitrate, a little free nitric acid, some glycoldiethyl ether and the chief reaction product is the ethyl ether mononitrate CH 2 OC 2 H 5 H 2 ON0 2 *Ber. 24, 1685 (1891). " Ann. SIS, 3 (1901). t + M *^ "Essex, Hibbert & Brooks, J. Am. Uhem. Soc. 38, 1369 (1916). 99 Camphene is the principal reaction product. Bull. Soc. Chim. S3, 566 (1881). ""Nef, Ann. S09, 150 (1899). CHEMICAL PROPERTIES OF SATURATED HYDROCARBONS 75 Alkyl halides, particularly chlorides, can be converted into the cor- responding alcohols by heating with alkali formate in methyl alcohol solution. Henry 102 first noted the ease with which certain alkyl for- mates react with methyl alcohol to give methyl formate and an alcohol. Nef prepared acetol in this manner and excellent yields of ethylene- glycol can be obtained from ethylene chloride. 103 (1) RCH 2 Cl + Na0 2 CH- ->RCH 2 2 CH (alkyl formate) (2) RCH 2 2 CH + CH 3 OH - -> RCH 2 OH + CH 3 O 2 CH + NaCl There is no appreciable difference in the behavior of alkyl and non- benzenoid cyclic halides toward magnesium and in the various appli- cations of the Grignard reaction. To cite a few examples among many, Borsche used the Grignard synthesis of sulfinic acids to convert cyclo- pentyl bromide into cyclopentanesulfinic acid, 104 and Bouveault used bromocyclohexane in the preparation of cyclohexanol. 105 Hesse has patented the conversion of bornyl chloride to borneol by the use of the Grignard reaction, 106 but in this case, as with the higher alkyl halides, the yields are very poor. Bromocyclohexane, like normal and iso- hexane monobromides, is unstable. Alcoholic caustic potash yields mainly cyclohexene. 102 Bull. acad. roy. *elg. 1902, 445. 108 Brooks & Humphrey, J, Ind. & Eng. CJtem. 9, 750 (1917). 104 Ber. 40, 2220 (1907). 106 Bull. soc. cMm. (3), 29, 1049 (1903). 106 U. S. Pat. 826,165; 826,166. Chapter III. The Paraffine Hydrocarbons. Methane. Methane is described in a special section on account of its com- mercial importance. One liter of methane (made by the action of water on magnesium-methyl iodide) weighs 0.7168 grams at and 760 mm. pressure. 1 Its melting-point is 184. 2 Its boiling-point under 760 mm. pressure is 164. 3 The critical temperature is 82.85, the critical pressure 45.60 atmospheres, and the critical den- sity 0.1623. 4 The coefficients of expansion x 10 6 are A = 3687 and B = 3681. 5 The liquefaction of methane has recently become of industrial im- portance in connection with the separation of helium from natural gas. Pure methane may be separated from ethane and other hydrocarbons in this manner, which is a matter of some importance in the industrial chlorination of methane. Although both the Linde and Claude proc- esses have been employed on a large scale for this purpose, little tech- nical information has been published. Satterly and Patterson 6 have determined the latent heat of vaporization of methane to be 130 calories per gram and ethane 260 calories per gram. Satterly 7 has shown that nitrogen dissolves in liquid methane at moderate pressures and Mc- Taggart and Edwards 8 have determined the temperature and compo- sition relations in the liquid and gas phases in the system methane- nitrogen. The flame of methane is not very luminous. When burned in an Argand burner at the rate of one cubic foot per hour it gives a flame of 5.2 candle power. Pure methane on combustion yields 1003 B.T.U. 1 Guye, Chem. Zentr. 1909 I. 977. 4 Baume & Perrot, compt. rend. 148, 39 (1909) ; also Wahl, Proc. Roy. Soc. 87A, 371. Moisaan & Chavanne, Compt. rend. Itf, 407 (1905) ; Olszewski, Compt. rend. 100, 'Cardoso, Arch. aci. phya. nat. 36, 97, 39, 400. Leduc, Compt. rend. 148, 173 (1909). Trans. Roy. Soc. Canada. 13. 123 (1919). 7 Ibid., IS, 109 (1919). Ibid., IS, 67 (1919). 76 THE PARAFFINS HYDROCARBONS 77 per cubic foot. 9 Values for natural gas vary from 950 to about 1250 B.T.U. per cubic foot. Methane has no physiological effect on men or animals except when present in sufficient per cent to produce the characteristic symptoms of oxygen deficiency. Mine gas and other mixtures of methane and air may, therefore, contain sufficient methane to form explosive mixtures and yet cause no physiological symptoms which might serve as a warn- ing to miners. Haber 10 has developed an interesting automatic warn- ing whistle. 'The largest explosive limits for methane and air are those deter- mined by Burrell and Oberfell, i. e., a minimum methane content of 4.9 per cent and a maximum of 15 to 15.4 per cent. 11 Initial pressures of 5 atmospheres do not appreciably effect these ratios, so that these values are practically independent of ordinary variations of barometric pressure. Burgess and Wheeler, 12 and Wheeler 13 find somewhat nar- rower limits. 14 Wheeler 15 also finds that moderate changes of pres- sure have only very slight effects on the explosion limits. Coward, Carpenter and Pay man, 16 give 5.6 per cent methane as the lower limit of explosibility. Methane and oxygen ignite at 667 and although this ignition point is somewhat lowered by certain metals, oxidation in the presence of palladium is not appreciable below 404. 17 This fact makes possible the quantitative determination of hydrogen in the pres- ence of methane by selective combustion. 18 Richards, "Metallurgical Calculations," 1918, p. 25, gives 970 B. T. U. per cubic foot as the net heat of combustion of methane: ethane 1719 B. T. U. and propane 2464 B. T. U. per cubic foot. 10 (The U. S. Bureau of Mines has recently demonstrated a highly efficient system of warning miners of danger by introducing butyl mercaptan in the air supply.) The Haber apparatus for the detection of methane in mine gases gives warning as the percentage of methane approaches the limit of explosibility. It is based on the principle that differences in the density of a gas are indicated by differences in the sound produced by blowing a whistle or pipe with the gas. The apparatus contains two stopped pipes, which are tuned to the same pitch when filled with the same gas. When one whistle is supplied, by piped connections, with a mixture of methane and air in the proportions corresponding to the lower explosive limit, and the other supplied with the mine air, then the simultaneous blowing of the two whistles produces a beat whose interval diminishes as the pitch of the two pipes approach the same value, or as the mine gas approaches the dangerous composition gas in methane content. When near the explosion limit the beat produces a charac- teristic shrill sound. Cf. Chem. Ztg. 57, 1329 (1913). 11 U. S. Bureau of Mines. Techn. Paper 119 and 121 (1916). 12 J. Chem. Soc. 105, 2591 (1914). 13 J. Chem. Soc. 105, 2606 (1914) ; also Mason & Wheeler, J. Chem. Soc. 113, 45 ( 1918) . 14 Mixtures of methane and air containing 9.6 per cent methane are the most flammable, and the rate of flame travel and explosion violence is greatest with mixtures of this composition. Methane and oxygen, in molecular proportions, gives a flame velocity of 7,616 feet per second. Mason & Wheeler [J. Chem. Soc. 117, 1227 (1920)1 give 5.4 per cent as the lower limit of methane and air mixtures for horizontal flame propagation. 16 J. Chem. Soc. Ill, 411 (1917). 18 J. Chem. Soc. 115, 28 (1919). "Denham, J. Soc. Chem. Ind. 24, 1202 (1905); Phillips, Am. Chem. J. 16, 163 (1894) . "Hempel, Z. anal. Chem. SI, 445 (1902); Richardt, Chem. Zentr. 1904, II. 364. 78 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The mechanism of the combustion of methane and other hydro- carbons has been studied by Bone and Wheeler 19 who found that formaldehyde is an intermediate product. Formaldehyde is then fur- ther oxidized, with the possible intermediate production of formic acid, to water and carbon dioxide. They represent the combustion of me- thane as passing through the stages indicated in the following: CD (2) (3) (4) 2CO + 2 2C0 2 Methane is exceptionally stable to heat. Bone and Coward 20 have shown that its decomposition at 700 is not appreciable, but at slightly higher temperatures it is decomposed directly into carbon and hydro- gen without the formation of ethylene or acetylene. Coward and Wil- son 21 showed that at 850 the equilibrium mixture contains 97.5 per cent hydrogen and 2.5 per cent methane. At 1000 the equilibrium mixture consists of 1.1 per cent methane and 98.9 per cent hydrogen. At 1200 Pring and Fairlie 22 found a gas mixture in equilibrium with amorphous carbon containing 0.36 per cent methane. The carbon formed by decomposing methane in hot tubes, hot furnace checker work and the like is not a good commercial black but is gray-black in color and usually gritty. Whitaker and Alexander 23 have called attention to the fact that in gas mixtures produced by the thermal decomposition of hydrocarbons, equilibrium corresponding to the temperature em- ployed is rarely, if ever, attained. The composition of the gas is not only dependent upon the temperature to which the mixture is sub- jected, but is also markedly affected by the time of heating, the pres- sure and the presence or absence of substances which may catalytically influence the tendency to establish equilibrium. When methane is decomposed in contact with metals, metallic car- bides are sometimes formed; in fact, it has been proposed to intro- J. Chem. 8oc. 81, 541 (1902) ; 83, 1074 (1903) ; Cf. Armstrong, J. Chem. Soc. 83, 1088 (1903). 20 J. Chem. Roc. 93, 1197 (1908). ai J. Chem. Soc. 115, 1380 (1919). **J. Chem. Soc. 101, 91 (1911) ; Bone & Jordan, J. Chem. Soc. 11, 41 (1897) : 79, 1042 (1901). u J. Ind. d Eng. Chem. 6, 383 (1914). THE PARAFFINS HYDROCARBONS 79 duce carbon into molten iron in this manner. Magnesium carbide is rapidly formed by heating the metal with methane at 760. Man- ganese also readily forms a carbide when heated to 800 in methane. 24 Chlorination of Methane: The industrial production of carbon tetrachloride, methyl chloride, chloroform and dichloromethane from methane or natural gas, is peculiarly an American opportunity on ac- count of the availability of natural gas. No process for the manu- facture of methane, as by the hydrogenation of carbon monoxide, has as yet been operated on an industrial scale. The problem of manufac- turing these chlorinated derivatives is an old one but recent interest in this direction is coincident with the steadily increasing value of the products of wood distillation, particularly methyl alcohol and acetone, and the rapid development of the electrolytic chlorine industry and relatively cheap liquid chlorine. Obviously, the maximum economic advantage would be secured by bringing natural gas and electrolytic chlorine production together. As pointed out elsewhere, natural gas varies considerably in the proportions of methane and other hydrocar- bons, but so-called dry gases containing very low percentages of ethane and higher methane homologues are widely distributed. According to reported analyses 25 such dry gas is available at numerous locali- ties in the Louisiana, Texas and California fields and, as has already been pointed out, pure methane can be separated from its homologues by liquefaction methods so that West Virginia or other gas could thus be employed. Chlorine and methane do not react in the dark at ordinary tem- peratures but Bedford 26 states that fairly good yields of methyl chlo- ride and carbon tetrachloride may be obtained, without explosions, by chlorinating at in strongly actinic light. Baskerville and Ried- erer 27 state that ultraviolet light has very little effect upon the re- action but that intense illumination by light strong in the visible blue rays is much more effective. Philips 28 heated the chlorine-methane mixture and prevented explosions by packing the heated zone with sand, asbestos or bone black, 29 very similar to the method of smoothly chlorinating acetylene. At 300 to 400, in the dark, the principal products are methyl chloride and carbon tetrachloride. Tolloczko "Hilpert & Paunescu, Ber. 46, 3479 (1913). 26 Cf. U. S. Bur. Mines. Techn. Paper #255,-ll, (1921). "Chlorination of Natural Gas" by Jones, Allison & Meighan. 28 J. 2nd. & Eng. Chem. 8, 1090 (1916). 27 J. Ind. & Eng. Chem. 5, 5 (1913). 28 Aw. Chem. J. 16, 361 (1894). 29 Yoneyama & Ban [J. Chem. Soc. Aba. 1321, I. 3] use bone black and fine calcium oxide at 250. 80 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS and Kling 30 obtained a yield of 78 per cent carbon tetrachloride by 'chlorinating at 400 in contact with pumice, and impregnation of the pumice with cupric chloride is said to favor smooth chlorination. Chlo- rination at 400 is also described by Mackaye. 31 The effect of cat- alysts upon these reactions is of interest, particularly as regards in- creasing the yield of partially chlorinated products. Passing a mix- ture of the two gases through active charcoal at 90 was proposed by Mallet 32 in 1879 and Damoiseau 33 states that methyl chloride may be chlorinated mainly to chloroform by passing the proper gas mix- ture through animal charcoal heated to 250-350. Garner and Clay- ton 3 * have recently patented a similar method, employing a specially activated charcoal as the catalyst. Recent experiments of Jones, Alli- son and Meighan 35 indicate that the carbons, particularly anthracite activated by steam at 700 F., are much more effective than silicious porous substances, such as pumice, asbestos, silica gels, porcelains and glass wool. Although the work of these investigators and others shows that chlorination occurs somewhat below 300 in the absence of cat- alysts, they employed temperatures within the range 375 to 400 in nearly all of their experiments with catalysts. Ferric chloride and antimony pentachloride give poor results 38 but the work of the U. S. Bureau of Mines indicates that coke impreg- nated with iron or nickel gives the highest yields of chloroform, that activated carbons give the best yields of carbon tetrachloride and that coke impregnated with nickel, tin or lead gives slightly better yields of methyl chloride, using larger proportions of methane in the latter case. A total yield of about 90 per cent of chlorinated products, based upon the gas used, can be obtained. Methyl chloride boils at 23.7, melts at 103 ; its critical tem- perature is 143, critical pressure 66 atmospheres, critical density 0.37. The densities and vapor pressures are given in the following table. 87 The latent heat of evaporation at C. is 176 B.T.U. per Ib. or 98 kilogram-calories per kilogram, or 4.94 kilogram-calories per gram molecule. 37 The reduction of carbon tetrachloride to chloroform by zinc and a *J. Soc. Chem. Ind. 32, 742 (1913). 11 U. S. Pat. 888,900. 82 U. S. Pat. 220,397. "Loc. cit. M Pfeifer, Mauthner & Reitlinger, ,J. orakt Chem (2) w 23Q no-iot Hoist, Refrigerating World, 1919, May^p 13 W ' ' 9 19) ' THE PARAFF1NE HYDROCARBONS 81 DENSITY AND VAPOR PRESSURE OP METHYL CHLORIDE. Density, referred Pressure, absolute C to water at 30 F. in atmospheres -50 0.27 -40 1.024 0.47 -22 1.008 0.76 -11 1.000 1.00 - 4 0.991 1.16 - 0.987 1.27 +14 0.972 1.73 +32 0.995 2.49 40 0.945 2.91 50 0.936 3.51 60 0.925 4.20 68 0.915 453 90 0.892 6.91 100 0.883 7.96 little aqueous hydrochloric acid 38 and by finely divided iron in the presence of water 39 has been carried out and some such method ap- pears to offer the best solution thus far proposed for the problem of manufacturing chloroform from methane. Tne conversion of methane into hydrocyanic acid by passing a mix- ture of methane, hydrogen and nitrogen through an electric arc has been tried out on an industrial scale but details of the process have not been published. 40 Rideal and Taylor have reviewed the hydrogenation of carbon monoxide to methane. 41 The industrial operation of the process would make illuminating gas much less toxic and increase its calorific value. The process might be of value in localities where natural methane is not available and where the methane could be utilized for a special purpose, for example, the manufacture of chlorinated methane prod- ucts. Elworthy 42 proposed to remove the carbon dioxide from water gas, add hydrogen sufficient to form the mixture, CO -J- 3H 2 , and effect the conversion to methane by passing over catalytic nickel at 250. At one time Sabatier 43 attempted the industrial solution of the problem in a somewhat different manner. He noted that the carbon deposited from the conversion of CO to C0 2 and carbon at 500 in the presence of nickel, readily reacts with steam to form C0 2 and methane. By superposing the two reactions, passing water gas and superheated steam over the catalyst at 500, mixtures consisting essentially of methane, hydrogen and carbon dioxide were produced. Reduced nickel at 250- 38 Chem. Rev. 1896. 88. "A. W. Smith, U. S. Patent, 753,325 (1904). Chem. Abs. 1914, 1659. ""Catalysis in Theory and Practice." 1919. p. 182. Brit. Pat. 12,461 (1902) ; 14,333 (1904). "French Pat. 355,900 (1905). 82 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS 300 in the presence of an excess of hydrogen has been found most ef- fective, 44 but in practice considerable difficulty was experienced by poisoning of the catalyst by substances containing sulfur, 45 and the de- position of carbon on the catalyst and it was also found that at least five volumes of hydrogen are required for one volume of carbon mon- oxide. Carbon dioxide is reduced to methane in the presence of nickel quite rapidly at 350. 46 The necessary excess of hydrogen can be ob- tained by the catalytic conversion of CO and steam to hydrogen and C0 2 and removing the latter, or by partially separating the carbon monoxide and hydrogen of water gas by liquefaction methods. Bed- ford finds that when the liquefaction process is carried out so that the uncondensed portion contains approximately 14 per cent carbon mon- oxide, the sulfurous impurities are removed with the liquefied CO and the resulting mixture has no appreciable poisoning effect on the cat- alyst. Bedford carried out the reaction in quartz tubes at 280-300 and owing to the strongly exothermic character of the reaction, CO + 3H 2 CH 4 + H 2 + 48,900 calories, the reaction maintains itself without external heating. In order to prevent the deposition of carbon the concentration of carbon monoxide was kept below 17 per cent, the resulting gas containing 28.3 to 31.8 per cent methane. By successive additions of carbon monoxide and repassage over the cat- alyst a gas mixture containing 76 per cent of methane can be obtained. Meredith 47 states that it is difficult to prevent the formation of nickel carbonyl in this process, although, as is well known, the decomposition of nickel carbonyl is rapid at temperatures as low as 200 C. Ethane: The simple derivatives of ethane are quite familiar to all organic chemists and their reactions have been most frequently em- ployed as type reactions in text books of organic chemistry. Yet ethane itself has never been a product of industrial interest, and the hydro- carbon has not been employed as the raw material for the manufac- ture of those derivatives which are so important. For example, ethyl- ene, ethyl chloride and ethyl ether are all manufactured from ethyl alcohol. Changed economic conditions conceivably may change a great many of these processes. That ethane can be separated in quite a pure state from methane and propane, was first shqwn, in an analytical way, by Burrell, Seibert and Robertson, 48 who made use of the large differ- 44 Jochum, J. Gaslel, 57, 73,103,124 (1914). "Gautier, Compt. rend, 150, 1564 (1910). 46 Sabatier & Senderens, Compt. rend, 13L, 514. 689 (1902) : Farbwerke M. L. & Br. Brit. Pat. 146,110; 146,114 (1920). 47 Gas Age, 47, 7 (1921). 48 U. S. Bureau of Mines, Techn. Paper 104 (1915). THE PARAFFINS HYDROCARBONS 83 ences of th'e vapor pressures of these several hydrocarbons at low tem- peratures. By the Linde or Claude methods of fractional distillations at low temperatures these gases may be easily separated ; in fact, the separa- tion of nitrogen and oxygen is considerably more difficult. Reference to the boiling-points of these several substances indicates this possi- bility. Boiling-Point, at 760 mm. Difference Nitrogen 195.84) 1Ofi , Oxygen -182.99 \ Methane 160. 7ft7 o Ethane 49 89.3 f Propane 80 44.1 452 The following vapor pressure curves of liquid ethane were deter- mined by Burrell and Robertson, Fig. I, and by Maas and Mclntosh, Fig. II. Natural gas and possibly oil gas and petroleum still gases ETHANE -/oo* -//ol VAPOR FIG. 2. 70 contain ethane in quantities sufficient for its separation on a large industrial scale; the high ethane content of many samples of natural gas, as determined by the old method of combustion and calculation, is undoubtedly inaccurate, as has been previously pointed out, but 10 49 This value found by Burrell & Robertson, J. Am. Chem. Soc. 37, 1893 (1915) ; Maas & Mclntosh give 88.5 as the boiling-point of ethane, J. Am. Chem. Soc. 36, 10, 497 (1913) ; found the value 84.1 ; give the value 88.3. (1915). Aiaas & Aicintosn give s.o- as tne Domng-pomt or 737 (1914) ; Cardoso & Bell, J. cMm. phys. 10, 497 (1 Maas & Wright, J. Am. Chem. Soc. ^3, 1102 (1921) giv M BurreU & Robertson, J. Am. Chem. Soc, 37, 2188 84 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS to 12 per cent ethane is not uncommon. Practically pure ethane, sepa- rated in this manner, and cheap chlorine, present to organic chemists a great opportunity. The manufacture of ethyl chloride, ethylene, ethyl ether and ethyl alcohol from ethane is entirely feasible by meth- ods now known but which are capable of great improvement. The chemical properties of ethane are nearly identical with those of methane ; it is less stable to heat and in contact with metallic nickel 800 TEMPER/JURE - DEGREES FIG. 1. at 325 carbon is deposited and methane and hydrogen are formed. 51 It is absorbed by fuming sulfuric acid somewhat more rapidly than methane. 52 Slow oxidation below the temperature of actual ignition yields chiefly water, carbon dioxide, carbon monoxide and formalde- hyde. 58 It is more readily chlorinated than methane and it is note- worthy, in the light of Michael's positive and negative theory of addi- tion, that ethyl chloride on further direct chlorination yields chiefly ethylidene chlorine but in the presence of antimony pentachloride n Sabatier & Senderens, Compt. rend. 124. 1360 (1897). "Worstall, J. Am. Chem. Soc. 21, 249 (1899). "Bone & Stockings, J. Chem. Soc. 85, 696 (1904). THE PARAFFINS HYDROCARBONS 85 ethylene chloride is the principal product. 54 No researches on the chlorination of ethane have recently been published. 55 Propane: The principal raw materials utilized for the prepara- tion of propane and its simple derivatives are acetone, glycerine, tri- methylene glycol, propyl alcohol from fusel oil, and propylene from oil gas or petroleum still gases. Crude pyroligneous acid contains allyl alcohol, but no industrial use for it has been found. The hydrocarbon itself is not used as such or separated from natural gas or other gas mixtures containing it. Natural gas is the only source from which it TEMPERATURE. DEGREES could be separated in quantity. For laboratory study it may conveni- ently be prepared by the catalytic decomposition of isopropyl alcohol, over alumina at 380-400, and the catalytic hydrogenation of the re- sulting propylene to propane. 58 The vapor pressure curves of liquid propane, propylene and butane are shown in the accompanying figure. 57 The chiorination of propane does not appear to have been studied since the work of Schorlemmer 58 in 1869, who noted the formation of n. propyl chloride (?), propylene chloride and more highly chlorinated products. Although the monochlorides of methane, ethane, pentane, and probably propane and butane can be converted into the corre- "D'Ans. & Kautzsch, J. prakt. chem. (2), 80, 3ie (1909); V. Meyer & Miiller, Ber. Zit, 4247 (1891) ; Kronstein, Ber. 5$B, I. (1921). 65 Cf. Lacy, U. S. Pat. 1,242,208 : chlorinates above 300. 88 Sabatier & Senderens, Compt. rend. 1*4, 1127 (1902). "Burrell & Robertson, J. Am. Chem. Soc. SI, 2188 (1910). 68 Ann. 152, 159 (1869). 86 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS spending alcohols with good yields and ethylene chloride may be con- verted satisfactorily to the glycol, the more highly chlorinated deriva- tive 1, 2, 3, trichloropropane behaves very differently. Caustic alkali yields p-epidichlorohydrine CHC1 = CH . CH 2 C1 and a-derivative CH 2 = CC1 . CH 2 C1, and alcoholic potash gives ethylchloroallyl ether, C 2 H 5 . C 3 H 4 C1. From what is known of the halogen derivatives of propane, it is very improbable that glycerine will ever be manufactured industrially by their means. Glycerine can be synthesized by adding HOC1 to allyl chloride and hydrolysing the product, but when it is attempted to prepare allyl chloride by decomposing propylene chloride, the principal products are found to be a and |3 chloropropylene. Butanes: Normal butane was made by Frankland in the attempt to isolate the hypothetical ethyl radical, by the reaction of ethyl iodide and metallic zinc. It has been prepared in a very pure state by Le- beau by treating n. butyl iodide with sodium amalgam in liquid am- monia. 59 Its boiling point at 755 mm. is 0.5 ; critical temperature 151 to 152. At 17 and 772 mm. pressure one volume of water dissolves 0.15 volumes; chloroform at 17 and 768 mm. dissolves 32.5 volumes of the gaseous hydrocarbon. The most convenient source of n . butyl compounds is n . butyl alco- hol from which a large number of simple derivatives may readily be prepared. 60 This alcohol is now a common commercial article, being obtained together with acetone by fermenting starch with a mould, Amylomyces rouxii studied by Fernbach and Strange, and by bacteria, probably Bacillus granulobacter pectinovorum, the latter process being developed by Weizmann. 61 The butyl alcohol produced in ordinary alcoholic fermentation and appearing in the fusel oil fore-runnings is isobutyl alcohol, (CH 3 ) 2 .CH.CH 2 OH. The most convenient source of crude butane is the very light gaso- line separated from natural gas or "casing head" gas. Garner and Cooper have described the isolation of crude butane from this source by the application of principles now well known in the industry. 62 Butlerow 63 pointed out that two isomeric butanes were possible and synthesized isobutane by treating acetyl chloride with zinc methyl, according to the well-known Butlerow synthesis, forming the car- "Bull. Ac. Roy, Belg. 1908, 300. w Kamm & Marvel, J. Am. Soc. 1920, 299. 91 Speakman, J. Soc. Chem. Ind. 38, 155 (1919) ; Weizmann, Brit. Pat. 4,845 lJS 1 ^ ) o ; -,H ernb ^ c ^ & stran & 6 ' Brit - Pat. 14,607 (1915) ; Fernbach, Biit. Patents, 109,- 960 (1917) ; 15,203, 15,204 and 16,925 (1911). TT. S. Pat. 1, 307,353 (1919). "Ann. m, 1 (1867). THE PARAFFINS HYDROCARBONS 87 CH 3 Isobutane : > CH CH 3 CH 3 binol (CH 3 ) 3 C.OH. This was converted into the iodide which on re- duction with zinc in the presence of water gave isobutane, an octane and isobutylene. Isobutane boils at 10.5 under 757 mm. ; its critical temperature is 134 to 135 . 6 * The butanes are readily chlorinated by moist chlorine at ordinary temperatures. 65 Bromine reacts much less readily and on heating with bromine in a sealed tube, it is decomposed forming tetrabromethylene, Br 2 C = CBr 2 as one of the products. 66 Isobutylene readily combines with hydrogen iodide to form tertiary butyl iodide. The Pentanes: Both normal and isopentane occur in petroleum, at least in certain petroleums which have been carefully examined. The difficulty of separating these two hydrocarbons by fractional distilla- tion is well shown by the work of Young, 67 whose results are ex- pressed by the following figure: Thirteen very careful fractional distillations and the use of a very effi- cient fractionating column were required to isolate these two hydro- carbons in fair degrees of purity. The vapor pressure curves of butane n.pentane, n.hexane, n. hep- tane and n . octane are given in the following table : 68 The pentanes are chlorinated very much more readily than methane and ethane, i. e., at in diffused daylight. The relative stability of 84 Lebeau, loc. cit. 5 Mabery & Hudson, Am. Chem. J. 19, 244 (1897). "Weith, Ber. 11, 2244 (1878). <"/. Chem. Soc. 73, 906 (1898). "Anderson, J. Ind. Eng. Chem. 12, 547 (1920). 88 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS the halogen derivatives of the series methane to pentanes inclusive has already been noted. No satisfactory method of preparing these simple hydrocarbons seems to have been developed. The preparation of n.pentane by heat- ing acetyl acetone with concentrated hydroiodic acid to 180 and by heating pyridine with the same reagent to 300 has been suggested. The action of amyl or iso-amyl bromides on magnesium, in ether, probably yields a little amylene and decanes, 69 but since n.pentane or isopentane can easily be separated from these by-products, the method of decomposing amyl or iso-amyl-magnesium bromides by water would undoubtedly prove the most satisfactory method of preparing these hy- drocarbons in a pure state. By the reduction of iso-amyl iodide by zinc fairly pure isopentane can be prepared, 70 and Ipatiev made isopen- tane 'by the hydrogenation of trimethyl ethylene. 71 Tetramethylmethane, C(CH 3 ) 4 , was prepared by reacting upon 2.2-dichloropropane with zinc methyl, CH 3 CH 3 CH 3 >CCl 2 + Zn(CH 3 ) 2 >C< +ZnCl 2 CH 3 CH 3 CH 3 This hydrocarbon 72 is remarkable for its relatively low boiling point, 9.5, and relatively high freezing-point, 20. 69 Cf. Tschelinzeff, J. Russ. Phys.-Chem. Soc. 36, 549 (1903) ; Tiffeneau, Compt. rend, m, 481 (1904). Frankland, Ann. 74, 53 (1850) ; Just, Ann. 220, 152 (1883). "Ber. 42, 2089 (1909). "Lwow, Z. f. Chetn. 1810, 520. THE PARAFFINS HYDROCARBONS 89 Aschan 73 has studied the chlorination of the pentane and hexane fractions of petroleum and also the chlorination of isopentane, which hydrocarbon, Aschan claims, is present in all petroleums. The best yields of monochloropentanes are obtained by chlorinating with dry chlorine but moist chlorination leads chiefly to the formation of the two possible primary chlorides, small proportions of secondary chlo- ride and no tertiary chloride. Dry chlorination yields all four pos- sible monochlorides the properties of which are given by Aschan as follows: Boiling-Point . D^-r~ lo 4-chloro-2-methyl butane 99. -102. 0.8692 3-chloro-2-methyl butane 90. - 93. 0.8752 l-chloro-2-methyl butane 96. - 99. 0.8818 2-chloro-2-methyl butane 85.5- 88. 0.8692 The primary iso-amyl chloride made from natural fusel oil is con- verted almost quantitatively to the acetate and alcohol by heating with alcoholic potassium acetate at 200. Isopentyl chloride is only very slowly acted upon by 2 per cent caustic potash at 60-70. The Hexanes: Like the pentanes, normal hexane may readily be separated, in an impure state, from light petroleum distillates, and the preparation of pure n . hexane depends upon standard laboratory meth- ods such as the reduction of secondary hexyl iodide (made from man- nite) or the condensation of normal propyl iodide by metallic sodium. 74 Like pentane it chlorinates readily and it also reacts rapidly with bromine in sunlight. The mixture of monochlorides contains about 10 per cent of the 1-chloro compound and about 45 per cent each of the 2 and 3 chloro derivatives. 75 Its reactivity to the halogens, to fuming sulfuric acid, to nitration by the dilute nitric acid method, and the properties of the simple derivatives, chlorides, .alcohols, amines, carboxylic acid derivatives, etc., is almost identical with the chemical behavior of cyclohexane. The Heptanes: Normal heptane enjoys the distinction of being the only saturated hydrocarbon, other than the solid paraffines formed by phytochemical processes. It is one of the major constituents in the volatile oil of the two American pines, Pinus sabiniana and Pinus jeffreyi, and also occurs in the "petroleum nuts," Pittosporum resini- ferum, of the Philippines. How such a saturated hydrocarbon is formed in the living plant is entirely obscure; it is accompanied by Chem. Abs. H, 3654 (1920). T * Michael, Am. Chem. J. 25. 421 (1901). "Michael & Turner, Ber. S9 t 2153 (1906). 90 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS no other substances, so far detected, which conceivably could have been the parent substance. Unsaturated hydrocarbons, the terpenes, are undoubtedly formed from alcohols, and it is well established that such decompositions occur in the leaves of plants. Pine needle oils commonly contain borneol and other alcohols although the oleoresins of these pines, when finally secreted in the resin ducts of the stem, con- tain only resin acids and unsaturated hydrocarbons. The fact that this particular hydrocarbon is one of an odd number of carbon atoms is also most unusual, since by far the great majority of the hydrocarbons, sugars and alcohols, fatty acids and ketones occurring in plants contain an even number of carbon atoms. Normal heptane probably occurs in most light petroleum fractions, as in commercial gasoline. 76 Its separation from gasoline, however, in a reasonably pure state is a matter of the greatest difficulty. The raw material which has been employed for the preparation of the best known derivatives of n. heptane is oenanthol or n.heptyl aldehyde. 77 This aldehyde undergoes the usual aldehyde reactions, yields 1 . 1-di- chloroheptane by treatment with PC1 5 , and on reduction gives n.heptyl alcohol from which n.heptyl chloride can be made by the action of hydrogen chloride. This is the only one of the four possible monochlor derivatives of n . heptane which has been prepared in reasonable purity and identified as such. Only three of the possible 17 dichlorides are known, i. e., the 1-1, 4-4 and 1-7 derivatives. Until the discovery of the Grignard reaction which serves to build up any desired carbon structure up to 10 carbon atoms and with limitations, larger mole- cules, and also the discovery of catalytic hydrogenation by which means unsaturated hydrocarbons may be readily converted at low tem- peratures and in neutral reaction media, to saturated hydrocarbons, our knowledge of hydrocarbons of the paraffine series containing more than six carbon atoms was very limited indeed. The physical properties of the known isomeric heptanes are as follows: Name Structure Boiling-Point Density n. Heptane CH 3 (CH 2 ) 5 CH 3 98.2-98.5 0.7006- 2-Methylhexane (CH 8 ) a CH.(CH 2 ) 3 .CH 3 89.0-90.4 0.7067- ^ 9ft 3-Methylhexane C 3 H 7 .CH(CH 3 ).C 2 H 5 90. -92. 0.6865-^ "Young, J. Chem. Soc. 73, 906 (1898) ; Engler & Hofer, "Das Erdol," Vol. I, 244 (1913). 77 This aldehyde is readily prepared by the well known method of destructive distillation of castor oil, enanthol and undecylenic acid being formed. THE PARAFFINE HYDROCARBONS 91 Name Structure Boiling-Point Density 3-Ethylpentane CH.(C 2 H S ), 95. -98. 0.689 -27 2, 2-Dimethylpentane (CH^CAH, 78. 0.6910 2, 4-Dimethylpentane (CH 3 ) 2 CH.CH 2 .CH.(CH,) 3 83. -84. 0.7002 5! 3, 3-Dimethylpentane (CH3) 2 C(C 2 H B ) 2 86. -87. 0.7111 The Octanes: Normal octane probably occurs in most light petro- leum distillates, or gasolines. It is most readily prepared in a pure state by treating n. butyl iodide with sodium, 78 a reaction which is said to give better yields with alkyl halides of higher molecular weight than with the simpler ones. As typical examples of methods which may be employed in the synthesis of hydrocarbons, the methods employed in the preparation of the known octanes are given in reaction outline, as follows: (1) Normal octane. 78 2CH 3 CH 2 CH 2 CH 2 I + 2Na - > Nal + CH 3 (CH 2 ) 6 . CH 3 (2) 2-Methylheptane. CH 3 CH 2 CH 2 CH 3 >CH.CH 2 CHO + Mg< - > CH 3 X CH 3 > CH . CH 2 CH . CH 2 CH 2 CH 3 CH 3 | OH - > corresponding iodide - > 2-methyl-heptane, by reduction with copperized zinc and hydrochloric acid. (3) 8-Methylheptane* CH 3 CH,CH 2 I + CH 3 COCH.C0 2 C 2 H 5 - Na 10% KOH CH 3 COCH . C0 2 C 2 H 5 - CH 3 COCH 2 CH 2 CH 2 CH, CH S C (OH) . CH 2 CH 2 CH 2 CH 3 -- corresponding iodide- C 2 H 5 3-methylheptane, by reduction as indicated above. "Zincke, Ann. 152, 15 (1869). L. Clarke, J. Am. Chem. Soc. SJ, 108 (1909). 80 IMd., 558 (1909). 92 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS (4) 4-Methylheptane. 81 CH 3 CH 2 CH 2 CH . CH 3 + CH 3 COCH . C0 2 C 2 H 5 > I Na 10% KOH CH 3 COCH . C0 2 C 2 H 5 - CH 3 CH 2 CH 2 CH . CH 2 COCH 3 ^01^ OTT f^TI OTT /^TT 3 V_yXl . V^Xl. 2 v>'JH 2 l^/Xl. 3 v_y.LJ- 3 by reduction by sodium in moist ether > alcohol -* iodide - CH 3 CH 2 CH 2 CH CH 2 CH 2 CH 3 CH 3 (5) 2 . 4-Dimethylhexane 82 CH 3 CH 3 CO . CH 2 CH . CH 2 CH 3 + Mg< -> CH 3 CH 3 C (OH) .CH 2 CH . CH 2 CH 3 CH 3 CH 3 > iodide, which by reduction >CH 3 CH.CH,CH.CH 2 CH 3 CH 3 CH 3 (6) 2.5-Dimethylhexane. (Di-isobutyl) (a) 83 2 (CH 8 ) 2 CH , CH 2 I + 2Na - (CH 8 ) 2 CH . CH 2 CH 2 CH (CH a ) 2 (b) 84 CH 3 CO.CH.C0 2 C 2 H 6 CH 3 H 2 CH< -^ CH 3 CO . CH 2 CH 2 CH . CH 3 CH 3 | CH 3 CH 3 by Mg< -> CH 3 C (OH) . CH 2 CH 2 CH . CH 3 I > hydrocarbon CH 3 CH 3 as above , 81 L. Clarke, Am. Chem. J. 89, 87 (1908) ; Ber. 40, 352 (1907) ; cf. Clemmensen, Chem. Ala. 6, 2919 (1912). L. Clarke, J. Am. Chem. Soc. SO, 1144 (1908). M Wurtz, Ann. 96, 365 (1855). 84 L. Clarke, J. Am. Chem. Soc. SI, 586 (1909). THE PARAFFINS HYDROCARBONS 93 (7) 3. 4-Dimethylhexane. 85 2CH 3 CH 2 COCH 3 (Methylethyl ketone) -> 2CH 3 CH 2 CH(OH) .CH 3 ^2CH 3 CH 2 CH.CH 3 | + 2N CH 3 CH (8) 2-Methyl, S-ethylpentane 8 * CH 3 CH 2 CH . CH . CH 2 CH 3 CHCH ; CH >CH.COCH 3 + Mg< by methods given above CH 3 CH, >CH.C(OH).CH, CH 3 CH 2 CH 3 CH, >CH.CH< tCH 3 CH 2 CH 3 )) 2.2.3. S-Tetramethylbutane* 7 (CH 3 ) 3 C.Br+ (CH 3 ) 3 C.MgBr >MgBr 2 + (CH 3 ) 3 C-C(CH 3 ) ighteen isomeric octanes are theoretically possible. The physical properties of the known octanes are as follows: Name n . Octane 2-Methylheptane 3-Methylheptane 4-Methylheptane 2 . 4-Dimethylhexane 2 . 5-Dimethylhexane 3 . 4-Dimethylhexane 2-Methyl, 3-ethylpen- tane 2.2.3.3.-tetramethyl- butane Structure CH 3 (CH 2 ) 6 CH, (CH3) 2 CH.(CH 2 ) 4 CH, Boiling-Point Density no I V^/flg CH 3 C 2 H 5 CH.CH 2 CHCH 3 CH, CH, (CH 3 ) 2 CH.CH 2 CH 2 CH(CH 3 ) 2 108.3-108.5 125.8 116. 0.7185^o- 0.7035 -[f; 117.6 0.7167^ 118. 0.7217i 109.8-110. 0.7083 15! 15 0.6993 C 2 H 5 .CH.CH.C 2 H S CH 3 CH 3 116. -116.2 0.7165 15 C CH C H (CH 3 ) 3 C.C(CH 3 ) 3 114. 106. M07. 0.7084^ E 'Remarkable for its high melting-point. 103-104. Norris & Green, Am. Chem. J. 26, 313 (1901). L. Clarke, Am. Chem J. 39, 572 (1908). 87 Henry, Compt. rend. 142, 1075 (1906). 94 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The Nonanes and Decanes: Pure normal nonane boils at 149.5 and n.decane at 173. The petroleum fraction boiling chiefly at 150- 170, therefore, consists chiefly of these two hydrocarbons, when de- rived from petroleums of the Pennsylvania type. 88 This particular fraction is a regular commercial article, being sold as a turpentine sub- stitute. Its volatility or rate of evaporation and solvent power for oils and resins is practically identical with that of turpentine. 89 CH 3 2.6-Dimethyl Octane, >CH. (CH 2 ) 3 CH.CH 2 CH 3 CH 3 | CH 3 is of interest in that it possesses the carbon structure of the aliphatic so-called terpenes, myrcene and ocimene, C 10 H 16 , and also the alcohols geraniol, linalool and citronellol and their corresponding aldehydes. Many of the terpenes proper and their alcohols are very probably re- lated genetically to these alcohols and aldehydes and it is therefore a matter of theoretical interest, what substance or substances in living plants yield alcohols or hydrocarbons of the carbon structure of 2.6-dimethyl octane. This saturated hydrocarbon has not been found in nature but is readily made by hydrogenation of myrcene or ocimene, or the alcohols geraniol or linalool. 90 Paraffines C 1Q H 22 to C 60 # 122 . Of the hydrocarbons of this series all the normal paraffines up to C 26 H 54 are known, and also a few hydro- carbons higher in the series have been definitely characterized. Most of what is known of the synthesis of the solid paraffines is found in three papers published by Krafft, 91 who employed the following meth- ods in their preparation: (1) yP* with hydriodic acid and phosphorus, in sealed Prprr ^ oeS meanS ^oium. Preparation of ketones by heating the calcium salts of fatty acids; con- InTphosphorls' ' tO dichlorides b ^ PC1 < and reduction by HI wHssCO . C 6 Hi3 _ dichloride > hydrocarbon - > 419, 482 (1897). over range ^n^eiSwy abo^^?! ^ ^ Varnish boils chiefl r of oan, linseed oil. However, the difference in tblgned wit t fresh,, oxidation or "drying" of <1908) - THE PARAFFINS HYDROCARBONS 95 Peterson 92 employed the method of electrolysing the fatty acid soaps and Mai 93 heated the barium soaps with sodium methoxide. Formates at 290-300 94 decompose to paraffines. Crystalline paraffines have been noted from a wide variety of sources but commercial paraffine is derived principally from certain petroleums and to a lesser extent from shale oil, ozokerite, and the dis- tillates obtained by the carbonization of coal or lignite at low tempera- tures. The constitution of the paraffines made by synthesis according to the methods indicated above, may reasonably be inferred from the methods employed in their preparation, but as regards the crystalline paraffines found in the various pyrolytic distillates and in natural waxes and essential oils we know practically nothing more than may be inferred from their melting-points, and these values may be very misleading. Thus Krafft 95 prepared a series of paraffines, by frac- tional crystallization, from the crude paraffine isolated from an oily distillate from lignite. On the basis of their melting points, varying from 22.5 to 76, the various crystal fractions are described as eight- een distinct substances but many of the specimens so prepared were probably mixtures. It is probable that many of these crystalline paraf- fines are not normal hydrocarbons, for example, n-eicosane, C 20 H 42 , melts at 36.7 (made by condensing n.decyl iodide by sodium) but an isomeric hydrocarbon melting at 69 has been reported from four dif- ferent natural sources. It is of interest to note the number of essential oils and other natural products which contain solid paraffines, and that most of them evi- dently bear no relation to the natural fatty acids, having many more carbon atoms than these acids. Source Melting-Point Kaempferia galanga (about 50% of the essential oil)** 10. ( Rose oil 22. j Jaborandi leaves * . " 28-29. I Rose oil .."..!.*!.*.!..* 40-41. | Birch buds '..'.'...*..'..'..'.'.'.'. 5o! ' Camomile oil 53-54] ' Orange blossoms ......!.!..... 55. I Eucalyptus oils w '.'.'.. 55-56. i Sassafras leaves 58. Bees-wax ; Virginia and Kentucky tobacco 98 ...'.'. '. '. '. '. '. '. 1 '. .........'. 59.5 82 Z. f. Electroch. 12, 144 (1906). 93 Ber. 22, 2134 (1889). * Fr. Bayer & Co. J. Chem. Soc. A6s. 1918, I, 209. Ber. 40, 4779 (1907). 66 Schimmel & Co. Semi-Ann. Rep. 1903, I, 43. Smith, Chem. Aba. 8, 399 (1914) ; in Eucalyptus acervula, E. paludosa and E. smithn. 98 Thorpe & Holmes, J. Chem. Soc. 79, 982 (1901). 96 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Source Melting-Point Verbena 62.5 Arnica, essential oil ; pelargonium 63. Camomile, Roman 63-64. Dill oil; Cistus, several species; Chrysanthemum cineraraefolium. . . 64. Wintergreen oil, from Betula and Gaultheria 65.5 Bees-wax; leaves of European olive; Kentucky and Virginia to- bacco; seeds of Brucea sumatrana; Lippia scaberrina; Micro- meria chamissonis; Grindelia robusta; Gymnene sylvestre 68.1 Eriodictyon calif ornicum ; leaves of European olive; Arthusa cyna- pium 74.7 Evodia simplex 80-81. Paraffines are also found in the mineral ozokerite, which is mined near Boryslaw, Galicia, and in Wasatch and Utah counties, Utah. Re- fining of ozokerite by concentrated sulfuric acid yields ceresine, which is valued for its relatively high melting point. When ozokerite is distilled crystalline paraffine, about 40 per cent, can be separated from the distillate, and the undistilled residue is ozokerite pitch or "oko- nite." Fractional crystallization of the solid waxes in Galician ozo- kerite gives a series of fractions the lowest melting at about 54 and the highest melting at 92.8-93.2." Practically nothing is known of the nature of these hydrocarbons in refined ceresine beyond the fact that their analyses indicate the composition C n H 2m+2 and that their chemical behavior is like that of other solid parafiines. The crystallization of paraffine is considerably affected by the vis- cosity of the oil from which it is crystallized and also the presence of asphaltic matter seriously interferes with the crystal growth. With increasing viscosity of the oil solvent the crystal size diminishes. 100 The exact nature of the so-called "amorphous wax" is not known but repeated distillation of oils containing much paraffine yields cleaner distillates from which large crystals are obtainable without difficulty. According to Rakuzin 101 crude petroleums contain crystallizable paraf- fine although its crystallization is greatly interfered with by asphaltic substances present. He is therefore opposed to Zaloziecki's views as to the presence of "protoparaffine" in crude petroleums, but there is no doubt that complex substances such as the "kerogen" of oil shale, peat and lignite, yield paraffine only when decomposed, as by heat. Paraffine is remarkably insoluble in most organic solvents. The solubility of a paraffine fraction, melting-point 64 to 65 from Ga- lician ozokerite, in petroleum ether is as follows: 102 M Engler, "Das Erdoel," Vol. I, 667. 'Cf. Fuchs, Petroleum u, 1281 (1919). ' chem - B c A *- ' THE PARAFFINS HYDROCARBONS 97 g. paraffine in Solvent 100 g. solvent Carbon bisulfide 12.99 Petroleum ether, boiling-point below 75 11.73 Acetic acid, glacial 0.06 Since petroleum ether and glacial acetic acid are miscible in all proportions, these two solvents are recommended for recrystallizing paraffine. In industrial practice the oil and low-melting wax is per- mitted to drain slowly from the crude crystals in warm chambers, i. e., the "sweating process." By several fractional distillations, at 40 mm., Mabery 103 separated commercial paraffine into several fractions, the lowest melting at 48 and the highest at 62-63. The dielectric constant of paraffine is such that large quantities are used for the purpose of electrical insulation, usually in cases where the material is not subjected to temperatures high enough to melt the wax. Comparisons of the dielectric constant of paraffine and other common insulating materials, are as follows: 104 E Paraffine, crude brown 2.07 melting-point 44-46 2.105 " 54-56 ...: 2.145 double refined 1.94 Asphalt 2.68 Amber 2.80 Shellac 3.10 Gutta-percha 4.43 Bees-wax 4.75 The specific heat of paraffine wax is a linear function of the tem- perature; at 100 it is 0.6307, at = 0.47, at 100 = 0.325 and at - 180 = 0.199. 105 The latent heat of fusion of commercial paraffine wax, calculated from the lowering of the freezing point on adding sub- stances of known molecular weight, ranges from 38.9 to 43.9 calories. 106 Paraffine wax is generally considered to be a very inert material but it is attacked by nitric acid and by sulfuric acid at slightly elevated temperatures, oxidation rather than nitration or sulfonation being the principal result. It reacts readily with sulfur on heating to about 200, evolving hydrogen sulfide ; in fact, this reaction serves as an admirable method for the laboratory preparation of hydrogen sulfide, particularly where the gas is not continually needed and the apparatus must stand 103 Cf.^m. Chem. J. S3, 251 (1905). 10 * Landolt-Bornstein, "PJiysikalisch-Chemische Tabellcn," 1912, pp. 1212. " Nernst, J. Chem. 8oc. Abs. 108, II, 263 (1910) ; Bushong & Knight, J. Ind. A Eng. Chem. 12, 1197 (1920). "Kozicki & Pilat, J. 8oc. Chem. Ind. 37, 681 A (1918). S" SfeSSte 2 3S 8 i-H rH ,< 1-1 >-^ il rH rH T 1 rH ss 'g S ^- Og^jg * > n feggi : ^ : ^ {g Jo S t>-i i> t ,Ui *G CO ^o s J n ooooo III I a I JS 98 rH^ C>d '000 o O C3-S -a 3 S3 J i i i? i (! !s I 11 * is JllH.LW 3 |s I ?tipii|i T^ W* . > ^ t- S ^ .583 .JSZ^fo ^c^-S^S^g^g M^SB^Brf^rfiirf ,:HJ gg.^CW^^^gl^fif^^l P^5^ rt o N S^ ;e*ftC -^llg-g, o8"B8*^^^ " - a ^c8^533 ^^cjS^ 3 a< N J-SosS'^ ^_^ .JHOfe^HHJ SH2S2SS525S5SS3S % O CO CO CO i 8 8" ass" o o co *^ o ooo o g g g ooo III i 12 8 2 .|| S t. * ^lo g g S S o e o o oo too o *O o ip o uS oscico'sj -J ^ C? Tf^t^cOC CO ^ O ^H ^H v*4 ^^ W^i " " fe i 1 1 i >t) X < ^ a^s-i - co eciak.5< I I " W 03 o P. A w" q s o >^x g o= I IS q d Iq t d* d d 99 lo lo Sp 1 co ^ eo|^n^r^ i i I O 3 woo TH* *" ll I 1 1 i 1 I q a n 100 og QJ o 9. fl .2 a o W W Q O ^ H H p a P o 55 ^j *^ *: o o s? CO CO 12 2 o o s I 000 I r-J r> O 3 S 8 3 & O 7j O'C fl> O W W OS5P W Q d d d d d d d d d* d" d d 101 Mil S25S3 ic2 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS idle 'for -long periods. 107 Chlorine reacts rapidly when passed into the 'melted wax at*about 125, or in solution in carbon tetrachloride. Such a chlorinated product, containing about 33 per cent chlorine, has been employed as a solvent for chloramine-T, about 10 per cent of this germicidal substance dissolving in the "chlorocosane" at ordinary tem- peratures. 108 The oxidation of paraffine by air or oxygen at 120-160 has already been referred to (see p. 52). Paraffine is also less stable to heat than is sometimes believed. Distillation, at ordinary pressure, of a wax melting at 52 causes a decrease in melting-point due to de- composition of about 4. In the old fashioned cracking process as car- ried out to increase the yield of kerosene, the wax often crystallizes from the distillate in fine large crystals, due largely to decrease in the viscosity of the distillate, but, according to Mabery, 109 paraffine is ac- tually decomposed during the process. Notes on the Refining of Petroleum Distillates. Petroleum distillates are refined with the object of removing offen- sive odors, removing or lightening the color and also rendering the oils more stable in the sense that certain constituents which oxidize readily with darkening of color and formation of acids or resinous substances are removed. The physical properties of the various fractions are but very slightly changed by refining, unless the lowering of the congealing point, or cold test, by the removal of paraffine wax may be considered as a refining operation. When first distilled from the crude oils the lighter fractions, including gasoline and kerosene, are nearly free from color and the lubricating oil fractions are clear shades of amber, brown or reddish brown, but on standing in contact with air, unrefined gaso- line and kerosene become yellow and the lubricating distillates become very dark in color. These color changes do not take place appreciably in well refined oils. Offensive odors are generally pronounced in the case of the more volatile oils, gasoline and kerosene, particularly when these are made from heavier oils by pyrolytic processes. The offensive odor of these distillates is commonly attributed to defines but, with the exception of conjugated di-olefines such as cyclohexadiene and cyclopentadiene present in light oil gas condensates, the odors of pure unsaturated hy- drocarbons are mild and not offensive. The conjugated di-olefines !M u f ] , 1 L lu rIcatin S oil also gives H 2 S on heating with sulfur. 'Dakin & Dunham. Chem. Abs. 12, 1079 (1918). 1M Proc. Am. Phil 800. 1897, 135. THE PARAFFINS HYDROCARBONS 103 have a sharp irritating odor suggestive of allyl alcohol or acrolein, but less pronounced. Unsaturated hydrocarbons generally develop objec- tionable odors on long standing due to oxidation, for example turpen- tine when fresh is very sweet and pleasant in odor, deteriorates by air oxidation, formic acid being one of the products formed. The con- stituents which are chiefly responsible for the objectionable odors of petroleum distillates are derivatives containing sulfur, nitrogen bases and naphthenic acids. These are very efficiently removed by the usual processes of refining with concentrated sulfuric acid and washing with caustic alkali, although special methods have to be resorted to in order to remove sulfur derivatives from oils derived from certain crudes, for example the Frasch copper oxide method as applied to petroleum of the Lima-Indiana field. Nitrogen bases in the more volatile distillates possess odors closely resembling pyridine. These simpler nitrogen bases are generally absent in the case of gasoline and kerosene distilled directly from crude petroleums, but are present in pyrolytic gasolines, unless made from nitrogen free oil. Petroleums of the Mid-continent, Gulf coast, California and Mexican fields on distilling under pressure yield volatile malodorous nitrogen bases. Mabery has investigated the nitrogen bases present in California petroleum and concludes that they are quinoline derivatives. When light distillates, e. g., motor fuel or kerosene, containing the simpler nitrogen bases, are treated with cop- per oxide, as by the Frasch method, the oxide combines with the or- ganic bases and treatment of the resulting copper oxide compound with caustic alkali liberates the nitrogen bases. In ordinary practice, how- ever, the organic bases are very efficiently removed by treating with concentrated sulfuric acid. When the acid sludge is diluted with water to precipitate oil and tarry matter, salts of the organic bases and a large proportion of the alkyl sulfuric esters, derived from the unsatu- rated hydrocarbons, remain in solution in the diluted acid. When this diluted acid is concentrated by the usual process of open pan heating and evaporation, this dissolved organic matter carbonizes and causes the destruction of a portion of the acid. The charring of this organic matter with the separation of carbon seriously interferes with the oper- ation of cascade evaporating systems by clogging of the overflow lips. The tarry matter precipitated by .diluting the sludge derived from treating lubricating oils, also generally contains nitrogen bases, as can readily be shown by heating or distilling with an excess of lime, but the quantity of ammonia thus obtainable is too small to be of indus- trial interest. 104 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Sulfuric acid is also very effective in removing naphthenic acids, as was first shown by Zaloziecki, 144 and Gurwitsch 145 later showed that this removal of naphthenic acids is not merely a solution effect and that far greater proportions of the naphthenic acids present pass into the acid layer than corresponds to the proportions required by the law of partition coefficients. These observations are in accord with the findings of Kendall and Carpenter who showed that a very wide variety of organic substances containing oxygen, e. g., aliphatic and aromatic acids, ketones, aldehydes, and phenols, form addition products with concentrated sulfuric acid and they regard these addition products as oxonium compounds. American petroleums do not contain conspicu- ous proportions of naphthenic acids, as do most of the Russian oils, but many of the Gulf Coast oils contain naphthenic acids of high boiling-point. These high boiling naphthenic acids are removed from the lubricating oil distillates by alkali. They are nearly odorless and their alkali soaps are very easily salted out of solution on account of their large molecular weight. They have apparently not been investi- gated and nothing definite regarding their empirical composition or chemical character is known. They are not recovered in present refinery practice. Practically nothing is known of the nature of the coloring matters in petroleum distillates. When such oils darken by air oxidation, amor- phous asphalt-like substances are formed. Sulfuric acid is very ef- fective in removing coloring matter, which is readily understood if the coloring matter consists largely of substances containing oxygen or nitrogen. It is improbable that these coloring matters are hydro- carbons, since the few colored hydrocarbons which are known contain conjugated unsaturated groups, as in the hydrocarbons of the fulvene series. Some writers regard the removal of such coloring matter by sulfuric acid as a purely physical or colloid phenomenon. 146 However, as all refiners know, it is necessary to use concentrated sulfuric acid in order to hold the tarry matter in solution, since in addition to the small amount of coloring matter present in the original unrefined lubri- cating oil, constituents are present which yield tar on treating with acid. Although water white gasoline and kerosene can be made with- out great difficulty, it is impossible entirely to remove the color from lubricating oils by sulfuric acid (or oleum) and alkali treatments. 144 Ghem. Ztg. 1892, 905. ] I PKvsM- Ghem. 87, 323 (1914). 148 Ubbelohde, Petroleum, 4, 1395 ; Schulz, Petroleum, 5, No. 4 and 8. THE PARAFFINS HYDROCARBONS 105 Pale yellow viscous oils can be made in this way which are practi- cally tasteless (liquid paraffine oil) but filtration through fuller's 147 earth, bone black or similar material, or distillation in vacuo, must be resorted to in order to obtain colorless oil such as is desired for phar- maceutical purposes. The fluorescence of petroleum distillates is due to substances which are largely removed by sulfuric acid, although several treatments with concentrated acid followed by treatments with oleum (15% S0 3 ) are necessary entirely to remove them. This property also has been re- garded by some writers as being due to particles of sulfur, carbon or other substance in a colloidal degree of dispersion, or due to the presence of substances having extremely large molecules. Although such mixtures are not optically homogenous and do show pronounced Tyndall effects, true fluorescence is not observed in aqueous or oil suspensions. Most petroleum distillates and certain crude petroleums which are sufficiently free from asphaltic matter, such as light Penn- sylvania crudes, exhibit green, bronze-green, bluish green or clear blue fluorescence. Examination of carefully filtered fluorescent oils in a quartz ultramicroscope of the Zsigmondy type shows no particles in suspension. 148 When sulfuric acid sludge is diluted with water and filtered to remove oil and tar the resulting aqueous solutions are usu- ally highly fluorescent. In other words, the fluorescent substances have been sulfonated to water soluble sulfonic acids. It is probable, therefore, that the extremely small quantities of fluorescent substances which are present in petroleum are highly condensed or benzenoid hy- drocarbons. 149 Such fluorescent substances are commonly formed when any organic substance is charred, for example, boiled linseed oil ex- hibits fluorescence if even slight carbonization occurs during the boiling process. The heavy waxy distillates obtained toward the end of the heating of an old-fashioned coking still are highly fluorescent. For various trade reasons it is sometimes desirable to modify the fluorescence and so-called "de-blooming" reagents are sometimes added to the oil. Thus nitronaphthalene is sometimes employed for this pur- pose. It is well known that the fluorescence of all organic substances which possess this property is greatly modified by various solvents 147 It is probable that the color absorbing qualities of fuller's earth are dependent upon the presence of partially dehydrated amorphous silica. Certain American refin- eries have recently manufactured a bleaching material superior to fuller's earth, by treating natural talc-like hydrated silicates with sulfuric acid, washing neutral f vatl , n * by dr y in S at not too high temperatures. 5 ks and B acon, J. Ind. & Eng. Chem. 6, 623 (1914). *u * ? hat the fluorescent constituents are not nitrogen derivatives is indicated by the fact that 80% sulfuric acid does not remove them. 106 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS and the fluorescence of petroleum is affected by the common solvents 15 in a way entirely parallel to the findings of Kauffman 151 in the case of the diaminoterephthalic acid methyl esters. Carbon bisulfide, nitro- benzene, and aniline diminish the intensity of the fluorescence and change its original bluish green character to dull green. Amyl alcohol and petroleum ether intensify the fluorescence and enhance its bluish character. Filtration of oils through fuller's earth does not remove the fluorescent constituents appreciably. Oxidizing agents destroy the con- stituents in question and sun-bleached oils which have thus been sub- jected to air oxidation are considerably altered in this respect, usually acquiring a brownish green or muddy fluorescence. The per cent of sulfur in the various petroleum fractions is very greatly reduced by treating with concentrated sulfuric acid, except in the case of highly unsaturated pyrolytic distillates when treated with a relatively small quantity of acid, in which case an increase in sulfur content may be observed. This is due to the formation of neu- RO tral esters of the type >S0 2 . No real explanation of the removal RO of sulfur compounds from mineral oils by sulfuric acid can be ad- vanced since our knowledge of the nature of these substances is so meager. Mabery and Smith 182 found that on treating a distillate from northern Ohio oil with sulfuric 1 acid the sulfur content was re- duced from 0.51% to 0.13%, and according to Robinson 153 sulfuric acid, 98% H.S0 4 , is much more effective than ordinary acid, a certain Ohio distillate containing 0.346% sulfur being thus refined to 0.05% sulfur. The reactions of sulfuric acid and pure olefines of different types have been discussed in another section. It is there shown that the hydration of the olefines to alcohols is important only with ole- fines of four to eight carbon atoms and that on standing in contact with the acid the proportion of polymers increases and the yield of alcohols decreases. With olefines of ten or more carbon atoms and containing one double bond, polymerization is the principal result; in certain instances being practically quantitative. In so far as the polymerizing action of sulfuric acid on unsaturated hydrocarbons is concerned, the specific gravity and viscosity of petroleum distillates should be increased by refining. Usually a slight decrease in these 180 Brooks & Bacon, loc. cit. 151 Ann. 393, 1 (1912). "'Am. Chem. J. 189^ 88. 1U Ohem. Ztg. Rep. 1907, 194. THE PARAFFINS HYDROCARBONS 107 values is observed after refining in this way, particularly in the case of lubricating oils. The effect of refining on the specific gravity of a number of pyrolytic gasolines, made by distillation of heavier oils under pressure of 100 to 150 pounds is indicated in the following: 154 25 Specific Gravity 25 Loss on refining, Original gasoline After refining % by volume 0.739 0.743 9.0 0.729 0.735 S3 0.727 0.748 9.8 0.737 0.754 10.1 0.730 0.749 10.6 Such oils refined and washed in the usual way become discolored on standing a few weeks, but if they are redistilled after refining, this discoloration does not take place, at least by no means rapidly. Such a redistillation also may serve the purpose of removing the heavy oily polymers formed by the acid treatment and which are commonly be- lieved to be objectionable constituents of gasoline when used as motor fuel or for extraction or cleaning purposes. One of the effects of treating highly unsaturated oils with relatively small proportions of sulfuric acid is to form alkyl sulfuric esters which remain dissolved in the oil and are not washed out by alkali. This is shown in the following treatment of a mixture of hexenes: SULFURIC ESTERS IN REFINED HEXENE. Vol. Residual g. SO, on % Cole, as Vol. Oil cc. VoLH 2 SOiCC. OiLcc. Distillation (RO)O> 50 25 32 0.284 4.9 50 50 28 0.146 2.9 50 100 26 0.094 1.8 The concentration of the sulfuric acid employed has a marked effect upon the sulfur thus introduced, as is shown by the following results of Condrea, 155 on a Roumanian kerosene, refined by a 2% volume of acid at 20: Acid concentration.. 90% 95% 97% 100% 5% SO, 10% SO, 20% SO. Color mm. to stand- ard 135 175 230 290 285 270 240 S0 2 .g. per 1 liter.... 0.157 0.294 0.426 0.67 1.30 1.71 257 Sulf onic acids in acid tar 1.30 2.57 4.20 7.30 12.45 16.77 35.00 Refining with sulfuric acid at low temperatures greatly reduces the oxidizing effect of the acid, with less attendant tar and color formation. 1M Brooks and Humphrey, loc. cit lu Rev. petrol. 1911, 61. 108 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS It is well known that lighter colored oils are produced by operating at low temperatures, but some difference of opinion exists as to the effi- ciency of the refining in other respects. Zaloziecki 156 gives the follow- ing data obtained by treating a Galician kerosene with sulfuric acid, 98.94% H 2 S0 4 , in the proportions of 50 grams per liter of oil. Acidity of Color in mm. Acid Tar Unused Sulfonic Acids Kerosene as to Match Temp.C Grams. HzSO* Calc.asHzSOt H 2 SO* Standard 61.6 47.91 1.45 0.86 193. 5 62.0 46.82 1.55 1.42 166. 10 62.5 46.53 1.65 1.56 143. 15 63.5 45.72 1.93 1.76 112. 20 64.3 44.37 2.22 2.45 89. 25 64.8 43.52 2.68 2.63 80. 30 65.2 41.87 3.72 3.65 52. 40 66.0 39.03 5.62 4.83 yellow 50 67.0 37.26 4.81 5.91 yellow Similar results have been noted by others. Mechanical agitation during the sulfuric acid treatment results in slightly lighter colored oils than when agitated by air. Generally, in practice, little attention is paid to temperature during the sulfuric acid treatment, particularly since cooling greatly increases the viscosity of oils of the lubricating class, and thus greatly prolongs the time required for the separation of the emulsified acid tar or sludge. Various mechanical means have been tried in the effort to increase the fineness of the emulsified oil particles, and also to decrease the time required for the tar laden acid to settle out. For the latter pur- pose centrifugal separation, and the addition of fine sand, infusorial earth and the like have been tried and while these methods give some- what better oils, these methods have not been adopted in large scale practice. Nitric acid or oxides of nitrogen in the sulfuric acid even in very small percentages, 157 e. g., .05 to 0.10 per cent, results in darker colored refined oils. Sulfuric acid made by the contact process is, therefore, much to be preferred to chamber acid, aside from the fact that the former acid is preferable on account of its greater concentration. The higher boiling distillates, for example, lubricating oils, require very much more acid for refining than kerosene or gasoline. The chemical reactions involved are fairly well known in the latter case but the chemical character of the substances removed from lubricating Ztff ' *v<* V 1 , 129 - For data on the rise in temperature on refining oils of (1908K 8ee Kissling ' Ch&m " Zt 9- 29 > 1086 < 1905 > ; Wischin, Petroleum 3, 1062 lw Schulz, Ohem. Rev. Fett u. Hwz. Ind. 20, 82 (1913). THE PARAFF1NE HYDROCARBONS 109 oils and why they react at all with sulfuric acid is not known. It is also possible that the large losses thus incurred are not necessary, that the per cent of substances present which are actually objection- able, malodorous substances, easily oxidized, color or acid forming substances, is really very small, as in the case of the lighter distillates. Naturally many other reagents have been tried, including benzensul- fonic acid, phosphoric acid, zinc chloride, aluminum chloride and the like. The latter, anhydrous aluminum chloride, is the only chemical refining agent other than sulfuric acid, which has shown great promise. The tar losses in this case are very high, but the quality of the prod- ucts produced, gasoline, lubricating oil or white medicinal oil, is re- markably fine. Anhydrous aluminum chloride polymerizes olefines energetically, decomposes sulfur derivatives and naphthenic acids. Color is very effectively removed. The oils so refined are extremely stable as regards oxidation by air. Interest in this reagent for refining has recently been revived by McAfee 158 and Grey. 159 In polymerizing amylenes by aluminum chloride Aschan obtained a series of saturated hydrocarbons and believed methylcyclobutane, cyclohexane and other cyclic hydro- carbons to be present in the lower boiling fractions. Liquid sulfur dioxide has been employed to some extent for refining kerosene, this method being based upon the marked difference in solu- bility of saturated and unsaturated and aromatic hydrocarbons in this solvent. With many oils the liquid sulfur dioxide method does not yield water white oils, and in such cases, refining with small proportions of sulfuric acid must be resorted to in order to get this result. The separation of the unsaturated and aromatic hydrocarbons from the paraffines is much more efficient at low temperatures, a temperature of - 12 being recommended. 160 While it is a fact that the removal of unsaturated and aromatic hydrocarbons improves the burning quali- ties of kerosene, and the Edeleanu process can, therefore, be considered as a rational method in this respect, there is nothing to indicate the refining value of the liquid sulfur dioxide method as regards naphthenic acids, malodorous sulfur compounds and the like. The method seems to be predicated mainly on the idea that unsaturated hydrocarbons should be removed from oils to be used as motor fuel, gasoline or naphtha solvents, lubricating oils, etc. On the other hand, there is con- 158 U. S. Pat. 1,277,328; 1,277,092; 1,277,329. Pat. T.322 *878*\ SZ21GJ2** '' 1>193 ' 541 ( essentiall y cracking processes). Cobb, U. 8. 16 'See' section on Physical Properties ; Solubility. 110 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS siderable evidence indicating that it is possible to refine motor fuel and lubricating oils to a satisfactory degree without the large losses at- tendant upon the removal of the unsaturated hydrocarbons and aro- matic hydrocarbons. That benzene can be satisfactorily used as a mo- tor fuel, particularly when mixed with gasoline, or gasoline and alcohol is now generally recognized. It is probable that the unsaturated hydrocarbons themselves, as removed from pyrolytic process motor fuel by the Edeleanu method, can be employed successfully in internal combustion engines, provided the resin forming conjugated diolefines, present only in very small proportions, be removed by fuller's earth according to Hall's refining process, or an equivalent method. 161 It is also probable that transformer oils and oils intended for the lubricat- ing of air compressors and internal combustion engines should be free from unsaturated hydrocarbons on account of the general tendency of such hydrocarbons to be readily oxidized by air. But it is possible that highly unsaturated but otherwise refined oils would prove satis- factory even in these instances. Considerable research needs to be carried out in order to determine precisely in what refining for par- ticular purposes should consist, and to develop industrially feasible methods of refining, which would remove the objectionable constituents with little or no loss of the valuable hydrocarbons. lei Th e writer has seen test runs of an automobile engine in which pure turpen- tine was used as the fuel, without abnormal deposition of carbon, with excellent thermal efficiency and without carburetor difficulties. A great many of our ideas as to what the characteristics of good motor fuel should be have apparently been derived from the commercial salesman, who had a certain article to sell. Chapter IV. The Ethylene Bond Theory of the Ethylene Bond and Cyclic Structures It is probably not exaggerating the relative importance of the mat- ter to state that the chemical behavior and physical properties of the unsaturated olefine, or ethylene group, is fully as important as the well differentiated properties of condensed or benzenoid structures. The chemical properties of the ethylene structure cannot properly be indicated by a few so-called type reactions and in the following dis- cussion it will be pointed out that all of the important chemical prop- erties of this group may be greatly influenced by structural configura- tion and proximity of other groups or substituents. The properties of this group as displayed in the enol form of tautomeric compounds is not discussed at length as this material has been well presented else- where and it is moreover not strictly germane to the subject of hydro- carbon chemistry. It will be of interest to examine the current theories regarding the atomic structure of such a linking. That the group >C = C< is rela- tively unstable, or under stress (Baeyer), is indicated by a wealth of experimental evidence. Our conceptions or theories of such carbon "linkings" have been greatly advanced by the general hypotheses re- cently published by Lewis and by Langmuir. First, Lewis l pointed out that "a study of the mathematical theory of the electron leads, I believe (irresistibly to the conclusion that Coulomb's law of inverse squares must fail at small distances." Like Parson, 2 Lewis believed that the most stable condition for the atomic shell is the one in which eight electrons are held at the corners of a cube. As regards the carbon atom Lewis may again be quoted. "Assuming now, at least in such very small atoms as that of carbon, that each pair of electrons has a tendency to be drawn together, perhaps by magnetic force if the mag- netic theory (of Parson) is correct, or perhaps by other forces which become appreciable at small distances, to occupy positions indicated by the dotted circles, we then have a model which is admirably suited to portray all of the characteristics of the carbon atom. With the 1 J. Am. Gliem. Soc. 38, 773 (1916). 'Smithsonian Inst. Public 65, 1915, p. 2371. Ill 112 CHEMISTRY OF THE NON-BEN ZEN01D HYDROCARBONS cubical structure it is not only impossible to represent the triple bond, but also to explain the phenomena of free mobility about a single bond which must always be assumed in stereochemistry. On the other hand, the group of eight electrons in which the pairs are symmetrically placed about the center gives identically the model of the tetrahedral car- bon atom which has been of such utility throughout the whole of or- ganic chemistry." Then two such tetrahedra, attached by one, two or three corners of each, represent respectively the single, double and triple bond. In the first case, one pair of electrons is held in common by the two atoms; in the second case two such pairs and in the third case, three such pairs. According to Lewis, the triple bond represents the highest possible degree of union between two atoms. Like a double bond it may break one bond producing two odd carbon atoms, but it may also break in a way in which the double bond cannot, i. e., to leave a single bond and two carbon atoms (bivalent), each of which has a pair of electrons which is not bound to any other atom. The three resulting structures, in the case of acetylene, may be represented as follows, H : C : : : C : H, H : C : : C : H and H : C : C : H. In addition we have a form cor- responding to Nef 's acetylidene and such forms as may exist in highly polar media, such as the acetylidene ion : C : : : C : H. The instability of multiple bonds, as well as the general phenome- non of ring formation in organic compounds, is admirably interpreted by the Strain Theory of Baeyer. This theory may, however, be put into a far more general form if we make the simple assumption that all atomic kernels repel one another, and that molecules are held to- gether only by the pairs of electrons which are held jointly by the component atoms. Thus two carbon atoms with a single bond strive to keep their kernels as far apart as possible, and this condition is met when the adjoining corners of the two tetrahedra lie in the line joining the centers of the tetrahedra. This is an essential element of Baeyer's Theory of stress in cyclic structures. When a single bond changes to a multiple bond and the two atomic shells have two pairs of electrons in common, the kernels are forced nearer together and the mutual re- pulsion of these kernels greatly weakens the constraints at the points of junction. This diminution in constraint, therefore, produces a re- markable effect in increasing the mobility of the electrons. In any part of a carbon chain where a number of consecutive atoms are dou- bly bound there is in that whole portion of the molecule an extraor- THE ETHYLENE BOND 113 dinary reactivity and freedom of rearrangement. This freedom usu- ally terminates at that point in the chain where an atom has only single bonds and in which, therefore, the electrons are held by more rigid constraints, although it must be observed that an increased mo- bility of electrons (and therefore increased polarity) in one part of the molecule always produces some increase in mobility in the neigh- boring parts. "There is much chemical evidence, especially in the field of stereo- chemistry, that the primary valence forces between atoms act in di- rections nearly fixed with respect to each other." 3 "Further evidence for the stationary electrons has been obtained by Hull, who finds that the intensities of the lines in the X-ray spectra of crystals are best accounted for on the theory that the electrons occupy definite positions in the crystal lattice." According to Langmuir's postulates carbon, atomic number six, has normally six electrons, two situated close to the nucleus or kernel as in helium, and the "four electrons in the second shell tend to arrange themselves at the corners of a tetrahedron for in this way they can get as far apart as possible." Langmuir regards the electrons in the atoms "as able to move from their normal positions under the influence of magnetic and electrostatic forces." It should be borne in mind when reviewing the chemical properties of the ethylene bond that there is no set of reactions which infallibly characterize this group as distinguished from other unsaturated types, particularly cyclopropane derivatives. This is in accord with Baeyer's strain theory and it is probably worth while to emphasize these rela- tionships and briefly review the theory. Baeyer was much impressed by the explosibility of the poly acety- lene compounds and endeavored to visualize the manner in which en- ergy could be absorbed in the formation of the acetylene bond, this energy being released as heat when such a substance explodes.. From the generalization of van't Hoff and LeBel, Baeyer inferred that "the four valences of the carbon atom act in directions which connect the center of the sphere with the corners of a (inscribed) tetrahedron, and which form an angle of 109 28' with each other. The direction of the attraction (or valence) can undergo a bending or distortion, which re- sults in a tension (Spannung) proportional to the amount of this bend- ing." 4 'Langmuir, J. Am. CTiem. Soc. &, 686 (1919). *Ber. 18, 2269 (1885). 114 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS "Ethylene is the simplest methylene ring, as it may be regarded as dmiethylene." In order to bend two of these hypothetical lines of valence direction to parallel positions would require that each of the pair be deviated one-half 109 28' or 54 44' from their normal direc- tions. In the same way the supposed deviations from the normal va- lence direction may be calculated for cyclopropane, cyclobutane, and so on. Ring structures containing more than five carbon atpms would require a spreading or widening of the normal angle, the angles of devi- ation of the simpler cyclic carbon structures being as follows: CH 2 CH 2 CH 2 CH 2 CH 2 CH 2 II I >CH 2 | | | >CH 2 OTJ OTI OTT O~H L>1 2 v^-tl-2 v^i2 ^-n_ _j_ 540 44' _|_ 24 44' + 9 44' + 44 H 2 H 2 C C H 2 C CH 2 H 2 C CH 2 H 2 C CH 2 H 2 C CH 2 V \ / vv vy ~ \j H 2 H 2 H 2 5 16' -9 33' cyclooctane, 12 46' cyclononane, 15 16' Cyclopropane and its derivatives are generally not as reactive as ethylene but the ring is broken by bromine, hydriodic acid, and by hydrogen in contact with nickel at 80. Cyclopropane is not oxidized by cold dilute permanganate. Cyclobutane is not reacted upon by bromine* concentrated hydroiodic acid or dilute permanganate solution. The ring is opened by hydrogen in the presence of nickel, forming butane at high temperature but is stable at 100. The stability of cyclopropane and cyclobutane rings toward oxidizing agents, bromine, halogen acids, dilute sulfuric acid and the like is very greatly modified by substituent groups, just as the chemical behavior of the ethylenes is altered by different groups. Thus 1 . 2-dimethylcyclopropane is acted upon by 1% permanganate 5 and the hydrocarbon 1, 1, 2-trimethyl Zelinsky, J. prakt. Chem. 84, II, 543 (1911). THE ETHYLENE BOND 115 cyclopropane combines with concentrated hydrochloric acid at 100. The derivatives CMe 2 CMe 2 CH 2 < I and CH 2 < | CH . CH 2 CHMe 2 CH . C0 2 H are stable to permanganate solution but the former is hydrogenated in contact with nickel at 125 and adds hydrobromic acid very slow- ly. 6 Ethylcyclobutane is extremely stable, being unaffected by per- manganate solution, concentrated hydrobromic acid at 100, concen- trated sulfuric acid at 25 and is only reduced by HI at 210. The cyclobutane derivative 1, 1, 3, 3-tetramethyl 2, 4-diethylcyclo- butane Me 2 C CHC 2 H 5 C 2 H 5 CH CMe 2 is also remarkable for its stability, its chemical behavior resembling that of a saturated hydrocarbon of great inertness. 7 The acid chloride CH 2 >CHCOC1 is sufficiently stable to anhydrous aluminum chloride CH 2 and hydrogen chloride 8 to react normally in the Friedel and Crafts CH 2 synthesis to give good yields of the ketone C 6 H 5 OC.CH< | . This CH 2 fact is somewhat remarkable in view of the ease with which the cyclo- propane ring in carane and sabinene arid the cyclobutane ring in the pinenes is ruptured by halogen acids, by bromine and by di- lute mineral acids. Wallach 9 has noted that the ketonic acid CH.COCH, CH.C0 2 H CH 2 < I is very unstable, but the acid CH 2 < | C CH 2 C0 2 H * C CH 2 C0 2 H C 3 H 7 C 3 H 7 is very stable. Although Baeyer's theory needs revision in the light of our present knowledge and theories of valence and atomic structure, it has passed Kishner, J. Chem. Soc. Aba. 1913, I, 1163. 'Wedekind & Miller, Ber. kk, 3285 (1911). Kishner, J. Russ. Phys.-Cliem. Soc. tf, 1163 (1911). Ann. 360, 82 (1908) ; 388, 49 (1912). 116 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS the test of usefulness and been of very great value. Experience is gen- erally in accord with the theory and the yields in analogous reactions of synthesis indicate that in the cyclopropane, cyclobutane, cyclopen- tane and cyclohexane series, derivatives of cyclopropane are produced with the greatest difficulty or poorest yields, and that while cyclobu- tane and cyclohexane derivatives are much more easily obtained, the tendency to form cyclopentane derivatives is so pronounced that quan- titative yields are frequently produced and, in fact, cyclopentane de- rivatives sometimes result during reactions which might be expected to yield other ring structures. As regards the relative influence of different substituent groups in such syntheses Perkin 10 states that it is clear that a useful generalization cannot be formulated until a much larger number of cyclic carboxylic acids and other derivatives have been prepared and investigated. J. von Braun xl states that 1,4 di- halogen alkyls and sodium malonic ester give good yields of cyclo- pentane derivatives but the same reaction applied to the synthesis of cyclohexane and cycloheptane compounds, from the 1, 5 and 1, 6 di- halogen derivatives, respectively, give very poor yields. The ease with which cyclohexanones are converted to cyclopentanones has been noted by Wallach 12 and the four carbon ring in cyclobutyldiethyl- carbinol, on decomposition with loss of water, forms the five carbon ring 1, 2-diethylcyclopentene. 13 However, a very large number of re- arrangements have been observed in which 'change to a system, less stable so far as the Baeyer theory and the number of carbon atoms in the ring is concerned, is brought about. 14 J. F. Thorpe 15 and his assistants reasoned that if two valences of a given carbon atom are under strain due to ring formation, the di- rections of the two remaining valences would be affected, for example, the angle formed by two side chains attached to a carbon atom in a ring such as cyclohexane, would be bent from the normal 109 28' re- quired by Baeyer's theory. In the case of cyclohexane these two side chains may be closer together than in a corresponding compound hav- ing an open chain structure. Their results are an interesting confirma- tion of the theory. Thorpe has compared the relative stability of the cyclopropane derivatives formed by the elimination of hydrogen 10 Cf. Goldsworthy & Perkin, J. Chem. Soc. 105, 2665 (1914). 12 J. Chem. Soc. Abs. 1916, I, 487. 13 Kishner, Chem. Zentr. 1912, I, 1001. 14 See chapter on Rearrangements. "Beesley, Ingold & Thorpe, J. Chem. Soc. 107, 1080 (1915) THE ETHYLENE BOND 117 bromide from the monobromo derivatives of cyclohexane-1 . 1-diacetic acid and P|3-dimethylglutaric acid, as follows, CH 2 CH 2 CHBr.C0 2 H CH 2 < >C< _* CH 2 CH 2 CH 2 C0 2 H. CH 2 CH 2 CH.C0 2 H CH 2 < >C< | CH 2 CH 2 CH.C0 2 H CH 3 CHBr.C0 2 H. CH 3 CH.C0 2 H CH 3 CH 2 C0 2 H. CH 3 CH.C0 2 H. Both of the resulting acids are remarkably stable towards boiling acid permanganate solution but the chief difference observed was in their behavior to concentrated hydrochloric acid in sealed tubes at 240 under which conditions the spiro acid, from cyclohexanediacetic acid, is unaffected but the other, trans-caronic acid, is completely changed to terebic acid, with rupture of the ring. The thermal measurements of Stohmann and Kleber are not in good agreement with Baeyer's theory. According to their work, the quantities of heat absorbed in the formation of similarly constituted compounds containing the cyclopropane, cyclobutane, cyclopentane and cyclohexane rings by the removal of two atoms of hydrogen from the corresponding open-chain substances, are as follows: Ring ............................. C 8 C 4 C 6 C, Angle of strain (Baeyer) .......... 24.7 9.7 0.7 5.3 Heat absorbed, calories ........... 38.1 42.6 16.1 ' 14.3 Ingold 16 has suggested that these calculated angles of strain may not be correct and that the normal tetrahedral angle of Baeyer (2 tan -1 V 2 = 109.5) may be modified somewhat according to the volume occupied by the four attached atoms or groups. (The distor- tion of the valency direction, as suggested by Ingold, has nothing in common with the theories of Guye and Brown, which refer to the ef- fect of the size of the substituent radicles upon the asymmetry of the molecule as measured by the molecular optical activity.) Ingold suggests that "the tetrahedron representing a carbon atom is approximately regular only when the carbon atom is attached to four atoms of a similar kind," for example, to four carbon atoms, M J. Chem. Soc. 119, 306 (1921). 118 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS (A) . However in cyclopropane, cyclobutane and the like, each carbon atom is attached to two hydrogen atoms and two carbon atoms, (B). C C H C v v / \ H/ \ (A) (B) Since the hydrogen atom occupies a much smaller volume than the car- bon atom, it is accordingly possible that in the >CH 2 group the two carbon atoms attached to the central one occupy more of the sur- rounding space than do the hydrogen atoms. If this is so, the angle be- tween the carbon-to-carbon valencies of a polymethylene chain will not be 109.5, as hitherto supposed, but will be some angle greater than this." Using Traubes values for the atomic volumes of carbon and hydrogen, Ingold calculates that this volume factor causes a change in the angles between each pair of carbon-to-carbon valencies in a polymethylene chain and that this angle may be nearly 6 greater than has hitherto been supposed. Employing this new angle 115.3 instead of 109.5, Ingold calculates "by how much" the terminal car- bon atoms of C 3 , C 4 , C 5 , and C 6 rings must approach one another and obtains values more nearly in accord with the thermal results of Stoh- mann and Kleber. The above stereo-chemical considerations afford an explanation of the effect of the gem-dimethyl group in promoting certain reactions and in other cases greatly increasing the stability of the substance. Thus, cux-dimethylbutane -apy-tricarboxylic acid is smoothly con- verted into the cyclopentanone derivative, on heating its sodium salt with acetic anhydride, but this change has not been observed with adipic acids which do not contain a gem-dimethyl group. The (CH 3 ) 2 C< group stabilizes certain lactones, for example, |3p-dimeth- ylglutaric anhydride may be boiled in water for hours without change, and a|3|3-trimethylglutaric anhydride may be crystallized from hot water in crystals containing water of crystallization, but ordinary glu- taric anhydride is easily decomposed by water. Hiickel 17 regards the heat of combustion of CH 2 as different in each polymethylene ring and points out that if the heats of combustion of these hydrocarbons are divided by the number of CH 2 groups ^er. 53 B, 1277 (1920). THE ETHYLENE BOND 119 tained in the hydrocarbon concerned, then values are obtained which are much better in accord with Baeyer's theory than the older com- parisons of Stohmann. In this way, the values for CH 2 in ethylene, cyclopropane, cyclobutane, cyclopentane and cyclohexane are calcu- lated to be 170, 168.5, 165.5, 159, 158 calories, respectively. In view of the fact that the chemical behavior of the cyclopropane group reveals a condition of unsaturation or strain (Baeyer) it is not surprising that ring closing in this case influences the physical proper- ties of substances containing this ring complex. This will be discussed more fully in the section dealing with physical properties and constitu- tion but it may be noted here that one of the most significant and use- ful properties, refractivity, is affected by the formation of the 3 carbon ring to almost the same degree as in the case of the ethylene bond, and that when the cyclopropane group occurs in a conjugated position to an ethylene bond substantially the same degree of exaltation is ob- served as is noticed in the case of two conjugated ethylene bonds. Chemical Properties of Unsaturated Substances of the Ethylene Type. Unsaturated substances of the ethylene type, e. g., substances con- taining one or more so-called olefine groups, are capable of a series of reactions which are very widely applicable to nearly all substances containing such an Unsaturated group and which have come to be re- garded as characteristic reactions of this type of unsaturation. These reactions are best exemplified by the addition of ozone, of halogens, particularly bromine, oxidation by potassium permanganate solution to the corresponding glycols, addition of nitrosyl chloride and oxides of nitrogen. Other reactions less widely applicable will be noted be- low. All of these reactions involve rupture of one of the ethylenic linkings or, in other words, one of the primary valences. In addition to these reactions, it has been noted that substituted ethylenes are capable of forming a large number of so-called molecular compounds with other substances. In these compounds the double bond is not broken and the formation of these molecular compounds is due to what is termed, for lack of a better name, "residual valence," "latent affin- ity," "secondary valence," and similar terms.' It should be pointed out, however, that the ability to form such molecular compounds is by no means limited to Unsaturated substances of the ethylene type. 120 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Well crystallized compounds of p-tetrabromotetraphenylethylene with acetone, ether, methylethyl ketone, carbon tetrachloride, ethyl acetate and benzene, have been described. These compounds are easily decomposed to the original constituent substances. Norris 18 has sub- mitted the following hypothesis in regard to substances of this kind. (1). The molecular compound is formed as a result of the com- ing into play of latent affinities residing in an atom in each of the constituents of the compound. (2). All atoms possess these latent affinities. If an atom in a compound reacts with difficulty when the latter is brought into con- tact with other substances, it is evident that a large part of its energy has been expended and but a little of it remains to take part in re- action. On the other hand, if the atom enters into reaction readily with other substances, it is evident that it still possesses available energy. It is probable, therefore, that such active atoms might be able to unite with atoms of a similar nature (with respect to residual energy) and form molecular compounds. A study of the literature confirms the view that compounds containing unusually active elements or groups form well characterized molecular compounds. (3). Substances which contain inactive double bonds may form molecular compounds. In most cases direct addition of atoms or groups at the double bond leads to the formation of ordinary saturated compounds. So-called unsaturated compounds are known, however, in which the unsaturation is so slight that they will not unite with such an active element as bromine. The chemical affinity latent in the dou- ble bond is so small that it cannot hold in combination other atoms or groups linked to it by primary valence bonds. Many such com- pounds form well characterized molecular compounds. In other words, the available energy of the double bond is not enough to neutralize the energy of atoms and form a true valence bond, but is suf- ficient to interact with a similar small amount of energy residing in another compound. For example, p-tetrabromotetraphenylethylene (BrC 6 HJ 2 C = C(C 6 H 4 Br) 2 will not react with bromine, as shown by Bauer, 19 but it does form a series of molecular compounds, as noted above. [There is apparently no way of determining whether the resid- ual energy which makes these combinations possible in this case is really inherent in the double bond or in the bromine atom. The latter possibility suggests itself in view of the fact that tetraphenylethylene 18 J. Am. Ohem. Soc. ft, 2086 (1920). "Ber. 37, 3317 (1904) ; Hinrichsen, Ann. 336, 223 (1904). THE ETHYLENE BOND 121 dichloride (C 6 H 5 ) 2 CC1.CC1(C 6 H 5 ) 2 also forms molecular compounds. B. T. B.] The hypothesis that unsaturated hydrocarbons are able to unite with acids, by virtue of "free partial valences," to form salt-like sub- stances was put forward and subsequently rejected by Baeyer but Kehrmann and Effront 20 have revived the hypothesis to account for the formation of two series of salts by the triphenyl methane dyes, and for the behavior of certain unsaturated ketones towards acids. For ex- ample, distyryl ketone gives mono- and di-acid compounds, lemon- yellow and orange-red respectively, from which it seems necessary to assume that combination can occur at one double bond in addition to the oxygen atom. The reaction of bromine with styrene and substituted styrenes shows that replacement of one of the methylenic hydrogen atoms by an aryl group increases the reactivity to bromine slightly and this difference is further accentuated by substituting both hydrogen atoms by alkyl groups. The introduction of one halogen atom decreases the reactivity toward bromine but the effect of a CN group is even more marked, the effect of the CN and carboxyl group being of about the same order. 21 Double compounds of ethylene and aluminum chloride have been isolated but energetic polymerization occurs with most olefines. Very little work has been done with the Friedel and Craft reaction as ap- plied to non-benzenoid hydrocarbons. Darzens 22 found that acetyl chloride did not react with cyclohexane in the presence of aluminum chloride but with cyclohexene the saturated chloro ketone was formed. H Cl X CH 3 COC1A1C1 3 / \_oo.ctt v Stannic chloride is a most efficient catalyst for this reaction. Norris and Couch 23 have recently noted that ethylene reacts with benzoyl 2 Bcr. 5k, 417 (1921). "Reich et al. Helv. Chim. Acta. 4, 242 (1921). 22 Compt. rend. 150, 707 (1910). The chloride noted above may be decomposed by alkali to give the unsaturated ketone, HC=C< COCH3 n J. Am. Chem. Soc. 42, 2330 (1920). 122 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS chloride in the presence of A1C1 3 apparently in a different manner, to give phenyl vinyl ketone C 6 H 5 CO . CH = CH 2 . The chloride, C 6 H 5 CO.CH 2 CH 2 C1, corresponding to Darzen's product, was not ob- served. There is no subject in organic chemistry to which it is more diffi- cult to give accurate expression than the modification of the chemical behavior of certain groups or siibstituent atoms by other groups or atoms in the same molecule. As pointed out by Bauer the substitu- tion in ethylene of strongly negative groups diminishes the ability of the substance to react with bromine. 24 ,0n the other hand, the substi- tution of halogens or the phenyl group very markedly increases the reactivity of the ethylene group in certain other respects. Thus ethyl- ene is polymerized only at high temperatures and pressures and in the presence of catalysts such as alumina or iron, or in the presence of very reactive substances such as anhydrous aluminum chloride or zinc chloride. On the other hand, styrene C 6 H 5 CH = CH 2 , vinyl bromide, CH 2 = CHBr, and vinyl chloride polymerize on standing at ordinary temperatures, and rapidly under the influence of light. The enhanced reactivity of the hydrogen atoms in styrene is also indicated by the fact that this substance yields the nitro derivative C 6 H 5 CH = CHN0 2 when treated with nitric acid. 25 That the double bond greatly influences the reactivity of the sub- stituent halogen atoms is also well known. Thus vinyl bromide and vinyl chloride are remarkably stable to alkalies and in many of their reactions closely resemble chloro-benzene and bromobenzene. Ad- vantage is taken of this unusual stability of chlorine substituted ethyl- enes, with respect to reactivity of the chlorine, in utilizing them as commercial solvents. For example, trichloroethylene, CHC1 = CC1 2 , is not appreciably hydrolyzed by hot water and is practically not af- fected by iron or copper and is therefore admirably adapted for use as a solvent in industrial apparatus made of these metals. 26 This sta- bility of halogen derivatives of ethylene is also indicated by the com- mercial methods of manufacturing trichloroethylene, e. g., treating tetrachloroethane with alkali or passing over thorium oxide at 390. 27 On treating trichlorocyclohexane with alcoholic caustic potash the prin- 24 Perkin has called attention to the fact that the stability of cyclopropane and cyclobutane derivatives is variable within wide limits depending upon the character of the substituent groups. 28 Recent work of Wieland, Ber. M, 201 (1920), shows that ethylene reacts with a mixture of nitric and sulfuric acids (20% oleum) to give a mixture of ethylene dinitrate and B-nitro-ethyl nitrate. 26 Gowing-Scopes, J. tSoc. Chem. Ind. S3, 160 (1914) ; Crudes, Chem. Aba. 1917, 544. 27 German Pat. 171,900; 206,854 (1906) ; 274,782 (1914). THE ETHYLENR BOND 123 cipal product is chlorodihydrobenzene, C 6 H 7 C1, the last chlorine atom being stabilized by the adjacent double bonds. Wohl 28 has shown that when tetramethylethylene (CH 3 ) 2 C = C(CH 3 ) 2 is treated with n-bromoacetamide, primary addition occurs through subsidiary valences of the bromoacetamide and of one or both of the unsaturated carbon atoms of the olefine. Acetamide is then formed, the bromine atom taking the place of the hydrogen re- moved to form acetamide, the final products being acetamide and (CH 3 ) 2 C = C . CH 3 . CH 2 Br. Free bromine reacts energetically with the unsaturated hydro- carbons and therefore solvents are usually employed in such re- actions, e. g., carbon bisulfide, carbon tetrachloride, glacial acetic acid and, less generally, alcohol or ether. The reaction is rapid and standardized solutions of bromine in acetic acid or carbon tetrachloride can often be used to titrate such hydrocarbons and determine the de- gree of unsaturation. 29 However, substitution of hydrogen sometimes takes place and the well-known analytical methods of Hiibl, Hanus and Wijs, which are of such value with unsaturated fatty oils, cannot be relied upon to give correct results in the case of the terpenes and the higher ethylene homologues derived from petroleum. 30 Bromine addition products are sometimes crystalline solids and thus serve for purposes of identification, as in the case of butadiene, the tetrabro- mide 31 melting at 118, and limonene and dipentene whose tetrabro- mides melt at 104-105 and 124 respectively. The addition of halo- gen acids has already been referred to in the section dealing with the preparation of halogen derivatives. The addition of bromine is made use of in the analytical chemistry of rubber and a chlorinated rub- ber 32 has recently appeared on the market. Hypochlorous acid reacts with ethylene bonds more readily than concentrated sulfuric acid, forming chlorohydrins. Thus ethylene re- acts readily with cold dilute solutions of hypochlorous acid, and also other substances, which are inert or react only very slowly with sulfuric acid at ordinary temperatures, yield chlorohydrins, for ex- ample, cinnamic acid, allyl bromide, maleic acid and the higher ethyl- ene homologues. Solutions of chlorine water give nearly theoretical **Ber. 52, B. 51 (1919). Cf. v. Soden and Zeitschel, Ber. S6, 266 (1903). 10 For description and details for carrying out these determinations see Leach, "Food Analysis," pp. 488-530, 4th Ed. Lewkowitsch, "Oils, Fats and Waxes," Vol. I, p. 393. See Faragher and Garner, J. Ind. & Eng. Chem. 13 f 1044 (1921). 31 A low melting modification melting at 37.5 is also known. " The product carries the trade name "Duroprene" and appears to have value as a varnish film resistant to corrosive vapors or acids. 124 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS yields of ethylene chlorohydrin but other unsaturated hydrocarbons that react energetically with chlorine also yield dichlorides. In such cases better yields of chlorohydrins are obtained by employing dilute solutions of alkali hypochlorite which yield free hypochlorous acid by hydrolysis and contain no free chlorine. Thus Walker 33 employs so- dium hypochlorite in the presence of sodium bicarbonate to prepare amylene chlorohydrins (carbonic acid is a stronger acid from the ioni- zation standpoint than hypochlorous acid) . The chlorohydrins of ethyl- ene, propylene, butylene, 34 amylenes, 35 and hexylenes are best known. Propylene and hypochlorous acid yields a mixture of the two isomers CH 3 CHOH.CH 2 C1 and CH 3 CHC1.CH 2 OH. By the action of hydro- gen chloride on propylene oxide both isomeric chlorohydrins are ob- tained as has been shown by an examination of their rate of hydroly- sis. 36 Isobutylene and the amylenes also yield a mixture of isomeric chlorohydrins. 37 All of these simpler chlorohydrins yield alkylene oxides when treated with concentrated caustic alkali, and slow hydroly- sis in the presence of sodium bicarbonate gives good yields of the gly- cols. The utilization of the ethylene and propylene in oil gas and petroleum still gases in this manner has recently been attempted on an industrial scale. On heating the simpler chlorohydrins with water, aldehydes, or ke- tones are formed. Thus 2-chloro-3-hydroxybutane is completely con- verted to methyl ethyl ketone in 3 hours at 120. Propylene chloro- hydrins give acetone and propionic aldehyde and the chlorohydrin of trimethyl ethylene similarly yields methyl isopropyl ketone. 38 The reaction of hypochlorous acid with other unsaturated sub- stances, for example, the terpenes, unsaturated petroleum oils and fatty oils has been very little studied. Pinene yields a mixture of products, 39 among which is pinol oxide, C 10 H 16 2 , which oxide, unlike cineol, is very easily hydrolyzed by dilute acids to a glycol. The di- chlorohydrine C 10 H 18 2 C1 2 is also formed, the bridged ring being opened. The substance cis-pinolglycol-2-chlorohydrin, C 10 H 17 2 C1, is very stable to aqueous alkalies as is also the chlorohydrin obtained from camphene, 40 C 10 H 16 HOC1. Large proportions of chlorination 88 U. S. Pat. 972,952 ; 972,954. 84 Henry, Bull. Acad. roy. Belg. 1906. 523 ; Compt. r&nd. 142. 493 ; Krassuski, Chem. Zentr. 1901, I, 995. "Carius, Ann. 126, 199 (1863) ; Umnowa, Chem. Zentr. 1911, I, 1278. "Smith, Z. physik, Chem, 93, 59 (1918) ; Cf. Michael, Ber. 39, 2785 (1906). 87 Henry, loc. cit. "Krassuski, Chem. Zentr. 1902, II, 20. 88 Wagner & Slawinski, Ber. S2, 2064 ; Henderson & Marsh. J. Chem. Soc. 119, 1492 (1921). 40 Slawinski, Chem. Zentr. 1906, I, 137. THE ETHYLENE BOND 125 products are also formed in the case of camphene and this together with the fact that these chlorohydrins are relatively stable and are not easily converted to glycols perhaps accounts for the fact that hypochlorous acid has not become an instrument of research in this series. The influence of constitution and the presence of substituent atoms or groups on the addition of water and behavior toward acids, or their aqueous solutions, is very pronounced. A few substances possess- ing double bonds, carbon to carbon, react with water energetically, for example, ketene H 2 C = CO and carton suboxide, OC = C = CO, whose behavior toward water resembles that of acid anhydrides. The unsaturated hydrocarbons themselves, however, do not react with wa- ter directly although Engelder observed indications that the dehydra- tion of alcohol to ethylene and water in the presence of alumina or kaolin, is reversible. 41 Aqueous solutions of organic acids, particularly formic and oxalic acids, effect hydration in certain instances, for example (CH 3 ) 2 C = CH . CH 3 + H 2 * (CH 8 ) 2 . C . OH . CH 2 CH 3 but the method is by no means general and is of no preparative value. The formation of esters of organic acids and defines on heating or in the presence of other substances, such as zinc chloride or sulfuric acid, often gives excellent yields. Heptylene and acetic acid heated in an autoclave or sealed tube to 300 yields heptyl acetate. 42 Amylene and acetic acid react at ordinary temperatures in the presence of zinc chlo- ride, but the yield is greatly diminished by the formation of polymers. (CH 8 ) 2 C = CH.CH 3 + CH 3 C0 2 H (CH 3 ) 2 C(0 2 C.CH 3 ).CH 2 CH 3 polymers In most cases, better results are obtained by the method of Ber- tram and Walbaum, in which process the olefine is dissolved in an excess of acetic acid and a relatively very small quantity of sulfuric acid is added. 43 The presence of water greatly retards the acetylation. This reaction does not appear to have been applied industrially to the acetylation of amylenes or other olefines derived from petroleum, but 41 J. Phys. Chem. 21, 676 (1917). Behal and Desgrez, Oompt. rend. 114, 676 (1892), J. prakt. Chem. W, 7 (1894). 126 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS has been conspicuously successful in the acetylation of camphene to bornyl acetate (see Artificial Camphor) . Barbier and Grignard ** rec- ommend benzene sulphonic acid instead of sulfuric acid to promote the reaction and state that the addition of acetic anhydride to the reaction mixture increases the yield of ester. Pinene yields mainly a-terpineol acetate. Anhydrous oxalic acid and pinene at 120 yield bornyl oxalate and formate, which process formed the basis of the first artificial camphor process to be attempted on an industrial scale. 45 A large number of patents have been issued covering the use of other organic acids in making borneol esters. Just as hydrogen chloride containing a little moisture yields chiefly dipentene dihydrochloride, concentrated formic acid (98-99%) yields mainly terpinyl formate, the bridged ring being opened in each case. The results of treating unsaturated hydrocarbons with sulfuric acid is of considerable interest in connection with the hydration of olefmes, including the terpenes, and also the refining of petroleum dis- tillates. The results include changes of the following nature: re- arrangement, or shift in the position of the double bond, polymeriza- tion, formation of mono and dialkyl sulfuric esters, and hydration to alcohols. The tendency of the defines and substituted ethylenes to react with sulfuric acid is distinctly less than their tendency to react with bromine. Thus cinnamic and fumaric acids readily yield dibromides but are not affected by ordinary concentrated sulfuric acid at 25. The substitution for the hydrogen of ethylene, of groups which impart a strongly electronegative character, results in decreased reactivity to sulfuric acid. Thus cinnamic and fumaric acids are inert, and dichlo- roethylene and trichloroethylene are only very slowly acted upon by sulfuric acid at ordinary temperatures. Allyl bromide is also more stable to sulfuric acid than is propylene. The substitution of groups which impart an electropositive character, such as methyl groups, re- sults in greatly increased reactivity to sulfuric acid. Isobutene, (CH 3 ) 2 C = CH 2 is rapidly and completely dissolved by sulfuric acid, 63% H 2 S0 4 , at 17. Also tetramethylethylene (CH 3 ) 2 C = C(CH 3 ) 2 reacts readily and completely with 77% acid at ordinary tempera- tures. Of the two amylenes "Compt. rend. US, 1425 (1907) ; Bull. Soc. Chim. (4), 5, 512 (1909). U THE ETHYLENE BOND 127 CH 3 C 2 H 5 > CHCH = CH and > C = CH CH 3 CH 3 the latter dissolves .more readily in 66% acid. 46 Results very closely parallel to these have been noted in the case of the reactions of amyl- enes and halogen acids. 47 Michael and Brunei believed that in the ali- phatic hydrocarbon series the tendency to form alcohols and alkyl sul- furic esters decreases with increasing molecular weight, this result ap- pearing to be maximum with the amylenes and hexylenes. With in- creasing molecular weight polymerization becomes the principal result, which result, however, may possibly be preceded by alcohol forma-' tion. 48 The difference in the final results may, therefore, be due in large part to the relatively greater stability of the simpler alcohols. Thus un- C 2 H 5 der the same conditions 3-ethylpentene (2) >C = C< yields C 2 H 5 H 72% alcohol and 12% polymers and 2-methylundecene(2) yields 97% polymers ancj only a trace of alcohol. Secondary octyl alcohol, octane-ol(2), treated with 95% sulfuric acid at 20 gives a yield of octene polymers C 16 H 32 and C 24 H 48 , increasing with the time of stand- ing. A mixture of octene (1) and octene (2) treated with sulfuric acid, with cooling, yields chiefly a mixture of the di- and tri-polymers. 49 In a study of a series of pure unsaturated hydrocarbons Brooks and Humphrey noted that the polymers were always more stable to sulfuric acid than the parent olefines. 50 Kondakow noted a closely parallel behavior in the reaction of hydrogen chloride and isobutene and its polymers. 51 These results can be sxpressed in another way, e. g., unsaturated hydrocarbons are more highly polymerized, to higher boiling, more viscous polymers, by 100% sulfuric acid than by 95% acid and the latter will produce a higher degree of polymerization than 85% acid. The mechanism of these changes is very obscure. It has generally been assumed that the alcohols, formed by treating unsaturated hydro- carbons with sulfuric acid or dilute sulfuric acid, were a result of the hydrolysis of the alkyl sulfuric esters first formed, 49 Michael and Brunei, Am. Chem. J, 41, 118 (1909).* "Eltekow, Ber. 10, 707 (1877) ; Konowalow, Per. IS, 2395 (1880). 48 Cf. Brooks and Humphrey. J. Am. Chem. Soc. 10, 822 (1918). 49 Rossolimo, Ber. 27, 626 (1894). 60 J. Am. Chem. Soc. 1,0, 822 (1918). 61 J. prakt. Chem. (2) 54, 449 (1896). 128 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS RCH RCH. RCH 2 | + H 2 > | RCHOSO.H RCHOH JLVV~/XJL* J.VV-/J.J-2 || +H 2 S0 4 I +H 2 > | +H 2 S0 4 RCH EOF Thus ethyl-hydrogen sulfate can be hydrolyzed to give ethyl alco- hol but the relative stability of this ester is indicated by the fact that a dilute solution of ethyl-sodium sulfate is hydrolyzed in 8 days at 60 only to the extent of 16 per cent. 52 The mono or acid sulfuric esters of amylenes, hexenes and heptenes are not appreciably hydrolyzed on diluting with water at ordinary temperatures and their hydrolysis in dilute solution at 100 is very slow. However, when these olefines are dissolved in cold sulfuric acid and the clear homogenous acid solu- tion diluted with water at the free alcohols are precipitated imme- diately in yields sometimes as high as 70 per cent of the theory. Fur- ther dilution or complete extraction of the alcohols remaining dissolved by means of an immiscible solvent causes no hydrolysis of the alkyl sulfuric esters which remain in the aqueous solution. The barium salts of these acid esters can be easily isolated by slow evaporation without appreciable decomposition. Although these alkyl sulfuric esters can be saponified by caustic alkali or hydrolyzed by prolonged boiling or steaming, they are not hydrolyzed to alcohols under the conditions which obtain in the separation of the alcohols from these sulfuric acid mixtures. Also the highest yields of alcohol are obtained when employing sulfuric acid containing water, greater yields of alcohol being obtained with 85 per cent acid than with 95 per cent or 100 per cent acid, or with benzene sulfuric acid. To account for these facts the theory has been proposed 53 that the addition of water to olefines with formation of free alcohols, in cold solutions, is due to reaction with the monohydrate of sulfuric acid H 2 S0 4 .H 2 0, or higher hydrates. The monohydrate, or orthosulfuric acid, is usually regarded as having the constitution HO OH Vo / \ HO OH B2 Linhart, Am. J. Sci. 35, 283 (1913) ; Evans and Albertson mention that in the system C 2 H B OH+H 2 SO45C 2 H B H.SO4 + H 2 O the dilution of the mixture by titration does not cause appreciable hydrolysis. [J. Am. Chcm. Soc. 39, 456 (1917).] 63 Brooks and Humphrey, loc. cit. THE ETHYLENE BOND 129 It is practically certain that esters of this acid would have quite differ- ent degrees of stability and quite different rates of hydrolysis than the known relatively stable esters of ordinary sulfuric acid. The hydration of pinene to terpin hydrate C 10 H 18 (OH) 2 .H 2 by dilute aqueous acids has long been known. Heating terpin hydrate with dilute sulfuric or phosphoric acids results in partial decomposition to terpineol, which process is carried out industrially. Wallach 54 has pointed out the marked effect of differences of constitution on the rate of hydration of five menthenols. CH 3 GIL c-OH CH 3 CH 2 /\ The menthenols I and II react readily with 5% sulfuric acid at ordinary temperature and III a little less rapidly. Menthenols IV and V react so much slower than I, II and III, that separation of these two groups can be effected in this way, taking advantage of the fact that the resulting terpins are not volatile with steam. Other substances having a methene group in a side chain are also very easily hydrated by dilute sulfuric acid, for example, dihydrocarveol and iso- pulegol, OH /\ dihydrocarveol "Ann. S60, 82 (1908). 130 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS H \ isopulegol The facility with which such unsaturated groups are hydrated af- fords an explanation of the rearrangement of many unsaturated sub- stances in the presence of dilute mineral acids, for example, = terpinenes THE ETHYLENE BOND 131 The rearrangement of 2-methylbutene-(3) to trimethylethylene by dilute acids is probably effected in the same manner, 55 CH 3 CH 3 OH CH 3 >CHCH = CH 2 -> >CHCH< -* CH 3 CH 3 CH 3 CH 3 Thus when commercial amylene is hydrated by sulfuric acid, the resulting alcohol is chiefly "amylene hydrate"- or dimethylethylcar- binol, 56 obtained in very pure condition from trimethylethylene, 57 - CH 3 CH 3 >C=CH.CH 3 > >C.OH.CH 2 CH 3 CH 3 CH 3 Amylene and alcoholic sulfuric acid yields amyl methyl ether. 58 Unsaturated Hydrocarbons and the Refining of Petroleum Oils. From the foregoing section, it is clear that treatment of petroleum distillates with sulfuric acid does not completely remove the unsatu- rated hydrocarbons but partly polymerizes them. The polymers thus formed are not removed with the "acid sludge," but are found in the treated and washed oil. This accounts for the relatively large pro- portions of high boiling fractions usually obtained when a so-called cracked gasoline is refined by sulfuric acid and then redistilled. 59 When the sulfuric acid from a refining operation is diluted with water an "acid oil" is precipitated which, in the case of gasoline and kerosene, has a pronounced odor due chiefly to the alcohols present. Acid oil from the lower boiling distillates, gasoline and kerosene, contain little tarry matter. Pure mono olefines of the aliphatic series do not yield 55 On account of this tendency of unsaturated substances to rearrange, in the presence of sulfuric or other mineral acids, the method of determining the constitution of unsaturated hydrocarbons by oxidation by chromic acid is not to be relied upon. The same consideration applies to the oxidation of certain alcohols, for example, a CH a substance containing the group > CH.CH 2 CHOH CHa R would undoubtedly CH 3 yield a mixture of oxidation products, among which acetone derived from > C = CH R would be found. CH 3 x 86 It will be noted that the alcohols derived from the hydration of ethylene double bonds are always tertiary or secondary alcohols ; the hydroxyl group becomes attached to the more "positive" carbon atom. The industrial manufacture of alcoholic solvents from low-boiling olefines, derived from petroleum or the commercial "amylene" obtained as a by-product of the manufacture of oil gas or Pintsch gas, has been attempted. The "acid oils" obtained by diluting the sulfuric acid used in refining gasoline made by pressure distillation or similar methods also contains secondary and tertiary alcohols. Although the tertiary alcohol, dimethyl ethyl carbinol, boiling point 102, is an excellent solvent for cellulose nitrate, it cannof be acetylated by ordinary methods. Like the majority of tertiary alcohols, it has a camphor-like odor. 67 Wischnegradsky, Ber. 10, 81 (1877) ; Ann. 190, 332, 366 (1878). 58 Reychler, Chem. Zentr. 1907, I, 1125 ; Henry, Bull. Acad. roy. Belg. 1906, 261. 59 Cf. Brooks & Humphrey, loc. cit. The proportions of such high boiling polymers contained in a refined oil will be greater if the duration of the treating operation is prolonged, or the mixture allowed to stand. 132 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS tars with concentrated sulfuric acid at ordinary temperatures, but diolefines, particularly those containing conjugated double bonds, re- act very energetically with sulfuric acid, forming tars and reducing the acid. Thus highly unsaturated oils, made at very high temperatures, such as crude benzene derived from oil gas or Pintsch gas manufac- ture, react violently with sulfuric acid on account of the cyclohexadiene and other di-olefines contained in such oils. When gasoline or kerosene containing unsaturated substances is re- fined by sulfuric acid and then redistilled, liberation of sulfur dioxide is always noted. This is present in the alkali washed oil, prior to dis- tillation, in the form of neutral or dialkyl esters of sulfuric acid, (RO) 2 S0 2 . These esters are decomposed on heating, yielding tarry matter and sulfur dioxide. Theory indicates that refining with a mini- mum of sulfuric acid leads to the formation of neutral or dialkyl esters, which partly remain dissolved in the treated oil, and greater proportions of sulfuric acid favor the formation of acid or mono-alkyl esters which are readily washed out. Practice confirms this suppo- sition; oils refined by relatively small quantities of acid contain more sulfur, in a form appearing as S0 2 on heating, than oils treated with relatively larger quantities of acid. It is also evident from the foregoing section that the per cent by volume of unsaturated hydrocarbons contained in a certain distillate cannot be accurately ascertained by treating with sulfuric acid. The usual practice has been to determine the loss on treating with concen- trated sulfuric acid but it is evident that the formation of polymers entirely destroys the quantitative character of such a determination. Such tests are of qualitative value only. The results obtained by em- ploying sulfuric acid, Sp. Gr. 1.84 are too low, at least for gasolines and kerosene, and the results obtained when fuming sulfuric acid is employed are too high since Worstall 60 has shown, and it is a matter of common experience that fuming sulfuric acid attacks saturated hy- drocarbons. Fuming sulfuric acid also sulfonates any aromatic hydro- carbons which may be present. No accurate quantitative method is now known for the determination of the percent by volume of un- saturated hydrocarbons in a mixture containing also saturated hydro- carbons (probably of various types) and aromatic or benzenoid hy- drocarbons. 60 Am. Chem. J. 20, 664 (1898). The original method as recommended by Kramer and Bottcher specified the use of fuming sulfuric acid. Worstall obtained yields of 30 to 40% of the sulfonic acids of n.hexane, n. heptane and n.octane. According to Markownikow naphthenes are simultaneously sulfonated and oxidized by fuming sulfuric acid. (J. Rusa. Phys.-Chem. Soc. 1892, 141.) THE ETHYLENE BOND 133 Other Reactions of Olefines. The oxidation of unsaturated hydrocarbons by air or oxygen is nearly as general a reaction as the reaction with ozone, although much less energetic than the latter. The oxidation of turpentine, and the formation of what are now recognized as peroxides, was noted by Schoenbein in his well-known studies of oxidation, hydrogen peroxide and ozone and, Berthelot like Schoenbein, wrote of ozone formation when turpentine is oxidized by air. Fudakowski 61 noted that light petroleum fractions acquired oxidizing properties similar to oxidized turpentine, when these oils were exposed to light and air. Kingzett 62 first proved that ozone was not present and attributed the ability of such oxidized material to effect the oxidation of other substances, to the presence of a peroxide or "hydrated oxide." A great deal of ex- perimental work on this subject was done many years ago, but the whole matter was greatly clarified by Engler and Weissberg, 63 Bach 64 and others and the general character of the "autoxidation" of these unsaturated hydrocarbons finds close parallels in the air oxidation and resinification of rubber, particularly prior to vulcanization, the oxi- dation and consequent deterioration of rosin, copals and varnishes, the drying of linseed and similar oils and the deterioration of many sub- stances by oxidation brought about by some second unsaturated sub- stance occurring with it, for example, the destruction of cellulose fiber when in contact with lignin or rosin sizing. Engler and Weissberg showed that "the oxygen combines as molecular oxygen," and that "a peroxide is formed which may then rearrange to ordinary oxides, or may react upon other unoxidized substance." In the case of turpen- tine, the per cent of peroxides present after oxidation at temperatures up to 160 decreases rapidly with rising temperature, and a sample rich in peroxides, formed at low temperature, is rapidly altered by heating, the peroxides being decomposed, with further oxidation of the turpentine. As surmised by Kingzett and later shown conclusively by Clover and Richmond 65 organic peroxides are hydrolyzed by water forming hydrogen peroxide, which accounts for the many positive re- actions for this substance obtained by the earlier investigators. Engler "Ber. 6, 106 (1873). 62 J. Chem, 8oc. 12, 511 (1874). ^Vorgange d. Autoxydation, 1904; Ber. SI, 3050 (1898). "Compt. rend. 124, 2951 (1897). *>Am. Chem. J. 29, 179 (1903). The oxidizing power of old oxidized turpentine has been utilized in medicine, as an antiseptic, as an antidote for certain poisons, acn as yellow phosphorus, and the more stable peroxides, such as benzoyl peroxide and benzoylacetyl peroxide studied by Clover and Richmond have been tried as anti- septics for diseases of the intestinal tract. 134 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS and Weissberg were able to isolate the peroxides of amylene, trimethyl ethylene, and hexylene in fair degree of purity. Rupture of the ethylene bond by autoxidation has been noted in many instances, aldehydes, ketones or acids being formed ; the methene group >C = CH 2 splits with the formation of formaldehyde or formic acid, as in the case of |3-pinene, limonene 66 and (3-phellandrene. 67 Willstatter sought a catalyst, in the hope that oxidation of unsaturated substances might be effected as easily as hydrogenation in the presence of fine nickel but, although metallic osmium appears to catalyse the reaction and cyclohexene was thus oxidized, in acetone solution, to cyclohexenol, the method has had no further extension. 68 However, the industrial use of catalysts in promoting air oxidation has long been known in the paint and varnish industry where salts or resinates of manganese, lead and cobalt are widely used. The effect of light in accelerating such oxidations has also long been known. In the autoxi- dation of styrene marked polymerization occurs, but in direct sunlight fission of the side chain occurs with the formation of benzaldehyde and formaldehyde. 69 The effect of sunlight in promoting autoxidation has been studied by Ciamician and Silber 70 whose investigations also show that oxidation under these conditions is by no means limited to substances containing an ethylene bond, but very stable ketones such as cyclohexanone and menthone are oxidized and their carbocyclic structure ruptured. Vanadium pentoxide 71 has come into vogue as catalyst for oxidizing a wide variety of substances by means of air at elevated temperatures, for example, naphthalene to phthalic acid or an- hydride. These conditions are quite different from those commonly un- derstood as autoxidation. The oxidation of olefines or saturated non- benzenoid hydrocarbons by this method has not been reported, but judging from their oxidation under very similar conditions the resulting products would probably be water, carbon dioxide, unchanged hydro- carbon and small yields of the simpler aldehydes and acids. Closely related to the subject of autoxidation is the method dis- covered by Prileshajew 72 who has shown that benzoyl peroxide, C 6 H 5 CO.O.OH, combines directly, in cold neutral solvents, with sub- "Blumann & Zeitschel, Ber. 47, 2623 (1914). For the oxidation of ethylene to formaldehyde see Ethylene, Willstatter, Ann. 422 (1921). 87 Wallach, Ann. 348, 30 (1905) ; 362, 291 (1908) ; Kingzett has noted the corrosion of metal containers, used for turpentine, due to solution of the metal by formic acid. *Ber. 46, 2952 (1913). 08 Stobbe, J. prakt. Chem. 1914, 551. Ber. 42, 1510 (1909) ; 46, 3077 (1913). 71 Senderens employed it for oxidizing alcohols. [J. Chem. Soc. 1913, I, 814.] Naphthalene and benzene are also oxidizable by its aid. "Ber. 42, 4812 (1909) ; J. Russ. Phys.-Chem Soc. 43, 609 (1911) ; 44, 613 (1912). THE ETHYLENE BOND 135 stances containing an ethylene bond. The initial product readily de- composes to give an oxide of the original olefine, and these oxides are generally very easily hydrolysed to glycols. The method was applied particularly to the oxidation of linalool, geraniol, citral and citronellal. The hydrocarbons di-isobutylene, decylene and the terpenes limonene and pinene yield oxides, which may be hydrolysed to glycols, which suggests that the autoxidation of other unsaturated hydrocarbons, for example, unsaturated petroleum hydrocarbons, may lead to the for- mation of glycols as one of the minor products, when moisture, suffi- cient for hydrolysis, is present. Probably the best known method of oxidizing the olefine group for the purpose of determining the constitution of organic substances is that of oxidizing by cold dilute potassium permanganate. Thus trimethyl ethylene gives a very good yield of the corresponding gly- col, 73 and diallyl yields a hexyl erythrite. An excess of permanganate results in further oxidation of the glycol with a break in the carbon atom chain, as in the rupture of the double bond in ct-pinene to form pinonic acid. 74 This break in the carbon atom structure of a substance does not always occur at the point at which the double bond was origi- nally located, as has been shown in the case of carvenone and ter- pinenol-(4). Nevertheless, this method of oxidation and the ozone method are the most reliable means yet discovered of determining the position of ethylene bonds in organic substances. The reaction of sulfur with unsaturated hydrocarbons has been little investigated. According to H. Erdmann 75 sulfur exists at 160 largely as S 3 or thiozone, and at this temperature he succeeded in forming a "thiozonide" of linalyl acetate C 12 H 20 2 S 3 and was unable to obtain a derivative containing less than three atoms of sulfur. Friedmann, 76 however, isolated a compound C 10 H 12 S by reacting upon dicyclopentadiene with sulfur. 77 By heating sulfur and turpentine together at 150 a viscous product containing 30 to 50 per cent of sul- fur can be obtained. 78 The reaction of sulfur with unsaturated hydrocarbons is of interest in connection with the vulcanization of rubber. In addition to the evi- dence furnished by the ozone reaction, the action of oxygen upon thin 'Wagner, Ber. U, 1230, 3343 (1888). *Baeyer, Ber. 29, 22 (1896). 3 Ann. 362, 133 (1908). * Ber. 1,9, 50, 683 (1916). 7 Koch, German Pat. 236, 490 (1909), prepares sulfur derivatives of terpenes by heating with sulfur until hydrogen sulflde is evolved. "Pratt, U. S. Pat. 1,349,909. 136 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS films of rubber indicates the presence of two double bonds for each C 10 H 16 complex, two molecules of oxygen being combined, 79 and when treated with sulfur chloride 80 the limit of the reaction corresponds more closely to Weber's (C 10 H 16 S 2 Cl 2 )n than to Hinrichson's [(C 10 H 16 ) 2 S 2 Cl 2 ] n . When pure Ceylon para, gutta-percha and ex- tracted and purified balata are treated with 37 per cent of sulfur at 135 the final products are apparently identical and correspond to the empirical formula (C 10 H 16 S 2 ) n . 81 In the ordinary hot process of vulcanization, using about 10 per cent of sulfur the first stage evi- dently consists in adsorption, followed by slow chemical combination, 82 and when sulfur chloride is employed adsorption followed by slow chemical combination appears to be the result. 83 Vulcanization is essentially an increase in the degree of polymeri- zation of the rubber and when this is effected by means of sulfur or sulfur chloride, it is probable that combination also occurs between sulfur atoms attached to different complexes or molecules, since the ten- dency of sulfur derivatives to polymerize is well known, as for example the thio-aldehydes. The literature on the subject of vulcanization is voluminous, and is burdened by much speculative matter which will not be reviewed here; the subject is complex and the effect of varia- tions in mechanical treatment, and the presence of other substances, is often very marked. These effects are of great importance to the rubber industry but' are not of general interest. The causes of the variability of the vulcanization of plantation Hevea rubber have been particularly well investigated 8 * and recently a large number of sub- stances have been investigated which promote further polymerization independently of sulfur or which greatly accelerate the vulcanization when sulfur is employed. Thus para nitrosodimethylaniline, one of the most potent accelerators, when added in amounts equivalent to 0.33 to 0.5 per cent, reduces the time required for vulcanization to about one-third that normally required and the proportion of sulfur may also be somewhat reduced. That many mineral substances, such as litharge, red lead, zinc oxide, magnesium oxide, etc., accelerate vulcanization by sulfur has long been known but a large number and variety of organic substances also function in this manner. A large number of aromatic nitro derivatives, piperidine and quinoline and "Peachey, J. 8oc. Chem. Ind. 31, 1103 (1909). 8 Kirchof, Kolloid Z. 14, 35 (1914). 81 Spence and Young, Kolloid Z. 13, 265 (1913). 82 Harries, Ber. 1,9, 1196 (1916). 13 Hinrichson, Chem. Abs. 12, 104 (1918) ; van Rossem, CTiem. Aba. 12, 2142 (1918). "Eaton & Grantham, J. Soc. Chem. Ind. S^ 989 (1915). THE ETHYLENE BOND 137 their derivatives, amines and substituted amines and ureas, have been found to have accelerating effects. 85 Barium peroxide alone has no vulcanizing effect but benzoyl peroxide does "vulcanize" in the absence of sulfur 86 but the product is markedly different from the commercial products made by the use of sulfur or sulfur chloride. 87 Dubosc has insisted that colloidal sulfur, which he assumes is formed by the inter- action of hydrogen sulfide and sulfur dioxide, produced in situ during vulcanization, is solely responsible for the vulcanization effects. This opinion is not commonly held but it is of interest in view of the fact that a process of cold vulcanizing has recently come into use which consists in treating rubber with a mixture of these two gases, sulfur being formed in an extremely finely divided state. Reychler 88 showed that rubber takes up nearly 25 times as much sulfur dioxide as C0 2 , under comparable conditions, and Peachey 89 has taken advantage of this fact in his process of vulcanization just alluded to. The saturation of the double bonds in rubber by sulfur explains the value of "hard rubber" in handling hydrochloric, hydrofluoric and other acids. The action of sulfuric or other mineral acids upon unvulcanized rubber has been but very little investigated. Addition of Ozone. That ozone is capable of reacting with unsaturated hydrocarbons has been known for many years, the reaction of ethylene and ozone to form formaldehyde, formic acid and carbon dioxide having been noted by Schoenbein ; 90 also the reaction between benzene and ozone was studied by Houzeau and Renard 91 but the reaction product was regarded as a peroxide rather than an ozonide. The true character of these reactions was first made clear by Harries, who pointed out that reaction with ozone in the absence of moisture gave thick viscous substances, which were very explosive, but which he was able to show by analysis consisted of products containing 3 for each double bond present in the original substances. These ozonides can break down in two ways as follows, 1 By reaction with water to form hydrogen peroxide and ketones or aldehydes accompanied by complete rupture of the double bond. 88 Twiss, J. Soc. Chem. Ind. 36, 782 (1917) ; King, Met. & CJiem. Eng. 15, 231 (1916). 88 Ostromuislenski, J. Russ. tf, 1462 (1915). 87 Twiss, loc. cit. 88 J. chim. phys. 8, 617 (1910). 89 Peachey & Shipsey, J. Soc. Chem. Ind. 1921, 4 T. 80 J. prakt. Chem. 66, 282 (1855). 91 Compt. rend. 76, 572 (1873). 138 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS >C C< + H 2 >>CO + OC< + HA 2 Decomposition can take place on warming or in solvents such as absolute alcohol or glacial acetic acid in the absence of water to give a peroxide and a ketone or aldehyde. >C - C< --- C< I + OC< The peroxides formed as in equation (2) can often react with water to form a carboxylic acid; for example, mesityl oxide ozonide breaks down in accordance with the two schemes just shown as follows: (a) (CH 3 ) 2 C CH.CO.CH | H,0 (CH 3 ) 2 CO + OCH . CO . CH, (b) (CH 3 ) 2 C CH.CO.CH 3 -> (CH 3 ) 2 CO |>CH.CO.CH, (bj Decomposition according to (b) and (b ) accounts for the fact that the yield of methylglyoxal is relatively small and formic and CH 3 ho"t~- pulegone ozonide p-methyl- adipic acid 1-methylcyclo- hexanedione-(8, 4) THE ETHYLENE BOND 139 acetic acids are formed. % This type of decomposition accounts for re- actions which were for a time considered abnormal, for example, pulegone ozonide 92 yields (3-methyladipic acid and not the substance which would be expected from the character of the great majority of ozonide decompositions, namely, 1-methylcyclohexanedione- (3, 4) Similarly camphene gives a little camphenilone and a relatively large yield of a lactone, whose formation is attended by rupture of the six carbon ring. 93 HCHO Harries regards these peroxides formed by the decomposition of ozonides as having the constitutions indicated in the above examples. Another type of peroxide is formed by the direct action of ozone upon carbonyl derivatives, aldehydes or ketones, thus nonyl aldehyde acted upon by ozone forms a labil peroxide melting at about 10, but a more stable peroxide of the same empirical formula CH 3 . (CH 2 ) 7 .CH0 2 is formed by the decomposition of the ozonides of substances containing the group CH 3 (CH 2 ) 7 CH = CHR. This more stable peroxide melts at 73 and can readily be recrystallized. The reaction of ethylene bonds with ozone is substantially as gen- eral a reaction as is the reaction of bromine. In fact, ozone reacts with many substances, which are commonly regarded as not having unsatu- rated bonds of the ethylene type, for example, benzene and naphtha- lene. It is of interest to note that the ethylene bond in fumaric acid, which substance is not hydrated by sulfuric acid, reacts only very slowly with ozone, but when prepared by employing very concentrated ozone the ozonide spontaneously decomposes on standing, yielding the original substance, fumaric acid. This reaction, which has been em- ployed so successfully by Harries in the investigation of various kinds of caoutchouc, has been an outgrowth of his studies of the reaction of ozone upon mesytilene, amylene, 2 . 6-dimethylheptadiene- (2, 5), diallyl, and similar substances. In this connection, it should be men- tioned that conjugated dienes react very energetically with ozone to K Harries, Ann. 374, 297 (1910). M Semmler, Ber. Jfi, 246 (1909) ; Palmen, Ber. *S, 1432 (1910). 140 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS form mono-ozonides, but the diozonides are formed only very slowly. Diallyl, or 1, 5 hexadiene, in chloroform solution readily gives a very explosive syrupy diozonide, which on hydrolysis yields succinic dialde- hyde and formaldehyde. A CH 2 CH = CH 2 CH 2 CH CH 2 CH 2 CHO > I > I +2HCHO r^TT r^Ti r^Ti /^TJ r^tr r^tr OTJ r^~crr\ U1 2 L/1 = O1 2 O1 9 O1 Oxlo Lyl 9 OilU V o 3 Geometrical isomers of the type of fumaric and maleic acids yield identical ozonides or rather identical hydrolytic products, which fact may serve to establish the structural similarity of such isomers. The work of Harries on the constitution of certain unsaturated hydrocarbons has clearly shown that most of them are in reality mix- tures of isomers, a fact brought out in the section on the preparation of unsaturated hydrocarbons. Thus the octadiene made by the action of methyl-magnesium iodide on succinicdiethyl ester and decompo- sition of the resulting glycol or its bromide was supposed to have the constitution, (CH 3 ) 2 C = CH.CH =C(CH 3 ) 2 but the ozone method clearly shows that this hydrocarbon is in reality a mixture chiefly con- CH 2 CH 2 sisting of the hydrocarbon C.CH 2 CH 2 C (2.5 di- CH 3 CH 3 methylhexadiene (1.5) ). When the alkaloid pseudo-pelletierin is decomposed by the method of exhaustive methylation, the basic nitrogen atom is removed and a cyclo-octadiene results which Willstatter and Veraguth 94 were inclined to regard as containing a pair of conjugated double bonds. Their cyclo-octadiene polymerized with remarkable ease. Nevertheless, Har- ries showed that this hydrocarbon forms a diozonide which is hydro- lyzed normally yielding succinic dialdehyde, and succinic acid, indi- cating that the hydrocarbon is cyclo-octadiene (1.5). CH 2 - -CH 2 CH 2 CH^CH CH 2 CH 2 HO . N (CH 3 ) 2 CH 2 > CH 2 CH 2 CH 2 CH 2 CH, CH, CH = CH "Ber. 38, 1975 (1905) ; 40, 959 (1907). THE ETHYLENE BOND 141 A CH 2 CH CH CH 2 CH 2 CHO I | 2 | > CH 9 - CHO H 2 CH CH CH, v 3 Harries has summarized his researches on ozonides in four general articles. 95 Some of the ozonides first prepared by Harries in the earlier period of his researches were not pure and consisted apparently of mixtures derived by the addition of 3 and also the hypothetical com- pound 4 or oxozone. After passing the crude ozone-oxygen mixture through 5 per cent caustic soda and then through sulfuric acid, the gas then gave pure ozonides. Polymeric forms of ozonides have frequently been noted, an oily volatile monomolecular form having usually a sharp disagreeable odor, and polymers in the form of solid gummy, glassy or crystalline substances having little or no odor, usually being observed. Thus monomolecular butylene ozonide can be distilled in vacuo and it readily dissolves in the common solvents. The dimeric form of butylene ozonide, however, is an almost odorless gummy sub- stance very sparingly soluble in water. Formation of ozonides at low temperatures, below 0, favors larger proportions of the polymeric forms. Unsaturated cyclic hydrocarbons behave toward ozone and sub- sequent hydrolysis generally like aliphatic defines. Cyclopentene ozonide, C 5 H 8 3 , is soluble- in the common solvents and is smoothly hydrolyzed by water resulting chiefly in the mono-aldehyde corre- sponding to glutaric acid, CH 2 CH )CH > H 2 CH 2 CH 2 CH-- ..\ I.... CH 2 COOH \CH > CH 2 CH 2 CHO CH 2 CH 2 M Ann. Stf, 311 (1905) ; 374, 288 (1910) ; S90, 236 (1912) ; W, 1 (1915). 142 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Cyclohexene ozonide is much more stable to water but on long boil- ing yields hexane dialdehyde and adipic acid. Limonene readily yields a mono-ozonide, the isopropenyl side chain 96 first being reacted upon, and at a much slower rate the cyclic double bond is attacked. 97 By the ozonization of natural Hevea rubber, Harries obtained lae- vulinic acid and laevulinic aldehyde, 98 from which observation he concluded that Hevea rubber is a polymer of the di-isoprene, 1.5- dimethy 1 cy clo-octadiene- (1.5), = C(CH 3 ).CH 2 CH 2 C = CH - CH CH 5 However, the real unit, which is polymerized to ring complexes an unknown number of times, is the group CH 2 C(CH 3 ) = CH.CH 2 . Artificial isoprene rubber, on treating with ozone and subsequent hy- drolysis, yields succinic acid and acetonylacetone in addition to lae- vulinic acid and aldehyde, 99 which products could conceivably be de- rived from 1 . 6-dimethy Icy clo-octadiene- (1 . 5) . The first work upon the treatment of petroleum distillates with ozone appears to have been done by Molinari and Fenaroli, 100 who ob- tained a yield of 32 per cent of an ozonide from a kerosene fraction, boiling at 295-300, derived from a Russian petroleum. The subject has not been pursued further but inasmuch as Harries observed that refined petroleum ether and hexane are not altogether unacted upon by ozone when used as solvents for unsaturated substances, the con- clusions of Molinari that a conjugated di-olefine, C 17 H 30 , was present in the relatively large proportions indicated by the yield stated, are hardly to be accepted. Ethane is reacted upon by dilute ozone at 100, the initial oxidation products being ethyl alcohol and acetalde- hyde. 101 During the recent war period Harries turned his attention 96 Prior to the researches of Harries, vanillin had been made by the action of ozone on iso-eugenol. (Otto, Ann. d. Chim. & Phys. [7] 13, 120 [1898] ; German Pat. 97, 620.) Better yields were, for a time, obtained by using crude ozonizing apparatus giving dilute ozone, about 1%, than when using more concentrated ozone made by improved apparatus. Harries later showed that 70% yields could be obtained by treating the ozonide with zinc dust and acetic acid (Ber. 48, 32 [1915].) The side chain in safrol also reacts readily, Semmler and Bartlett (Ber. J t l, 2751 [1908]), obtain- ing homopiperonylic aldehyde. 97 Critical examination of Harries' work is apt to elicit the fact that he frequently paid little attention to the history or purity of his original material and also that more definite results might often have been obtained, in the terpene series, in the hands of other well-known specialists in this field. *Ber. 38, 1195, 3986 (1905) ; 46, 733 (1913) ; Ann. 406, 173 (1914). "Steimmig, B&r. 47, 350 (1914). 100 Ber. 41, 3704 (1908). 101 Bone and Drugman, Proc. Chem. Soc. 20, 127 (1904). THE ETHYLENE BOND 143 to the oxidation by ozone of the highly unsaturated oily distillates obtained by the low temperature carbonization of lignite. Although the method had a large scale trial in Germany during the stress of conditions imposed by the war, the yields of fatty acids obtained were very small and the project was soon abandoned. 102 Harries identified stearic, palmitic and myristic acids among the reaction products, to- gether with relatively large proportions of simpler, water soluble acids, including formic, acetic, propionic and oxalic acids. The reaction of unsaturated hydrocarbons with sulfur trioxide is naturally a very energetic one leading, under ordinary experimental conditions, to oxidation of the hydrocarbon and formation of S0 2 . A definite reaction product is easily obtainable with ethylene, the crystal- line anhydride carbyl sulfate being formed. CH 2 CH 2 -S0 2 \ || +2 S0 3 >| >0 CH 2 CH 2 0-S0 2 No further work on this reaction seems to have been done since its discovery in 1838 103 and whether ethylene homologues can form similar derivatives (at low temperatures in a neutral solvent) is not known. The anhydride carbyl sulfate reacts energetically with water 10 * to CH 2 .S0 3 H form ethionic acid, I which substance is then rapidly CH,.O.S0 3 H CH 2 .S0 3 H. hydrolyzed to iso-ethionic acid Sulfonic acid groups in CH 2 .OH which sulfur is bound directly to carbon, as in iso-ethionic acid, are not easily displaced and alcohols or glycols cannot be made from them by any known methods. The propane derivative, propanol-(l) sulfonic acid- (3), is formed when allyl alcohol reacts with an alkali bisulfite, CH 2 OH. CH 2 OH. CH + KHS0 3 CH 2 CH 2 CH 2 .S0 3 K. 102 Ozone, as an oxidizing agent to be employed in industrial operations, Is usually much too costly compared with other methods of oxidation, although its cost may be expressed largely in terms of the cost of electrical power. 103 Regnault, Ann. 25, 32 (1838) ; Magnus, Pogg. Ann. 47, 509 (1839). '"Claesson, J. prakt. Chem. (2), 19, 253 (1879). 144 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS The behavior of ethylene bonds to sulfur dioxide and aqueous sul- furous acid is very different from sulfuric acid in that hydration to alco- hols does not occur, but addition to form very stable sulfonic acids is frequently the result. This reaction follows the general rule that other adjacent groups exert a very great influence upon the reactivity of the unsaturated bond. Anhydrous sulfur dioxide has not been shown to react with unsaturated hydrocarbons, although the very marked solu- bility of such hydrocarbons in liquid sulfur dioxide and the complete- ness with which they may be extracted from paraffine hydrocarbon mixtures, as in the Edeleanu refining process, might be considered as an indication of the formation of such labil compounds. When solu- tions of amylenes or butylene in sulfur dioxide are subjected to the action of heat and light, amorphous hornlike solids are formed, 105 the butylene compound having the composition (C 4 H 8 S0 2 ) n , and when the conjugated diene, isoprene, is allowed to stand two days in liquid sulfur dioxide a crystalline substance C 5 H 8 S0 2 , is formed. 106 Sulfonic acid derivatives of sabinene, sabinol and pulegone are formed when S0 2 is passed into their cooled alcoholic solutions 107 but the formation of sulfonic acid derivatives has been most frequently ob- served in cases where the ethylene bond is adjacent to a carbonyl group, >CH = CH C . Thus acrolein 108 and crotonic aide- hyde 109 react with sodium bisulfite normally so far as the aldehyde group is concerned but the ethylene bonds react also, to form stable sulfonic acid derivatives, which are not affected by treating with alkali. The aldehyde group generally reacts more readily with bisulfite than the ethylene bond and advantage is taken of this fact in isolating un- saturated aldehydes, such as citral and citronellal, from mixtures containing them. Citral contains two double bonds, one of them adja- cent to the aldehyde group, and both ethylene bonds may react yield- ing the stable disulfonic acid salt, C 9 H 1T . (S0 3 Na) 2 .CHO, from which citral cannot be regenerated. When cold neutral sodium sulfite is em- ployed and the alkali, formed by the reaction, is neutralized as fast as formed, C 9 H 15 .CHO + 2Na 2 S0 3 + 2H 2 -* C 9 H 15 (S0 3 Na) 2 CHO + 2NaOH 10B Mathews and Elder, J. 8oc. Chem. Ind. 1915, 670. 10 de Bruin, Chem. A 6s. 9 t 623 (1915). 1OT Wallach, Nachr. Wiss. Ges. Goettingen. 1919 t 321. "Muller, Ber. 6, 1442 (1873). "Haubner, Monatsh. 12, 546 (1891). THE ETHYLENE BOND 145 then an unstable dihydrosulfonate is formed from which citral is easily regenerated. 110 Citronellal contains only one double bond and this is far removed from the aldehyde group and accordingly less reactive. Under the conditions just described citronellal is not reactive; 111 with cold concentrated bisulfite, in the presence of sodium bicarbonate, the aldehyde group only reacts, and normally, but when warmed with an excess of bisulfite (containing a little sulfite) the stable sulfonate is formed. 112 CH 3\ OH C = CH.CH 2 CH 2 CH.CH 2 CH< I, H oso * Na citronellal \ CH 3 s x OH hot \ _ CH2 C H 2 CH 2 CH . CH 2 CH < / 1 I OS0 2 Na S0 3 Na CH 3 cannot be regenerated. Nitrosyl chloride has been a most useful reagent in the investiga- tion of the terpenes but has not been used in the investigation of unsaturated hydrocarbons derived from petroleum, although the first thorough study of the addition of nitrosyl chloride to defines was carried out with amylene. 113 When Wallach first undertook the study of the terpenes the literature had become confused with a va- riety of names for hydrocarbons which were not clearly differentiated, one from another, and the names adopted usually referred to various particular sources. The reaction with nitrosyl chloride, which had been discovered by Tilden and Shenstone, 114 proved to be a most valuable reagent for the preparation of characteristic crystalline de- rivatives of these unsaturated hydrocarbons and the multiplicity of names began to diminish as the identity of differently named terpenes was established. The addition products formed by the reaction of these hydrocarbons with nitrogen trioxide and nitrogen tetroxide also proved useful in this connection. Crystalline tetrabromides, dihydro- chlorides, etc., also assisted in this work of identification and "cit- rene," "hesperidene," and "carvene," for example, were shown to be identical and are known as limonene. 110 Tiemann, Her. SI, 3306, 3315 (1898). m Tiemann, B&r. S2, 816, 818 (1899). 112 Tiemann, Ber. SI, 3306 (1898). 118 Wallach, Ann. 245, 246 (1888). 114 J. Chem. Soc. 1877, I, 554. 146 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Among the simpler olefines the ability of defines to combine with nitrosyl chloride increases with molecular weight (or introduction of the "positive" methyl groups) : ethylene forms only ethylene chloride, and propylene forms both the dichloride and nitrosochloride. 115 As indicating the variety of ethylene types which form nitrosochlorides, the following may be mentioned, trimethyl and tetramethyl ethyl- ene, cyclohexene, methene derivatives R 2 C = CH 2 , and also hy- drocarbons having a semicyclic double bond and a side chain, as CH 2 (CH 2 ) X < >C = CH.R. Unsaturated hydrocarbons having the CH 2 groups >C = CH 2 , and CH = CH 2 , do not usually yield crystalline nitrosochlorides: 116 the type R 2 C = CHR usually does yield crystal- line nitrosochlorides. 117 The terpene, (5-fenchene forms a crystalline nitrosochloride and Wallach has obtained such crystalline derivatives from other hydrocarbons whose double bond is similarly situated. 118 The crystalline nitrosochlorides, nitrosites and nitrosates are gen- erally bimolecular 119 and hence called bis-nitrosochlorides, 6is-nitro- sites and fo's-nitrosates, but in solution many 12 of these derivatives are blue in color and are monomolecular. Many of the nitrosochlorides, in monomolecular form, are volatile with steam without decomposition ; for example, the blue modifications derived from the hydrocar- bons, 121 - 122 ""Tilden & SudbVough, J. Chem. Soc. 63, 479 (1893). "Meyer, "Analyse & Konstitutioniren. org. Verb," Ed. 2, Berlin, 1909, p. 939. 11T Weyl, "Die Methoden d. org. Chemie," II, 639 (1911). 118 Awn. Slil, 322 (1906) ; 865, 267 (1909). 118 Baeyer, Ber. 28, 641, 650, 1586 (1895) ; 29, 1078 (1896). 120 Wallach & Sieverts, Ann. S06, 279 (1898), 332, 309 (1904), showed that pinol nitrosochloride may exist in a colorless monomolecular form. 121 Wallach, Ann. 353, 308 (1907) ; 396, 280. 122 The preparation of nitrosochlorides is best carried out by dissolving the hydro- carbon in an equal volume of glacial acetic acid, adding one volume of ethyl nitrite, cooling to 10, and then adding one-third volume of concentrated hydrochloric acid. In most cases, where a crystalline nitrosochloride is possible, an abundant crystalline deposit of the nitrosochloride forms in a few minutes. Acetone is generally the best solvent for recrystallizing these derivatives. Nitrosobromides are also easily prepared but are less stable than the corresponding chlorides : 5 CC tetramethyl-ethylene and 5 CC ethyl nitrite, cooled to C and treated with 5 CC concentrated HBr solidifies in a few minutes to the solid bisnitrosobromide. THE ETHYLENE BOND 147 =C(CHJ. The value of the nitrosochlorides has been chiefly their ready con- version to oximes, from which ketones may be made, and to other more stable substances suitable for identification purposes, for example, condensation with benzylamine. All three types of nitroso derivatives may be converted into the isomeric oximes by carefully warming with alkalies, (CH 3 ) 2 C-C1 (CH 3 ) 2 C.C1 CH-CH.NO * CH 3 -C = N. 3 It has been proposed to utilize the reaction with nitrogen tetroxide in determining the constitution of defines, since the addition product RCH.N0 2 .CHN0 2 .R, is split by heating with concentrated hydro- chloric acid to give fatty acids. 123 Ammonia and aliphatic amines react with the ethylene bond in cer- tain instances where the very reactive ethylene-carbonyl group >CH = CH C occurs, as in mesityl oxide, 124 (CH 3 ) 2 C = CH.CO.CH 3 + NH 3 CH 2 COCH 3 Vinyl chloride, which polymerizes on GtapHin* but. is nnite stable to sulfuric acid, reacts with ammonia to give ethylenediamine, 125 CH 2 = CHC1 + 2NH 3 - ^NH 2 . CH 2 CH 2 NH 2 HC1 The unsaturated hydrocarbons themselves are not reactive to am- monia or amines. The ethylene-carbonyl group is also reactive to hydroxylamine in a fairly large number of substances. Allyl ketones, CH 2 = CH . CH 2 COR, react normally to give oximes but the ethylene bonds in propenyl and vinyl ketones also react, 126 123 Jegorow, J. prakt. Chem. S6, 521 (1912). 12 * Sokoloff, Ber. 7, 1387 (1874) ; Kohn, Monatsh. 25, 135 (1903) ; Blaise & Maire, Compt. rend. 142 f 215 (1906). 125 Engel, Compt. rend. Wk, 1621 (1887). 128 Blaise, Compt. rend. 138, 1106 (1904) ; 1J,2, 215 (1906). 148 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS R CH 2 = CH . COR - -4 NH . OH . CH 2 CH 2 C N.OH The ethylene-carbonyl group also reacts with aniline, phenyl hydra- zine, urea, semicarbazide, mercaptans, hydrogen sulfide, hydrocyanic acid, malonic and acetoacetic esters. C0 2 R C0 2 R aniline 127 + RCH = C< > RCH CH< C0 2 R I C0 2 R NH.C 6 H 5 CH.COOH NH CO CH 2 urea l28 +\\ >| CH 2 CO NH CH 2 semicarbazide 129 + (CH 3 ) 2 C = CHCOCH 3 CH 3 > (CH 3 ) 2 C CH 2 .C = N.NH.CO.NH 2 NH.CO.NH.NH 2 ethyl mercaptan 13 + (CH 3 ) 2 C = CH.CO CH S SC 2 H 5 SC 2 H 5 hydrogen sulfide 131 + carvone > (C 40 H 14 0) 2 H 2 S hydrocyanic acid 132 + (CH 3 ) 2 C = CH.COCH 3 -->(CH 3 ) 2 C CH 2 COCH 3 CN Blank, Ber. 28, 145 (1895). 128 Fischer & Roeder, Ber. 8J h 3751 (1901). 129 Important in connection with the use of semicarbazld for the identification of such ketones ; citronellal and two molecules of semicarbazid gives the crystalline semi- carbazino-semicarbazone immediately. Cf. Rupe, Ber. 86, 4377 (1903) ; Semmler, B&r. 1^1, 3991 (1908). Posner, Ber. 85, 799 (1902) ; 37, 502 (1904). 181 Cf. Wallach, Ann. 279, 385 (1894) ; 31,3, 32 (1905) ; mesityloxide and the hydro- carbon menthene also form compounds with H 2 S. 182 Lap worth, Jour. CTiem. Soc. 85, 1214 (1904). THE ETHYLENE BOND 149 acetoacetic ester 133 + CH 2 = CH.C0 2 R CH 3 COCH.C0 2 R CH 2 CH 2 .C0 2 R The above reactions cannot be said to be general reactions even for a, (3-unsaturated ketones or acids (the "ethylene-ketone" group), and none of them have so far been found applicable to hydrocarbons. In fact, until the mechanism of such reactions, and the part played by the carbonyl group, is understood, it is questionable whether this last group of reactions should really be considered as reactions of the un- saturated bond ; it would be more correct to consider them as reactions of the>CH = CH C or "ethylene-carbonyl" group. These con- siderations apply also to the hydrolytic rupture of the ethylene bond of this group which is noted when many substances, for example, mesityl oxide, citral 134 and pulegone, are treated with dilute alkalies or min- eral acids, thus + H 2 (CH 3 ) 2 C = CH.CO.CH 3 (CH 3 ) 2 CO + CH.COCH 3 (CH 3 ) 2 C = CH.CH 2 CH 2 C = CH.CHO CH, citral + H 2 (CH 3 ) 2 C = CH.CH 2 CH 2 C = O CH 3 + CH 3 CHO Metallic sodium has been observed to combine directly with un- saturated hydrocarbons 135 only in a few cases where a negative group is present, as in stilbene, C 6 H 5 CH = CH 2 . Metallic sodium in a very finely divided or colloidal form is employed for the purpose. Preparation of defines. As regards the preparation of olefine hydrocarbons, it may be pointed out that most methods of preparation yield a mixture of isomers 183 Vorlander, Ann. 94, 317 (1897). ls *Verley, Bull. soc. chim. (3), 17, 175 (1897). Effected best by heating with potassium carbonate ; pulegone may be hydrolyzed by heating with water in an auto- clave. Wallach, Ber. 32, 3388 (1899). ""Schlenk, Appenrodt, Michael & Thai, Ber. )ft, 473 (1914). 150 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS and that often when a single product is theoretically probable a mix- ture of isomers results owing to rearrangement. Thus Eltekow 136 showed that isobutyl alcohol or the corresponding halides decompose to give a mixture of three butylenes, CH 3 CH 3 >CH.CH 2 X -- > CH 3 CH 3 CH 3 CH 2 CH = CH 2 Haber m showed that on heating normal hexane to 600 to 800, it is decomposed, methane and amylene being found among the products formed and assumed that the amylqne was alpha amylene, as expressed by the equation, C 3 H 7 . CH 2 CH 2 CH 3 -- > C 3 H 7 CH = CH 2 -f- CH 4 But, as indicated in the case of the butylenes, olefines of the type RCH = CH 2 are prone to rearrangement under the influence of heat. CH 3 Thus the amylene >CH.CH = CH., rearranges under the influ- CH 3 ence of heat, or mineral acids, to trimethylethylene, CH 3 >C = CH.CH 3 CH 3 Noorduyn 138 has made a study of the constitution of the olefines formed by heating barium fatty acid salts with sodium ethoxide or methoxide. Very little work of this kind, critical examination of the constitution of acyclic olefines, has been done. Five methods have been used to determine the constitution of unsaturated substances. (1) The method of Varrentrap; fusion with caustic potash which causes a change of position of the double bond toward the carboxyl group (of fatty acids). (2) Oxidation by potassium permanganate, in which method 1:2 glycols are former as intermediate products. (3) Beckmann's transposition, in which method the formation of the dibromide is the first step. Ber. 13, 2404 (1880) ; Cf. Nef, Ann. SIS, 1 (1901). m Ber. 29, 2691 (1896). Rec. trav. ch4m. S8, 317 (1919). THE ETHYLENE BOND 151 (4) Jegorow's method based upon the addition of N 2 4 to the double bond. (5) Harries' method consisting in reaction with ozone and hy- drolysis of the resulting ozonide. Noorduyn used the ozonide method. Examination of the decylene made by heating the barium salt of undecylenic acid with sodium ethoxide and hydrolyzing the ozonide of the hydrocarbon yielded for- maldehyde, acetic, propionic, butyric, valeric and hexylic acids show- ing that the hydrocarbon is a mixture of isomers. Similarly the hepta- decylene, from oleic and elaidic acids and sodium methoxide was shown to be a mixture of isomeric hydrocarbons. Nonylenic acid, from oenan- thole and malonic ester, yielded a mixture of octylenes and "(3-octyl- ene," boiling-point 124-126 from secondary octyl alcohol was also shown to be a mixture. Primary alkyl iodides or alcohols invariably yield a mixture of hy- drocarbons, as a critical examination of the physical properties of the hydrocarbons described in the literature as having been prepared in these ways, shows. Thus Morgan and Schorlemmer 139 prepared a hex- ene, boiling-point 68 to 70, from a monochlorohexane and Zelinsky and Przewalski 14 heated n-hexyl iodide with quinoline and obtained a liquid mixture boiling from 35 to 67. On oxidizing the fraction boiling from 63.5 to 65 they obtained a mixture of butyric and valeric acids indicating that this hexene fraction was probably a mix- ture of the a and (3-isomers. Van Beresteyn 141 also obtained a hexene boiling at 67.7 to 68.1 by decomposing n-heptyl alcohol by heating in contact with nickel at 220. CH 3 (CH 2 ) 3 CH 2 CH 2 CH 2 OH - CH 3 (CH 2 ) 3 CH = CH 2 ? + CO + 2H 2 However, von Braun 142 obtained a hexene, by gently heating n-hexyl trimethyl ammonium hydroxide, which showed a boiling-point of 62 to 63 and which he regarded as a-hexene, although he was unable to prove the constitution of it on account of the small quantity made. Brooks and Humphrey 143 confirmed the character of von Braun's a-hexene by synthesizing it by means of a reaction which had been applied by Tiffeneau 144 to the synthesis of allyl derivatives of ben- Ann. m, 305 (1875). 140 Ghent. Zentr. 79, II, 1854 (1908). 141 Ibid. 1911, II, 1017. 142 Ann. SSS, 22 (1911). 143 J. Am. Chem. Soc. \0 t 833 (1918). M Compt. rend. I39 t 481 (1904). 152 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS zene. 145 They treated n-propyl magnesium bromide with allyl bromide, the reaction taking place smoothly at room temperature. C 3 H 7 MgBr + BrCH 2 CH = CH 2 > C 3 H 7 CH 2 CH = CH 2 + MgBr 2 CH 3 Iso-a-heptene, >CH.CH 2 CH 2 CH = CH 2 and iso-a-octene were CH 3 also made in a similar manner. This reaction is undoubtedly appli- cable to the preparation of a large number of alpha defines of this series. Pure olefines of other types may be prepared in certain cases by making use of the symmetry of the parent alcohol or halide. For ex- ample, the tertiary alcohol triethylcarbinol readily yields pure y-ethyl- (3-pentene (3-ethyl pentene-2). C 2 H 5 C 2 H 5 C 2 H 3 C OH- --*C 2 H 5 C C 2 H 5 CH. CH, and 5-iodo heptane, C 3 H 7 CHI C 3 H 7 , by virtue of its symmetry, yields pure y-heptene when decomposed by caustic alkali. The simpler primary alkyl chlorides and bromides on treatment with caustic alkali in methyl or ethyl alcohol yield chiefly methyl or ethyl ethers but alkyl iodides of five or more carbon atoms, particu- larly secondary and tertiary derivatives, yield olefines almost quanti- tatively. 146 Organic bases, aniline, quinoline and the like have frequently been employed to remove halogens with success. Tertiary halides, like ter- tiary alcohols, are easily decomposed and Klages 147 has employed pyri- dine for tertiary chlorides. Decomposition of alkyl halides by heat is usually attended by rearrangements and often with rupture of the car- bon structure, and, in some cases, by condensation or polymerization, but many substances catalyze this decomposition of the halides, so that lower temperatures may be employed and subsequent changes of the olefines may be minimized. Nearly quantitative yields of ethylene 145 Auster well later synthesized isoprenes in a similar way, treating vinyl-magne- sium bromide with beta-chloropropylene, / CH = CH 2 / Cl / CH = CH 2 '*' ! "1 Mg +CH 3 C MgBrCl + CH 3 C \ Br \CH 2 \ CH 2 J. Chem. Soc. Abs, 102, 525 (1912). 149 Nef, Ann. S09, 126 (1899) ; 318, 1 (1901). Ber. 35, 2633 (1902). THE ETHYLENE BOND 153 may be obtained from ethyl chloride by heating in contact with barium chloride and chloropentanes can be decomposed to amylenes in this manner. 148 A large number of processes for the decomposition of bornyl chloride to camphene have been described in connection with the syn- thesis of camphor. Bornyl chloride is remarkably stable and most of the successful reactions are carried out within the range 160 to 190. Usually an alkaline substance or mixture is sought which will dissolve the bornyl chloride forming a homogenous reaction mixture and pat- ented methods refer to the use of sodium phenolate, sodium soaps such as oleate, linoleate, etc., sodium acetate in acetic acid, aniline, quino- line, etc. Zinc chloride catalyzes the decomposition of bornyl chloride but rapidly polymerizes the camphene formed. Certain defines are often most readily made by the decomposition of halogenated, hydroxy or unsaturated fatty acids but these methods are by no means generally applicable. Pure p-butylene is easily made from bromotiglic acid by heating with soda in aqueous solution. 149 CH 3 CH 3 CHBrCH > CH 3 CH = CHCH 3 + C0 2 + NaBr C0 2 Na The p-halogen derivatives of the fatty acids are decomposed very easily by alkalies and the resulting unsaturated acids frequently lose C0 2 on heating or distilling to give an olefine. p-bromoisobutyric acid is quantitatively decomposed by aqueous barium hydroxide to methyl CH 3 acrylic acid, 'CH . CO 2 H > C . CO,H and a-bromo- CH/ 7 " butyric acid, considerably more stable, also yields methyl acrylic acid by treating with 25% caustic soda. 150 Normal p-bromo fatty acids on heating with water, dilute alkali, or by destructive distillation, yield more of the a-(3-unsaturated acid than the p-y-unsaturated acid. Wallach 151 has made use of the instability of the p-hydroxy acids to " 8 Badische, German Pat. 255,519, J. Chem. Soc. 104, 438 (1913); German Pat. 268,100, Chem. Zentr. 1914, I, 308; Sabatier & Mailhe, J. Chem. Soc. 104, 330 (1913) ; Braun and Deutsch, Ber. 45, 1271 (1912). According to Mathews, Bliss and Elder, the decomposition of alkyl halides within tne range 100-700 is catalyzed by water, either in the presence or absence of other catalysts. Brit. Pat. 16,828 (1912) ; 17,234 (1912). 6 Pagenstecher, Ann. 195, 112 (1879). "Engehorn, Ann. 200, 68 (1880) ; Bischoff, Ber. 4, 1041 (1891). lsl Ann. S65, 257 (1909). 154 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS synthesize defines containing the methene group. Using the method of Reformatsky of condensing ketones or aldehydes with bromoacetic ester by means of zinc, Wallach proceeded as indicated by the follow- ing reactions, the hydroxy acids being dehydrated by heating with acetic anhydride. and from nopinone (3-pinene was synthesized, \ C=CHC0 2 H C=cii Tertiary alcohols decompose so readily that they are difficult to acetylate. Glacial acetic acid at 150 to 155, or acetic anhydride con- taining a little zinc chloride, or sulfuric acid, yields chiefly unsaturated hydrocarbon. Thus trimethyl carbinol yields isobutylene, 152 diethyl- propylcarbinol yields an octene, 153 etc. Primary and secondary alcohols of high molecular weight may be converted to olefines by the method of Krafft, 15 * i. e., treating with palmityl chloride and distilling the palmitic ester slowly at ordinary pressure. In the terpene series the method developed by Tschugaeff 155 consisting in heating the methylxanthogenate ester of the alcohol, has given excellent results. Very little heat is usually required and the probability of rearrangement or decomposition is greatly lessened; in fact, the methylxanthogenate esters of tertiary alcohols, if formed, de- compose spontaneously at ordinary temperatures. Henderson 156 pre- 162 Menschutkin, Ann. 197, 204 (1879). 163 Mason, Compt. rend. 132, 483 (1901); Henry, Compt. rend. 1U, 552 (1907); 147, 1260 (1908). 154 Ber. 16, 3020 (1883). 166 Ber. 32, 3332 (1899). 166 J. Chem. Soc. 91, 1620 (1910) ; 99, 1903 (1911). THE ETHYLENE BOND 155 pared very pure bornylene from the methylxanthogenate ester of bor- neol. Heating primary or secondary alcohols with mineral acids rarely gives good results except with the simpler members, as in the well- known methods of preparing ethylene and propylene, using sulfuric or phosphoric acids. Wallach showed that concentrated formic acid 15T or oxalic acid 158 give better results in the terpene series than mineral acids. Potassium acid sulfate has been employed with good results, as in the conversion of borneol, which is relatively quite stable, to cam- phene. 159 When potassium acid sulfate or phosphorus pentoxide is used to dehydrate cyclohexanol-1-acetic acid, cited above, the resulting product is A 1 - 2 cyclohexene acetic acid instead of the A 1(7) acid which is obtained with acetic anhydride. Sabatier and Mailhe 16 have shown that phosphorus, carbon, anhy- drous calcium sulfate, basic aluminum sulfate and many metallic oxides promote the dehydration of alcohols. The corresponding olefine hydro- carbon is usually produced, although alumina at 210 causes some ether to be formed. Ipatiev found that under higher pressures the formation of ether was considerably increased. Baskerville was unable to detect ether in ethylene resulting from the decomposition of alcohol in con- tact with thoria at temperatures as low as 250. 161 Sabatier and Mailhe studied a series of catalysts and, within the temperature range 300-350, ethyl alcohol gave varying yields of ethylene and hydrogen, the latter being formed together with acetaldehyde, CH 3 CH 2 OH >CH 3 CHO + H 2 . Thoria, alumina and blue oxide of tungsten at 340-350 gave practically quantitative yields of ethyl- ene and the other catalysts gave the results indicated in the following table: Per cent ethylene Th0 2 : 100. A1 2 O 3 98.5 W 2 3 98.5 Dehydration and dehydrogenation Cr 2 0, 91. ^ SiO 2 . 84 TiO 2 63 BeO 45 ZrO 2 45 U 3 8 24. 23. FeoO 3 . 14. V 2 O 3 . 9 ZnO . 5. J Ann. 291, 361 (1896) ; S56, 243 (1907). 168 Ann. 275, 106 (1893). 169 Wallach, Ann. 230, 239 (1885). 190 Ann. Ohim. Phys. VIII. 20, 289 (1910). 191 J. Am. Chem. Soc. S5, 93 (1913). 156 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS SnO 0. ^ CdO 0. MgO 0. f Dehydrogenation Cu ..! 0. Ni 0. J Engelder 162 showed that in the presence of A1 2 3 , Si0 2 , Zr0 2 and Ti0 2 the equilibrium, alcohol ? water + ethylene, could be displaced by the addition of water vapor to the incoming alcohol. Kaolin within the range 350-400 is particularly efficient in producing ethylene. 163 Ipatiev, prior to the work of Sabatier and Mailhe quoted above, had shown the wide applicability of alumina as a dehydrating catalyst. 164 He prepared isobutylene and pure propylene using alumina as a cat- alyst and butylene has been made from n. butyl alcohol, on an indus- trial scale, by the same method. 165 Senderens found that amorphous silica is much more active than ground quartz and aluminum phos- phate is also an excellent catalyst; Senderens obtained noteworthy re- sults by decomposing cyclohexanol at 300, to cyclohexene and men- thol to menthene. 166 Pinacone gives excellent yields of dimethylbuta- diene when passed over alumina at 450, but the best yields are ob- tained in vacuo. The decomposition of tertiary amines has been employed as a laboratory method but the preparation of these amines is compara- tively costly and difficult though several processes for the preparation of dienes, leading to the synthesis of rubber, which involve this ter- tiary amine method, have been patented. 168 As pointed out above, the decomposition of the tertiary amines usually takes place at compara- tively low temperatures, thus lessening the probability of decompo- sition or rearrangement of the resulting olefine. Willstatter and Schmaedel, 169 made cyclobutene in this way. CH 2 CH NH 2 CH 2 CH N (CH 3 ) 3 CH 2 CH CH 2 CH 2 CH 2 CH 2 OH CH 2 CH + N(CH 3 ) 3 + H 2 The Grignard reaction has sometimes been applied to the synthesis of olefines in ways other than noted above. In rare instances the Grig- 182 J. Phys. Chem. 21, 676 (1917). 163 The activity of these catalysts is gradually diminished as they become impreg- nated with carbon, evidently formed by the decomposition of ethylene to methane and carbon. ^Ber. 36, 1997 (1903) ; 3J,, 596, 3579 (1901). MS Newman, Can. Chem. J. 1920, 47; King, J. Chem. Soc. 115, 1404 (1919). 166 Compt. rend. 11^, 1109 (1907) ; 1J,6, 125 (1908) ; Bayer & Co. Brit. Pat. 4,076 (1913). 167 Badische, French Pat. 417,275 (1910). 168 J. Chem. Soc. 102, I, 821 (1912). 169 Ber. 38, 1992 (1905). THE ETHYLENE BOND 157 nard complex RC< OMgX R, breaks down spontaneously, but heat is usually required to effect decomposition to the olefine. 170 The decomposition of hydrocarbons by heat ha& not been employed for the preparation of pure olefines, but it is well known that the pyro- lytic products of paraffine, petroleum oils and the like are rich in olefines. Pressure, as in distillation of heavy oils under pressure, di- minishes the proportion of olefines in the product and decomposition by heat under vacuum increases the proportion of olefines. Dilution of the original hydrocarbon vapors with an inert gas or steam also has this effect. 171 The reduction of the ketone group to the CH 2 group, in the pres- ence of a cyclopropane ring or unsaturated bonds of the ethylene type and without reducing these double bonds, may, in certain instances, be accomplished by forming the hydrazine derivative of the ketone and then decomposing this by solid caustic potash. Thus carone gives carane, without rupture of the cyclopropane ring, and ionone yields the corresponding unsaturated hydrocarbon. 172 CH. CH=CH C=N.NH 2 CH 3 CH: "o Harries & Weil, Ann. 348, 363 (1905) ; Klages, Ber. S9, 2306 (1906) ; Barbier & Locquin, Chem. Zentr. 1913, II, 28. 171 Greenstreet, U. S. Pat. 1,110,925. "'Kishner, J. Russ. Phys.-Chem. Soc. 43, 1398, 1563 (1911). Chapter V. Acyclic Unsaturated Hydrocarbons. Remarkably few hydrocarbons of this series are known. Many which have been described are undoubtedly mixtures and the constitu- tions assigned to many of them are undoubtedly incorrect. This is particularly true of olefines of the type RCH 2 CH = CH 2 . The simpler olefines are very reactive and the most promising outlook for the chemical utilization of petroleum is undoubtedly in the direction of these simpler olefines, including the gaseous olefines, ethylene and propylene, and the low boiling highly reactive olefines such as the butylenes, amylenes and hexylenes. Ethylene One liter of ethylene under standard conditions weighs 1.2519 grams. 1 Its boiling point at 760 mm. according to Cailletet 2 is 105 and according to Ladenburg and Kriigel 3 is 105.4 ; Burrell and Robertson 4 give 103.9 as the boiling-point. Its melting-point is 169. Its critical temperature is 9.5 0.1, critical pressure 50.65 0.1 atmospheres. 5 Its heat of combustion is stated to be 333,350 and 341,400 calories (Thomsen 6 ) and 345,800 calories (Mixter 7 ). Data on the compressibility of ethylene and the extent of its deviation from the behavior of a perfect gas under pressure at ordinary atmos- pheric temperature have recently been published, compressed ethylene, in steel cylinders, for welding and cutting now being commercially available. 8 Water at dissolves approximately 0.25% ethylene. The gas is markedly soluble in ammoniacal cuprous chloride, but not in am- 1 Ct. Malisoff & Egloff, J. PTiys. Ghem. 23, 65 (1919). 'Compt. rend. 94, 1224 (1882). Ber. 32, 1818 (1899). *J. Am. Chem. Soc. 37. 1893 (1915). *J. Chim. Phya. 10, 504 (1913). Thermochem, Unters, 4, 64. 7 Am. J. Sci. (4), 4, 51 (1897). According to British Patents 146,332 and 147,051 (1920), ethylene may be separated from coal gas or oil gas by passing the gas through a series of absorbent materials at low temperatures. 158 c: E 3 S-Q 300 <3 -er ID Nef, Ann. 298, 202 (1899). 21 J. Ghent. Soc. 115, 144 (1892). 22 J. Gliem. Soc. 83, 1074 (1903). *'Ann. 422, 36 (1921). "Harries, Ann. &$, 288 (1910). ACYCLIC UNSATURATED HYDROCARBONS 163 and formic acid results. 25 According to Taylor 26 ethylene and oxygen react at ordinary temperatures in the presence of activated charcoal. Chromic acid oxidizes ethylene, with difficulty to C0 2 , formic and acetic acids. 27 Potassium permanganate in dilute sulfuric acid yields C0 2 , formic and acetic acids, but neutral or alkaline permanganate yields glycol and oxalic acid. 28 Ethylene combines directly with a large number of substances and while many of these reactions have been known for a great many years, a few of them have become industrially important only within very recent years. Ethylene chlorohydrin was made by Carius 29 and others by treating ethylene with dilute aqueous hypochlorous acid. Gomberg 30 has recently shown that the reaction of ethylene and hypo- chlorous acid takes place so rapidly that practically quantitative yields of the chlorohydrin are obtained by agitating ethylene with cold chlor- ine water, although free chlorine is also present. Methods suit- able for large scale manufacture of ethylene and propylene chlo- rohydrins, using chlorine and cold aqueous solutions of sodium car- bonate or bicarbonate, have recently been described. 31 Ethylene bro- mohydrin has recently been made by passing ethylene and bromine vapor separately into ice water, keeping the concentration of the bro- mine in the solution very low. 32 The bromohydrin had previously been made by the action of HBron ethylene glycol or by the action of PBr 3 on the glycol. The bromohydrin boils with slight decomposition, at 146-150 and has a density, 20, of 1.7629. Ethylene chlorohydrin reacts with sodium azide, the chlorine being replaced by the triazo group. 33 (Vinyl bromide does not react with sodium azide.) By converting the triazoethyl alcohol to the bromide and replacing the bromine with iodine, the resulting triazoiodine deri- vative can be decomposed by alkali, removing HI and yielding triazo- ethylene. The boiling-point of triazoethylene is 26, or 10 higher than the corresponding bromide CH 2 = CHBr. CH,C1 CH,N 3 CH 2 N 3 CHN 3 | +NaN 3 J >| >|| CHOH CH 2 OH CHJ CH 2 "Berthelot and Gaudechon, Compt. rend. 150, 1327 (1910). 29 Trans. Am. Electrochem. Soc. 1919, 167. 27 Chapman & Thorpe, Ann. V&, 182 (1867) ; Othmar & Feidler, Ann. 197, 243 (1879). 28 Ann. 150, 373 (1869) ; Ber. 21, 1234 (1888). Cf. Evans on oxidation of ethylene glycol by permanganate, J. Am. Chem. Soc. 41, 1385 (1919). "Ann. 126, 197 (1863) ; Butlerow, Ann. 144, 40 (1867). so J. Am. Chem. Soc. 41, 1414 (1919). 31 Brooks, Chem. d Met. Eng. 22, 629 (1920). 82 Read & Hook, J. Chem. Soc. in, 1214 (1920). "Forster & Newman, J. Chem. Soc. 97, 2570 (1910). 164 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Ethylene chlorohydrin gave promise of becoming of some impor- tance during the recent war, as an intermediate in the manufacture of dichloroethyl sulfide (mustard gas), but another reaction of ethylene, i. e., its reaction with sulfur chloride, first discovered by Guthrie, 34 proved to be more suited to large scale production and was adopted in all the Allied countries. The experimental conditions, of which Guth- rie's work gave little more than a hint, were worked out by Pope 35 and the large scale operations worked out by Levinstein. Gibson and Pope 36 showed that when the reaction between ethylene and sulfur chloride is carried out above 70 considerable decomposition occurs and Pope and his assistants showed in a later paper 3T that practically quantitative yields are obtained when the ethylene contains a little alcohol vapor but when pure ethylene is employed the product is not so pure, thus explaining the discrepancies reported by other workers. The sulfur liberated in the reaction appears to be retained largely in a colloidal condition and may be separated by dissolving the dichloro sulfide in kerosene and then separating the mustard gas from the kerosene solution by chilling. Distillation in vacuo readily yields pure (3(3- dichloroethyl sulfide. In Germany the chlorohydrin method of making mustard gas was employed. Ethylene reacts with selenium monochloride 38 to give free selenium and the product Cl 2 Se(CH 2 CH 2 Cl) 2 . The reactions of the two manufacturing processes for mustard gas are as follows: (1) CH 2 || +HOC1 CH 2CH 2 OH +N a 2 S H 2 C1 | C CH 2 C1 (2) 2 34 Ann. 119, 91 (1861) ; 121, 108 (1862). 36 J. Soc. Chem. Ind. 38, 344R, 434R (1919) ; Green, J. Soc. Chem. Ind. 38, 363R, 469R (1919). M J. Chem. Soc. 117, 271 (1920). 87 J. Chem. Soc. 119, 634 (1921). 88 Bausor, Gibson & Pope, J. Chem. Soc. 117, 1453 (1920) : Heath & Semon J Ind d Eng. Chem. 12, 1100 (1920). ACYCLIC UNSATURATED HYDROCARBONS 165 CH 2 C1 CH.C1 + S CH 2 S CH 2 Phosgene reacts with ethylene under the influence of light as fol- lows. 39 CH 2 CH,C1 _j_COC! 2 > I !H,COC1 v^ j-j>2 v>i -f COC1 2 > I (Chloropropiony 1 chloride) . CH CI Norris and Couch 40 have shown that benzoyl chloride reacts with ethylene in the presence of anhydrous aluminum chloride to give phenyl vinyl ketone, a reaction probably capable of considerably wider appli- cation. Ethylene is readily absorbed by anhydrous aluminum chloride and benzene to form mono- and poly-substituted ethyl benzenes. 41 Chlorine and bromine 42 react smoothly with ethylene to give the symmetrical dihalides. The addition of chlorine to ethylene to form ethylene chloride or "Dutch liquid" was first carried out in 1796 and Faraday later treated oil gas 43 with chlorine obtaining ethylene chlo- ride together with other chlorinated products. Although this reaction has been known since this early date, no very thorough study of it has been made. Dry chlorine and ethylene react exceedingly slowly. Ethyl- ene passed into chlorine water yields ethylene chlorohydrin almost exclusively; cold dilute bromine water yields both ethylene bromide and ethylene bromohydrin. In chlorinating ethylene, it is difficult to limit the chlorination to the dichloride, trichloroethane and still more highly chlorinated products being formed. The introduction of ethyl- ene into liquid chlorine in the cold, under pressure, gives excellent yields of ethylene chloride, 44 and chlorination in the presence of char- coal, alumina or other very porous material is stated to give good yields. 45 Another patentee employs solid calcium chloride as a cat- alyst. 46 Higher defines, amylenes and hexylenes, yield dichlorides when treated with sulfuryl chloride below 30. 47 When chlorine is absorbed in cold bromine in the proportions required by the hypothetical sub- "Lippman, Ann. 129, 81 (1864). 40 J. Am. Chem. Soc. J&, 2330 (1920). 41 Balsohn, Bull. Soc. chim. (2), 31, 529 (1879). 42 According to Plotnikov, Chem. Abs. 1917, 48. ethylene and bromine react even at 80 in the dark. 43 The oil gas used by Faraday was probably made from fatty oils, which, however, closely resembles oil gas made from mineral oils in its general character. 44 Curme, U. S. Pat. 1,315,545 ; 1,315,547. 45 Harding, Brit. Pat. 126,511 (1918). 46 Sniythe, Gas. J. 149, 691 (1920). A yield of 50% ethylene chloride by this method is reported. 4r Badische Co., J. Soc. Chem. Ind. 1912, 151. 166 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS stance BrCl and ethylene is then introduced, the principal product is CH 2 Cl.CH 2 Br, which appears to be the only real evidence of the ex- istence of the compound Cl.Br. 48 Bromine has practically no action on ethylene bromide. Iodine reacts slowly with ethylene in direct sunlight or on heating to 60. 49 Iodine monochloride and ethylene yields ethylene chloride and free iodine. Concentrated hydriodic and hydrobromic acids combine slowly with ethylene at 100. Boron trifluoride reacts with ethylene forming the product C 2 H 3 BF 2 boiling at 125. 50 The reaction of sulfuric acid and the simpler gaseous olefines has recently become a matter of industrial interest since the alkyl sulfates may be saponified or hydrolyzed by steam to the corresponding alco- hols. As first described by Faraday 51 in 1827, the absorption of ethylene by concentrated sulfuric acid, with the formation of ethyl hydrogen sulfate, is not rapid below 160 and according to Butlerow 52 the absorption is rapid at 160-170. The question of an industrial synthesis of alcohol from gases containing ethylene was investigated in 1855 by Berthelot 53 and later by P. Fritzsche n4 who states that 100 kilos of concentrated acid are required to produce 18 kilos of alcohol. The chief difficulties have been the handling and re-concentration of relatively large quantities of sulfuric acid and loss of acid by oxidation and charring of other olefines, which were not completely removed prior to the absorption of ethylene. One patentee claims that vana- dium or uranium salts facilitate the absorption of ethylene. 55 Ferrous sulfate and cuprous salts are said to promote the absorption of ethyl- ene and under these conditions 56 the gas is treated with acid at 100- 120. Very recently, Bury and Ollander 57 have carried out these re- actions on an industrial scale, in England, and state that one ton of Durham coal yields sufficient ethylene to produce 1.6 gallons of 95 per cent ethyl alcohol. After removing benzene vapors and olefines other than ethylene, the gas is scrubbed by hot 95 per cent sulfuric acid and the resulting ethyl-hydrogen sulfate is hydrolyzed by steam. Since "Delepine & Ville, Bull. Soc. chim. 27, 673 (1920). "Faraday, Phil. Trans. IS, 118; Regnault, Ann. d. Chimie. (2), 59 (1835). ^Landolt, Ber. 12. 1586 (1879). ^Pogg, Ann. 9, 21 (1827). 52 Ber. 6, 196 (1873). 63 Compt. rend. 40, 102 (1855) ; Ann. Chim. (3), 43, 385 (1855). <*Chem. Ind. 20, 266 (1897) ; 35, 637 (1912). 55 Lattre, French Pat. 468,244 ; J. Soc. Chem. Ind. 1914, 953 ; Loisy, Compt. rend. 170, 50 (1920). 66 Brit. Pat. 152,495, J. 800. Chem. Ind. S9, 833 A (1920). "Brit. Pat. 147,360 (1914) ; Chem. Weekblad. 17, 478 (1920). ACYCLIC UNSATURATED HYDROCARBONS 167 coal gas ordinarily contains not over 2.5 per cent ethylene, it would be reasonable to assume that such a process would be more successful with oil gas or waste gas from petroleum stills, particularly cracking or coking stills, gas from the latter source containing 5 to 6 per cent ethylene. Propylene and butylenes are absorbed by sulfuric acid at ordinary temperatures and according to Hunt 58 and Ellis 59 good yields of isopropyl and secondary butyl alcohols are obtainable in this way. By the Ellis process isopropyl alcohol is obtained from propylene con- tained in the gases from Burton stills used in cracking petroleum oils to make gasoline. The gases are allowed to bubble through cool sul- furic acid of specific gravity 1.8 until the gravity falls to 1.3 or 1.4. ^_ if 30 3r c * CH 3 CC1I.CH 3 CH 3 CC1 = CH 2 + HBr > CH 3 CC1 . Br . CH 3 ___ S CH,Cl.CH CH,Br * ' ( CH 2 Cl.CHBr.CH 3 dark CH 3 CH = CHC1 + C1 2 > CH 2 C1 . CH = CHC1 light CH 3 CH = CHC1 + C1 2 - > CH 3 CHC1 . CHC1 2 dark CH 3 CC1 = CH 2 + C1 2 - > CH 2 C1 . CC1 . = CH 2 CH 2 C1 . CC1 = CH 2 + HC1 - -* CH 2 C1 . CC1 2 CH 3 rT r rT T _ P TTp> r , TT R 5 CH 3 CHBr . CH 2 Br * * I CH 3 CH 2 CHBr 2 CH 3 C . Br = CH 2 + HBr - -* CH 3 CBr 2 CH 3 -PTT -LTTR i CH 2 Br.CHBr,CH 3 ^ i -"-t5" ^ ) r^TT T>T^ r^TT OTT "R*, ( ^yXigJor . oxi v_yXi 2 r>r "Plant & Sidgwick, J. Roc. Chem. Jnd. W, 17 T. (1921). "Butlerow, Ann. Itf, 275 (1868) ; Michael, J. prakt. Chem. (2). 60, 445 (1899). "Reboul, Ann. Chim. (5) l|, 461 (1878) ; Michael, Ber. 39, 2787 (1906). ACYCLIC UNSATURATED HYDROCARBONS 171 p(3'-dichloro-n-propyl sulfide, analogous to mustard gas, has been described by Coffey, 75 who obtained it easily from propylene chloro- hydrin by means of Clarke's 76 modification of Victor Meyer's meth- od. 77 Coffey was unable to make the dichloro sulfide from sulfur chlo- ride and propylene although in the case of ethylene the results leave little to be desired. With propylene, condensation to dark colored semi-solid material results, when the reaction is carried out at 50 to 60. - The Butylenes and Amylenes: There are three butylenes, i. e., CH 3 CH 3 CH 2 CH = CH 2 , >C = CH 2 and CH 3 GH = CHCH 3 , the lat- CH 3 ter hydrocarbon being known in cis and trans form, 78 HC.CH 3 HC.CH 3 HC.CH 3 CH 3 C.H cis, boiling-point 1 to 1.5 trans, boiling-point 2.5 When primary or secondary butyl alcohol or the corresponding halides are decomposed, all three butylenes are formed. 79 The diffi- culty of preparing pure olefines has repeatedly been emphasized in these pages. The butylenes occur in oil gas, in the light liquid, con- densed under pressure, from Pintsch-gas, and in the fore runnings of the distillation of crude benzene, particularly whe?^ made by low tem- perature carbonization of coal or from water gas tar. The butylenes are not at present utilized industrially. Their physical properties are very imperfectly known but their boiling points, as recorded, are as fol- lows, Butene- ( 1 ) , boiling-point 5 Butene-(2), " " cis + 1 to 1.5; trans + 2.5. Isobutylene, " " 6. Isobutylene can readily be prepared by dropping tertiary butyl iodide into boiling water, the hydriodic acid being retained by the water. 81 76 J. Ghent. Soc. 119, 94 (1921). 76 J. Chem. Soc. 101, 1583 (1912). 77 The writer experimented with this method, in cooperation with the U. S. Chem- ical War Service in 1917 and 1918, in the effort to utilize the ethylene and propylene in oil gas. The yields of the dichloro sulfide are good, in the case of propylene, but the product is much less toxic than the ethylene derivative. 78 Wislicenus, Ann. 313, 228 (1900). 78 Faworski, J. prakt. chem. (2), 42, 153 (1890); Senderens, Compt. rend. 144, 1110 (1907) ; Newman, Can. Chem. J. 1920, 47 and, King, J. Chem. Soc. 1919, 1404, describe the catalytic decomposition of n. butyl alcohol to butylene, which is then treated with 80% sulfuric acid to obtain secondary butyl alcohol, which in turn may be catalytically dehydrogenated to obtain methyl ethyl ketone. 80 Wislicenus, loc cit. 81 Nef, Ann. 318, 23 (1901) ; Cf. Ipatiev, Ber. 40, 1829 (1907). 172 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Isobutylene is rapidly dissolved by 70 per cent sulfuric acid in the cold; when such a solution made up with 50 per cent acid is warmed to 100 di-isobutylene C 8 H 16 is formed, and when acid of stronger concentration, 80 per cent, is employed tri-isobutylene is formed, illus- trating a very general behavior of olefines, i. e., that the more con- centrated the acid the further the polymerization proceeds. 82 The butylenes form characteristic crystalline nitrosates, or rather 6is-nitrosates, when nitrogen peroxide is passed into cold ether solu- tions; 83 reduction of these nitrosates yields the corresponding diamines. Isobutylene reacts with acetyl chloride in the presence of zinc chloride to form a chloroketone. 84 CH 3 CH 3 > C = CH 2 + CH 3 COC1 > > CC1 . CH,COCH 3 CH 3 CH 3 which decomposes on heating to mesityl oxide. The above reaction is analogous to the reaction between ethylene and benzoyl chloride in the presence of anhydrous aluminum chloride, discovered by Norris and Couch. 85 As noted in connection with the action of sulfuric acid on olefines, the butylenes and amylenes are much more reactive than their higher homologues, and it is therefore probable that in the presence of aluminum chloride the rate of polymerization may greatly exceed that of condensation with other substances as in Norris's reaction. Very probably the^dgher olefines such as the decylenes will give bet- ter yields of condensation products, in the presence of aluminum chlo- ride, than butylene- or amylenes. The amylenes have probably been more thoroughly studied than any of the olefines with the exception of certain of the terpenes. This is perhaps to be explained by the availability of the raw materials, amyl alcohol and petroleum pentane. Of the five possible amylenes, four are definitely known but pentene-(l) certainly never has been prepared in a pure state and it is doubtful if the material supposedly isolated by Brochet, 86 from the distillate of bog head coal, contained any of this hydrocarbon at all. It is also doubtful if pentene-(2) has 82 Butlerow, Ann. 180, 247 (1876) ; 189, 48 (1877) ; Ber. 12, 1482 (1879) ; Brooks & Humphrey, J. Am. Chem. Sac. J f O, 822 (1918). 83 Ssiderenko, Chem. Zentr. 1907, I, 399. "Kondakow, J. Russ. Phys.-Cliem. Roc. 26 } 12 (1894). 85 J. Am. Chem. Soc. >$, 2329 (1920). 88 Bull. chim. & Phys. (3), 7, 567 (1892) ; Wurtz, Ann. 148, 136 (1868), and Wag- ner & Saizew, Ann. 179, 304 (1875), attempted to prepare this hydrocarbon by the reaction of allyliodide and zinc ethyl ; reaction of magnesium ethyl bromide and ally! bromide should yield this hydrocarbon in a pure state, analogous to the preparation of hexene-(l) by Brooks and Humphrey, J. Am. Chem. Soc. 40, 822 (1918). When this hydrocarbon is prepared its boiling point, by analogy from the hexenes, will probably be found to be below 35 instead of 39-40 as given by Brochet, ACYCLIC UNSATURATED HYDROCARBONS 173 been prepared in a fairly pure condition. It is idle therefore to com- pare the physical properties of these isomeric amylenes. The most stable of the amylenes is trimethylethylene and it is formed when any of the other amylenes are prepared at high temperatures. According to Ipatiev 87 2-methylbutene-(3), is almost quantitatively converted into trimethylethylene by passing over heated alumina, CH 3 CH 3 ^ /"ITT /^TT OTT -v ^ O /^TT OTJ ^> \jFL . *utt \jJLL 2 ^ ^ \^i . \jtt. 3 CH 3 CH 3 When ordinary amyl alcohol is passed over alumina at 340-350 all three of the methylbutenes are formed, trimethylethylene being the principal product. 88 CH 3 CH, \ CH.CH 2 CH 2 OH CH 3 \ CH 3 CH, \ = CH.CH, CH 3 CH, C.CH 2 CH 3 CH, Commercial "amylene" is accordingly a mixture of these hydrocarbons containing trimethylethylene as the principal constituent. When such amylene is treated with 70 per cent sulfuric acid in the cold, the prin- cipal product is dimethylethylcarbinol, boiling at 102. CH 3 C.CH 2 CH 3 +H 2 CH 2 CH 3 ( CH 3 = CH.CH H0 CH,. \ C.CH,CH, CH, 87 Ber. S6, 2004 (1903). 88 Senderens, Compt. rend. Ill, 916 (1920). 174 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Sulfuric acid in methyl alcohol and trimethylethylene gives the methyl ether. 89 It is worthy of note, that most of the reactions of the amylenes are applicable to the terpenes, and vice versa. The chemical behavior of the two groups of hydrocarbons is entirely similar, but the use of the word "hydro-aromatic" for the cyclohexane derivatives has probably done a great deal to prevent the full realization of the similarity, one might say homogeneity, of the chemistry of the non-benzenoid hydro- carbons. For the purpose of emphasizing this similarity a number of reactions of amylenes and terpenes will be noted. (1) Addition of HC1 and HBr (in acetic acid solution). CH 3 C.CH 2 CH 3 (2) Addition of nitrosyl chloride. 90 91 CH 3 CH 3 CH 3 = CH.CH, CH 3 A H >C CH.CH 3 CH 3 | Cl NO CH 3 NOC1 H /\ in certain terpenes NO 'H 89 Reychler, CTiem. Zentr. 1907, I, 1125. 80 J. Schmidt, Ber. 35, 3732 (1902) ; 36, 1765 (1903). "Wallach, Ann. 2+5, 245 (1888) ; Ber. 24, 1535 (1891). ACYCLIC UNSATURATED HYDROCARBONS 175 (3) Behavior of nitroso chlorides. 92 Both amylenes and terpenes, FR 2 C CHR-| bimolecular, T R 2 C CR L Cl NO J, crystalline J Cl N.OH monomolecular (4) Behavior of nitrosochlorides and nitrosates; formation of oximes. 93 R Cl R / 'X A H,C C = N.OH. x \ * "1 1 \ nitrosochloride CH. \ //\ // oxime H 2 C C = NOH. nitrosate (5) Behavior of nitroso chlorides: formation of nitrolamines. 94 R Cl V H 2 C C = NOH + H 2 NR -i- 1 A. nitrosochloride nitrolamine (6) Addition of N 2 4 amylenes > bimolecular or bis-nitrosates terpenes - > "Baeyer, Ber. 28, 1586 (1895) ; Ber. 29, 1078 (1896). 93 Best carried out by heating with sodium acetate in acetic acid. Wallach, Ann. STU, 202 ; 379, 135. M Wallach, Ann. 241, 296 (1887) ; 262, 327 (1891) ; Ann. ttf, 253 (1888) ; Stf, 143 (1906). 176 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS (7) Oxidation, parallel behavior, e. g., KMn0 4 trimethylethylene > glycol. terpenes > glycols. (8) Dilute sulfuric acid, CH 3 CH 3 >C = CH.CH 3 - > >C.CH 2 CH 3 CH 3 CH 3 | OH terpenes -- > terpene alcohols (9) Concentrated mineral acids, amylenes - -- > diamylenes -- > triamylenes, etc. terpenes -- > diterpenes - -- > trimerides, etc. (10) Action of zinc chloride, amylenes -- > polymers terpenes - > polymers (11) Behavior on heating, amylenes, rearrangement to more stable form chiefly trimethylethylene, terpenes, rearrangement to more stable forms, e. g., pinenes - -- > dipentene, terpinene phellandrenes - > " " (12) Halides, heated with sodium acetate in acetic acid, chloropentanes --- > amyl acetate -f- amylenes bornyl chloride - -> bornyl acetate + { Also, the behavior of the amylenes and the terpenes to bromine, ozone, catalytic hydrogenation, air oxidation, and many other reactions, is very closely parallel. Pentadienes: The preparation and polymerization of isoprene, pi- perylene and dimethylallene have been discussed in the chapter on poly- merization and the problem of synthetic rubber. Piperylene, ACYCLIC UNSATURATED HYDROCARBONS 177 CH 3 .CH = CH.CH = CH 2 , may be identified by its physical proper- ties (noted in the table on page 231) and by its tetrabromide, 1.2.3.4- tetrabromopentane, known in two stereo- isomeric forms (1) crystalline form, melting point 114.5 and (2) a liquid, distilling at 115-118 (4 mm.). 95 Oxidation of piperylene by permanganate yields formic and acetic acid. Harries 96 endeavored to prove its constitution by means of its reaction with ozone but without success ; it combines only slowly with ozone but the diozonide was so explosive that no definite results were obtained. Auwers 97 concludes from the exaltation of its refractive index that the double bonds are in the conjugated position. The isomer 1 .4-pentadiene, CH 2 = CH.CH 2 CH = CH 2 , is one of the products of the decomposition of pentamethylenediamine nitrite but it has only been isolated in the form of its tetrabromide, 98 melting- point 86-87. The preparation and polymerization of isoprene is also discussed in connection with the subject of synthetic rubber. Isoprene in glacial acetic acid solution combines with two molecules of hydro- gen bromide to form CH 2 Br.CH 2 CBr(CH 3 ) 2 , and with hypochlorous acid to form a dichlorohydrin melting at 82. It condenses with ben- zoquinone when the two are heated together at 120-180, the product melting at 234, and since it yields a dioxime and a tetrabromide Euler and Josephson " conclude that the combination has occurred through the double bonds in the isoprene, and the quinone, the product probably having the following constitution. * CH 3 \/ N/\ \ CH, According to Ostromuislenski isoprene may be estimated when pres- ent in a mixture of butylenes, amylenes, benzene, etc., by shaking with about ten volumes of fuming hydrochloric acid for six hours; the prod- "Magnanini, Gazz. Chim. Ital. 16, 391. "Ann. WO, 1 (1915). " Ber. 49, 827 (1916). 88 Demjanow, Ber. 40, 2590 (1907). "Ber. 53 t 822 (1920). 178 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS uct is washed with cold brine, dried over calcium chloride and distilled. The fraction distilling at 50-90 contains butyl and amyl chlorides, the fraction from 90-130 is separately collected and then, the tem- perature rising rapidly from 130 to 142, the 2.4-chloro-2-methyl- butane fraction in a fairly pure condition is collected at this tempera- ture. Refractionation of the fraction boiling at 90-130 will yield a further small proportion of the dichloride. Ole fines, Six to Nine Carbon Atoms: Very few of the many pos- sible hexenes, heptenes, octenes arid nonenes have ever been prepared, but their properties may be roughly assumed from the behavior of the impure mixtures, which have been prepared and from the properties of olefines of the terpene class, many of which have been carefully investi- gated. Certain hydrocarbons of this series are incorrectly described in the literature, for example, hexene- (1) boils at 62-63, and the hy- drocarbon described by Brochet 10 boiling at 67 which he separated from a distillate from bog head coal, is probably a mixture containing chiefly hexene- (2), (see pp. 151-152). High temperatures, and many chemical reagents, particularly acids, cause such a-olefines to rearrange or the double bond to shift its position. Only reactions employing low temperatures and absence of isomerizing reagents can be expected to produce these a-olefines in any degree of purity, for example, CH 2 CH 2 CH 3 Mg < + BrCH 2 CH = CH 2 - CH 3 CH 2 CH 2 CH 2 CH = CH 2 Br , + MgBr 2 or von Braun's method of decomposing trialkyl ammonium hydrox- ide. 101 The hexene obtained by treating secondary hexyl iodide (from man- nite and HI with alcoholic caustic potash is a mixture of hexene- (1) and hexene- (2). Tetramethylethylene (CH 8 ) 2 C=rC(CH 8 ) 2 , is the best known of the hexenes and is probably the most stable. Of the heptenes only three are known in fairly pure state, and only two of the many possible nonenes are definitely known. These hydrocarbons have been relatively of such little importance that they will not be described in detail. Most of them, as described, are obviously impure and so few of the many possible hydrocarbons are known that it is impossible to learn anything from a study of their physical properties. Bull. Ghim. & PUys. (3) 7, 568 (1892). 101 Ann. 382, 22 (1911). ACYCLIC UNSATURATED HYDROCARBONS 179 As regards their chemical properties, it should be kept in mind that in the majority of cases one is dealing with mixtures. Nearly all of the defines of this series combine with hydroiodic acid in the cold, form nitrosochlorides and nitrosates (which have been definitely de- scribed in but a' few cases) , and behave normally toward most of the reagents affecting olefines. Sulfuric acid yields varying proportions of polymers, alcohols and alkyl sulfuric esters. None of the known chemical reactions of these olefines offer much promise that the unsaturated hydrocarbons in unrefined gasoline will be utilized. They can be removed practically unchanged by extraction with liquid sulfur dioxide and their conversion to alcohols, ketones and acids would not be matters of great difficulty. Such products, if made, would be mixtures and, therefore, entirely unsuitable for certain uses, for example, perfumes, flavoring materials and pharmaceuticals. When one reviews the chemical reactions of such olefines, it is evident that these reactions have been devised and applied chiefly for the purpose of isolation and identification, or for their removal as a nuisance, as for example, the usual method of refining with concen- trated sulfuric acid. 102 It is, therefore, entirely possible that the dis- covery of new reactions will render these petroleum olefines industrially valuable. 103 Octadienes: Conylene, C 8 H 44 . By distilling the ammonium base obtained by exhaustive methylation of coniine, an octadiene is ob- tained boiling at 126 (738 mm.). When benzoylconiine is treated with phosphorus pentachloride 1 . 5-dichlorooctane is obtained. 104 2.5-Dimethylhexadiene-(l.5) 105 is of interest as illustrating a property, quite general among dienes of eight or more carbon atoms, of forming an oxide, the anhydride of the 2 . 5-diol, when treated with 70 per cent sulfuric acid. This substance also illustrates the labil char- acter of the a-olefine or > C = CH 2 group, being converted into diiso- crotyl by the action of alcoholic alkali, 102 That large proportions of the olefines remain in the refined oil as polymers, has previously been pointed out. 103 Tne wr iter suggests that it would hardly be worth while, at least for one who greatly values his time, to enlarge our knowledge of the many possible hydrocarbons between pentane and the terpenes by proceeding along the old preparative lines, and examining the various derivatives by old reactions, and measuring the usual physical properties. Pending the possible development of new reactions, or greatly improved old ones, and the discovery of uses for such products as can now be made, it would seem that the best utilization of such olefines would be their polymerization, perhaps by aluminum chloride or zinc chloride to their much more stable polymers, of value as lubricants. 1M v. Braun & Schmitz Ber. 39, 4866 (1906). 105 Pogorzelsky, J. Rus8. Phys.-Chem Soc. SO, 977 (1898) ; J. Chem. Soc. Abs. 1899, I. 785. 180 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS H 2 SO 70% CH 3 CH 3 - >CH. \ / >C.CH 2 CH 2 C< C.CH 2 CH 2 .C CH 3 CH 3 CH 2 CH 2 CH 3 CH 3 > >C = CH.CH = C< KOH CH 3 CH 3 The latter hydrocarbon also yields this oxide when treated with 70 per cent sulfuric acid. (Oxides containing five or six atoms in the ring are very much more stable than the three membered ring oxides such RCH as ethylene or propylene oxides | >0 (Cf. Cineol.) CH 2 ' Nonadienes: Geraniolene, 2.6-Dimethylheptadiene-(1.5). This hydrocarbon, boiling-point 142-143, is of interest on account of its relation to geraniol and citral, and its conversion to cyclogeraniolene when treated with 65 per cent sulfuric acid. When the oxime of citral is dehydrated by acetic anhydride the nitrile is formed which readily yields geranic acid, C 9 H 15 .C0 2 H. On distillation at ordinary pres- sure, geranic acid loses a molecule of C0 2 and forms "geraniolene." 108 The constitution of this hydrocarbon follows from the structure of citral and, if we accept the structure of citral as found by Barbier and Bouveault 107 the relations between geraniolene and a and p-cycloger- aniolene are as follows: CH 3 CH 3 CH 3 CH 3 CH 2 C ^C = CH. CH 2 CH 2 C - > C CH CH 3 CH 2 CH 3 CH 2 CH 2 CH 3 CH 3 CH = C and >C< >CH 2 10 Tiemann & Semmler, Ber. 26, 2708 (1893). 107 Compt. rend. 122, 393 (1896). ACYCLIC UNSATURATED HYDROCARBONS 181 Tiemanns' conclusions 108 as to the constitution of geraniolene and the cyclogeraniolenes are confirmed by Crossley and Gilling 109 by the synthesis of the supposed intermediate alcohol, and the conversion of the corresponding bromide into a and (3-cyclogeraniolene. CH 3 ' LlBr Decadienes and Decatrienes : Dihydromyrcene, 2 . 6-Dimethy locta- diene (2.6). Boiling-point 166-168, D 150 0.7792. 110 This hydrocar- bon is obtained by the partial hydrogenation of ocimene or myrcene, by means of sodium and alcohol, 111 or by slowly distilling methylger- anic acid. 112 Like geraniolene it is converted into a cyclic hydrocarbon by sulfuric acid (in acetic acid). 113 Kishner's method of converting aldehydes and ketones to hydrocarbons 114 converts citral to an isomer of dihydromyrcene, boiling-point 164.5. Kishner's method reduces the carbonyl group to CH 2 without affecting the ethylene bonds present. CH 3 >C = CH.CH 2 CH 2 C = CH.CHO CH 3 CH 3 CH 3 > > C = CH . CH 2 CH 2 C = CH . CH, CH 3 | CH 3 108 Ber. 31, 816 (3898) ; S3, 3711 (1900). 109 J. Chem. Soc. 97, 2218 (1910). The above structures are also confirmed by the work of Wallach, Ann, 32$, 97 (1902) but are not accepted by Harries and Turk, Ann. 3$3, 331, 362 (1905). 110 Enklaar, Kec. trav. cMm. 26, 164 (1907). 111 Semmler, Ber. 3j f 3126 (1910). U2 Tiffeneau. Compt. rend. U,6, 1154 (1908). 113 J. Russ. Phys.-Chem. Soc. tf, 951 (1911), u *Enklaar, Ber. 41, 2083 (1908). 182 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS 'Myrcene, Ocimene and Alloocimene: C 10 H 16 . These hydrocarbons are isomeric decatrienes, two of the double bonds being in conjugated positions. All three hydrocarbons yield 2 . 6-dimethyl octane on hydro- genation. 115 They are sometimes called "aliphatic terpenes" perhaps because of their empirical formulae C 10 H 16 and the fact that myrcene and ocimene are constituents of essential oils. Myrcene was discovered by Power and Kleber 116 in oil of bay, Pimento, acris (Myrcia acris) , one of the myrtaceae, and thus named by them. It also occurs in oil of hops 117 and in oil of verbena, Lippia citriodoro.* 18 The physical properties of these three hydrocarbons are as follows: Myrcene Power & Kleber Semmler 119 Enklaar 12 Ocimene Van Romburgh m Enklaar 122 Allo-Ocimene Auwers & Eisenlohr 123 Enklaar 122 Boiling -Paint Density 15 167 : 67-38 20mm. 0.8023 171-172 : 67-80 20mm. 166-168 0.8013 Density 15 .176-178 : 73-74 21mm. 0.801 81 30mm. 0.8031 188' : 81 12mm. 0.8119 0.8133 15.6 C n D 1.4673 1.4673 1.4700 n D 1.4861 1.4857 1.54558 1.5447 Myrcene and ocimene, on partial hydrogenation, yield the same di- hydromyrcene (dihydromyrcene tetrabromide melting-point 88), and of the two original hydrocarbons myrcene is much more rapidly resini- fied. Enklaar proposed the following structures for myrcene, ocimene and dihydromyrcene. CH 3 >C=CH . CH 2 CH 2 C CH=CH 2 CH, myrcene ;i H CH CH >C=CH . CH 2 CH,C=CH . CH. " CH 3 t >C=CH . CH 2 CH=C CH=CH CH 3 | C CH 3 reaction-product H ocmene Enklaar, Ber. LI, 2083 (1908). Pharm. Rev. (New York), is. 61 (1895). 117 Semmler & Mayer, Ber. 44,, 2009 (1911). 118 Barbier, Bull. Soc. Chim. (3), 25, 691 (1901). 119 Ber. 3k, 3126 (1901). 120 Rec. trav. chim. 26, 157 (1907) ; Schimmel & Co. Semi-Ann. Rep. 1906, I, 109. 121 Chem. Zcntr. 1901, I, 1006. 123 Rec. trav. Chim. 26, 157 (1907) ; Schimmel & Co. Semi^Ann. Rep. 1906, I, 109. 123 J. prakt. Chem. (2) 81,, 37 (1911). ACYCLIC UNSATURATED HYDROCARBONS 183 Ocimene derives its name from its presence in the essential oil of Gcimum basilicum. Allo-ocimene was thought to be a geometrical isomer of ocimene, being obtained from this hydrocarbon by heating. Enklaar 12 * later studied the ozonides and the resulting decomposition products of these hydrocarbons and concludes that allo-ocimene is CH 3 >C = CH.CH = CH.C = CH.CH 3 CH 3 CH 3 This structure having all three ethylene bonds in conjugated positions, as in n . hexatriene, accounts for the high refractivity of this hydro- carbon. Both ocimene and myrcene yield alcohols, ocimenol and myrtenol, on treating with acetic acid and a trace of sulfuric acid, according to Bertram and Walbaum. Barbier 125 believes myrcenol to be different from linalool, and Enklaar noted the following constants: Boiling- point 99 (10 mm.), d 150 0.9032, nl 5 1.4806, phenylurethane melting- point 68. Ocimenol gives a phenylurethane melting at 72. Enklaar is of the opinion that myrcene, ocimene and allo-ocimene are not obtain- able in a state of purity, an opinion held by Wallach with regard to the terpinenes and phellandrenes. The instability of the former hy- drocarbons probably accounts for the fact that the physical constants of the myrcene investigated by Lebedew and Mereshkowski 126 was found, after "repeated purification," to be quite different from the con- stants observed by others. 127 Other Derivatives of 2 . 6-Dimethyloctane. The Citral Group: Several well-known alcohols and aldehydes be- long to this group. Their occurrence in essential oils is very wide and includes a very large number of plant species. Many of the most valu- able essential oils owe their fine aroma chiefly to substances of this group, for example, the essential oils of the rose, Rosa damascena and Rosa centijolia, lavender and orange blossoms. Some of the cheaper oils such as lemon grass, citronella and palmrosa oils are used as raw . trav. cJiim. 36, 215 (1916). 125 Bull. Soc. chim. (3) 25, 687 (1901). 128 J. Russ. Phys.-Chem. Soc. !, 1249 (1913). 127 The polymerizing action of metallic sodium on conjugated dienes is now well known ; such hydrocarbons give a brown resinous deposit after repeated distillation over sodium. 184 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS materials for the isolation of certain constituents such as citral and geraniol, which are further utilized, as in the manufacture of ionone from citral. The chemical behavior of these alcohols and aldehydes has been well established but in most cases it has been impossible defi- nitely to distinguish between the groups CH 3 CH 3 C = CHR and C.CH 2 R. CH 3 CH 2 Instead of outlining the historical development of the subject, the gen- eral relationships of the substances in this group will be indicated, fol- lowed by a description of the individual substances and some of their more important reactions. The chemical behavior and constitution of substances in the citral series is intimately associated with methylheptenone, 128 or, as Tiemann and Semmler 129 showed it to be, 2-methylheptene-(2)-one-(6), (CH 3 ) 2 = CH.CH 2 CH 2 COCH 3 . A little later, Verley 13 confirmed this struc- ture by synthesis. Oxidation, first by Wagner's method, using cold dilute permanganate, followed by chromic acid, yields acetone, and levulinic acid. CH 3 : >C = CH.CH 2 CH 2 COCH 3 > (CH,),CO + CH COCH 3 CH 3 : | : CH 2 C0 2 H On boiling an aqueous solution of potassium carbonate with citral methylheptenone and acetaldehyde are formed, and on oxidizing with chromic acid methylheptenone is also produced. The empirical for- mula of citral is C 10 H 16 and its chemical behavior and physical prop- erties indicate that it is an aldehyde containing two double bonds. If methylheptenone condensed with acetaldehyde, splitting off a molecule of water as in the condensation of acetaldehyde to croton aldehyde, or acetone to mesityl oxide, CH 3 CHO + CH 3 CHO > CH 3 CH = CH . CHO 128 Methylheptenone is usually associated with citral, and is a constituent of lemon grass, lemon, palmarosa and linaloe oils. It is best prepared by boiling a 10% solution of potassium carbonate with citral, Verley, Bull. S&c. chim. (3) 17, 176 (1897). Its boiling-point is 173-174 ; density 20 0.8602. Hydrogen in the presence of nickel at 180-190 saturates only the double bond ; sodium and alcohol reduces the ketone group forming methyl heptenol. It reacts normally with alkyl magnesium halides. 129 Ber. 28, 2115, 2126 (1895). w Bull. Soc. chim. 17, 192 (1897). ACYCLIC UNSATURATED HYDROCARBONS 185 CH 3 CH 3 > CO + H 2 CH . COCH 3 > > C = CH . COCH 3 CH 3 CH 3 the result would be citral. Such a reaction would be the reverse of the hydrolytic reaction brought about by aqueous .potassium carbonate. CH 3 CH 3 >C = CH.CH 2 CH 2 C = CH.CHO -> >C = CH.CH 2 CH 2 C = CH 3 | CH 3 | CH 3 CH 3 citral + CH 3 CHO That this is the structure of citral is indicated by the synthesis of geranic acid from methylheptenone 131 and the conversion of geranic acid to citral by heating its calcium salt with calcium formate. 132 Methylheptenone and iodoacetic ester condense in the presence of zinc to give the hydroxy acid and heating this with acetic anhydride yields geranic acid. CH 3 >C = CH.CH,CH 2 C = O + CH 2 I.C0 2 R > CH 3 | CH 3 CH 3 OZnl -> >C = CH.CH 2 CH 2 C< > hydroxy acid CH 3 I CH 2 C0 2 H CH 3 CH 3 > > C = CH . CH 2 CH 2 C = CH . CO.H > CH 3 CH 3 CH 3 >C = CH.CH 2 CH 2 C = CH.CHO CH 3 CH 3 citral Tiemann 133 discovered that purified natural citral yields mainly a semicarbazone melting at 164, and from the mother liquors of these crystals a second semicarbazone melting at 171 was isolated; mixtures of the two melt as low as 130. The aldehyde yielding the low melting 131 Barbier & Bouveault, Compt rend. 122, 393 (1896). 132 Tiemann, Ber. SI, 827 (1898). 134 Ber. SI, 3331 (1898) ; S2, 115 (1899) ; 53, 877 (1900). 186 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS semicarbazone was designated "citral a" and the other "citral b." Cit- ral a condenses more readily with cyanacetic acid, forming a citry- lidene cyanacetic acid melting at 122 and the corresponding deriva- tive of citral b melts at 94-95. Tiemann considered these isomeric crystalline derivatives as geometrical isomers and Zeitschel 134 states that citral a and citral b probably correspond to the geometrically isomeric alcohols geraniol and nerol. Me 2 C = CH . CH 2 CH 2 C CH 3 Me 2 C = CH . CH 2 CH 2 C CH 3 CH,OH. HC.CHO geraniol citral a Me 2 C = CH . CH 2 CH 2 C CH 3 Me 2 C = CH . CH 2 CH 2 C CH 3 II -> II HO.H 2 C C H. OHC C H nerol citral b Citral a and citral b have practically the same chemical proper- ties 135 and their physical properties differ only very slightly. As a rule, the boiling-points of such geometrical isomers differ only very slightly, for example, the two (3-butylenes, and dibromobutylenes HC CH 3 boiling-point HC CH 3 boiling-point HC CH 3 -f 1 to 1.5 CH 3 C H + 2 to 2.7 Br C CH 3 boiling-point Br C CH 3 boiling-point II II Br C CH 3 146-146.5 CH 3 C Br 149-150 That small differences in structure may greatly affect the melting point has previously been pointed out, and the different melting points of certain derivatives of these isomeric citrals is a case in point. Con- version of citral a to citral b and vice versa takes place readily, and, according to Bouveault, 136 alkalies convert a to b. Ordinary natural citral gives nearly pure condensation products of citral a. Further confirmation of the above relationship of geraniol and nerol is found in the behavior of these two alcohols on oxidation, first by dilute permanganate thus oxidizing the double bonds to glycols, and followed by oxidation with chromic acid. Both alcohols yield the same oxidation products and in the same proportions, i. e., acetone, levulinic 134 Ber. 39, 1780 (1906). 185 Tiemann, Ber. S3, 877 (1900). w Bull. Soc. chim. (B), 21, 423. ACYCLIC UNSATURATED HYDROCARBONS 187 Boiling-point Specific gravity , Refractive index Diphenylurethane, M. P. Tetrabromide, M. P acid and oxalic acid. 137 As in the case of the two citrals, geraniol and nerol have nearly identical physical properties but the melting-points of some of their condensation products differ markedly. Geraniol 138 Nerol 13 230 226-227 .. HOMirUOmm.) 111 (9mm.) 0.8812 to 0.883" 0.8813 150 1.4766 - 1.4786 1.468 82.5 52-53 70-71 118-119 Separation of geraniol and nerol is best carried out by means of anhy- drous calcium chloride which forms a crystalline product with geraniol but not with nerol. According to a recent paper by Verley, 140 citral a is mainly the A 1 isomer. When it is boiled with one per cent aqueous caustic soda 2-methyl-A 1 -heptenone is produced, which when oxidized first by per- manganate and then by chromic acid gives only traces of acetone, OH CH. CH,OH. \ / C-CH,CH 9 CH 9 CO.CH, \ / -CH 9 CH 9 CH,COCH, CH 3 CH 3 This methylheptenone is rapidly converted into the isomeric, ordinary 2-methyl-A 2 -heptenone, by warming with dilute sulfuric acid. Verley therefore favors the corresponding A 1 formula for geraniol and points out that this structure better explains the conversion of geraniol to dipentene. CH, HC. CH CH 2 OH Clf H H 5 H CM, CH, '"Blumann & Zeitschel. Ber. U, 2590 (1911). 138 Bertram & Gildemeister, J. prakt. Chem. (2) 56, 508 (1897) ; Erdmann, J. prakt. Chem. (2) 56, 3 (1897) ; Stephan, ibid., 58, 110 (1898). 139 Soden & Treff, Chem. Ztg. 27, 897 (1903). Bull. Soc. chim. (4), 25, 68 (1919). 188 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Two ketones occurring in artemisia oil appear to have a carbon structure different from the citral group but these two isomeric ketones are supposed to bear the same relation to each other as the A 1 and A 2 isomers discussed above. 141 As is indicated in the foregoing discussion of the constitution of citral, the constitution of geraniol and nerol are shown by their rela- tions to citral. Citral is formed from geraniol by oxidation with chro- mic acid, 142 and reduction of citral yields geraniol. Apparently only the groups CH 2 OH and CHO are affected. On more energetic oxidation the citral first formed is oxidized as indicated above, to methylheptenone, acetone, levulinic acid, etc. These relations are, therefore, expressed by the following constitutions of geraniol, CH 3 C = CH.CH 2 CH 2 C = CH.CH 2 OH CH ; CH OH, or, C . CH 2 CH 2 CH 2 C = CH . CH 2 OH CH 2 ' CH 3 When geraniol is heated with water in an autoclave to 200 linalool is formed, 143 and the conversion of linalool to geraniol, or geranyl ace- tate is brought about by heating with acetic anhydride. 144 By warm- ing a solution of linalool in toluene with hydrochloric acid, geranyl chloride is formed. 145 These changes are readily understood from the structure of linalool deduced by Tiemann and Semmler 146 by a study of the oxidation products of linalool. Oxidizing first with dilute per- manganate, followed by chromic acid gave acetone and levulinic acid (equivalent to methylheptenone) and oxalic acid. CH 3 : OH . CH 3 >C = CH.CH 2 CH 2 C<. > >CO + CH 3 : | .CH = CH 2 CH 3 CH 3 . 14l Asahina & Takagi, J. Chem. 8oc. Abs. 1921, I. 9. " 2 Semmler, Ber. 23, 2966 (1890). 143 Schimmel & Co.'s Ber. 1898, I, 25. 144 Bouchardat, Compt. rend. 116, 1253 (1893). Terpineol is also formed. 145 Tiemann, Ber. 31, 832 (1898) ; Dupont & Labaune, Roure-B&rtrand, Fils. Bull. 1909, II. 27; Forster & Cardwell, J. Chem. Soc. 103, 1338 (1913). 148 Ber. 28, 2126 (1895). ACYCLIC UNSATURATED HYDROCARBONS 189 C0 2 H.CH 2 CH 2 CO C0 2 H CH 3 C0 2 H It was also pointed out that the chemical and physical properties of linalool agree with the structure of a tertiary alcohol, and that when oxidized by chromic acid direct, to citral, isomerization by the acid to geraniol, or the glycol, first takes place, by acid CrO 3 linool - geraniol - > citral OH OH CH 3 I | > C = CH . CH 2 CH 2 C CH = CH, > RC CH 2 CH 2 OH CH 3 | | CH 3 CH 3 linalool R _ c = CH.CH 2 OH - > RC = CH.CHO CH 3 geraniol CH 3 citral Linalool has recently been synthesized by Ruzicka 14T who employed a reaction discovered by Nef, 148 i. e., condensation of acetylene with ketones by means of metallic sodium. The first condensation product gives good yields of linalool on reducing by moist ether at low tem- peratures. \ CH 3 >C = CHCH 2 CH,C = CO + HC = CH CH 3 CH 3 CH 3 OH > > C = CH . CH 2 CH 2 C < CH 3 | C = CH > linalool CH 3 The constitution of citronellol and the corresponding aldehyde, cit- ronellal, is shown by the following reactions; citral may be oxidized to the corresponding acid geranic acid and on reducing this by sodium and amyl alcohol citronellic acid is obtained; also the aldehyde citronellal may be converted to its oxime and this, by loss of H 2 0, to the nitril, which yields citronellic acid. Therefore, citronellol is dihydrogeraniol, Helv. CMm. Acta. 2, 182 (1919). 148 Ann. 308, 264 (1898). 190 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS and citronellal is dihydrocitral. 149 As to the location of the remaining double bond in citronellic acid, citronellol and its aldehyde, the evi- dence was at first confusing, but the facts are best explained by the reduction of the RC = CH . COOH group, which is in harmony with CH 3 the well known highly reactive character of the >C = C CO group. The aldehyde citronelall can be reduced mainly to the corre- sponding alcohol, citronellol, by sodium amalgam in acetic acid. 150 As with the other substances of this group, some doubt remains as to whether the double bond in citronellol and its aldehyde is in the po- sition, shown but according to Harries 151 both isomers are present in natural citronellol, i. e. ; CH 3 C = CH.CH 2 CH 2 CH.CH.CH 2 OH CH 3 CH 3 and CH 3 C CH 2 CH 2 CH 2 CH . CH 2 CH 2 OH. CH 2 CH 3 Citronellol and rhodinol appear to be isomers differing only in the CH 2 position of the double bond, C.CH 2 R, or (CH 3 ) 2 C = CH.R. CH 3 The question of the existence of rhodinol has been the subject of con- siderable controversy, the difficulty of deciding such questions being that, as in all such cases, the chemical behavior and physical prop- erties are so nearly identical, and conversion of the one isomer into the other takes place with great ease. In discussing the simple aliphatic olefines, such as hexene(l), it was pointed out that double bonds of the type RCH 2 CH CH 2 frequently shift their position very readily, and the work of Verley, noted above, shows that warming with dilute sulfuric acid changes the group 149 Tiemann, Ber. SI, 2899 (1898) ; Bouveault, Comvt rend. 138, 1699 (1904). 160 Dodge, Am. Chem. J. 11, 463 (1889). 161 Ber. 41, 287 (1908). ACYCLIC UNSATURATED HYDROCARBONS 191 CH 2 C.CH 2 R to the isomer (CH 3 ) 2 C = CH.R. CH 3 German chemists continued to regard rhodinol as a mixture of cit- ronellol and geraniol but Harries 152 and his assistants have shown that natural citronellbl and the aldehyde citronellal consists of a mixture of the two isomers, confirming the contention of Barbier, Bouveault 153 and Locquin 154 as to the existence of rhodinol. According to Harries ordinary citronellal, derived from oil of citronella, contains approxi- mately 60 per cent "rhodinal," the aldehyde corresponding to rhodinol. Methods of oxidation have not clearly shown the structure of these isomers but rhodinol appears to be the more stable of the two alcohols. Both alcohols, in the form of their acetates, combine with hydrogen bromide, and when this is removed by heating with sodium acetate, rhodinol is the product. Also, according to Barbier and Locquin, 155 citronellal may be converted into its oxime, which on dehydrating by acetic anhydride yields the nitrile, but the oxime of the aldehyde, made by the oxidation of ^-rhodinol or d-rhodinol, does not yield the nitrile but acetylmenthone oxime. Citronellol may be converted into rhodinol by the addition of water, brought about by treating with 30 per cent sulfuric acid. CH 2 C . CH 2 CH 2 CH . CH 2 . CH 2 OH / citronellol CH 3 I CH 3 CH 3 C . CH 2 CH 2 CH 2 CH . CH 2 CH 2 OH CH 3 OH | CH 3 CH 3 C = CH . CH 2 CH 2 CH . CH 2 . CH 2 OH / rhodinol (according to CH 3 CH 3 Barbier). 153 Ber. 41, 2187 (1908) ; Ann. 410, 1 (1915). 163 Bull. Soc. cMm. (3), 23, 458, -465 (1900). Compt. rend. 157, 1114 (1913). 158 LOG. ait. 192 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Prins 156 endeavored to separate natural citronellal into its two isomers, by repeated fractional distillation 157 followed by repeated fractional crystallization of the semicarbazone and semioxamazone, but without success. Prins also studied the conversion of citronellal to isopulegol, by treating with 85 per cent formic acid and by 80 per cent phosphoric acid but was unable to detect the formation of any substance, which could be derived directly from rhodinal. The formation of isopulegol acetate by heating ordinary citronellal with acetic anhydride is prac- tically quantitative, 158 which, in the light of Harries' work, indicates that under these conditions rhodinal must be converted into its isomer, true citronellal. Barbier and Bouveault believed that they had ob- tained small yields of menthone from rhodinal, but Tiemann and Schmidt 159 were unable to confirm this. The ready conversion of cit- ronellal to isopulegol, however, favors the structure purposed by Bar- bier for this aldehyde, According to Semmler 16 aldehydes of the types R 2 CH . CHO and RCH 2 CHO are converted to enolic forms by acetic anhydride and that in the case of citronellal this change precedes ring formation. The above review illustrates how difficult it is to distinguish be- tween isomers of this kind. Geraniol. The importance of the alcohols and aldehydes of this group to the essential oil industry warrants further description of them and their chemical behavior. Geraniol is present to a large extent in palmarosa oil, ginger grass, citronella and oil of sweet geranium, partly 156 Chem. Weekbl. IJf, 627, 692 (1917). The maximum difference in boiling-points observed by Prins was 198-200 for the low-boiling fraction and 203-204 for the higher boiling portion. 167 Schimmel & Co.'s Rep. 1910, I, 155. Schimmel & Co.'s Rep. 1896, 34; Semmler,. Ber. 42, 584, 963, 1161, 2014 (1909). 160 Ber. 30, 38 (1897). 160 Ber. ^2, 584, 963, 1161, 2014 (1909) ; kk, 991 (1911). ACYCLIC UNSATURATED HYDROCARBONS 193 in the free state and partly as the acetate. In oil of geranium small proportions of the geraniol ester of tiglic acid are present. 161 Com- mercially geraniol is isolated from either palmarosa or citronella oil by means of finely ground anhydrous calcium chloride, the mixture being chilled to about 5 for several hours. Other oils are removed by means of petroleum ether and the crystalline calcium chloride com- pound decomposed by water. Small percentages of geraniol cannot be separated from essential oils in this manner. It is readily identified by its diphenylurethane, 162 melting-point 82, or its naphthylurethane, melting-point 47-48. Geranyl chloride is of particular . interest as filling a niche in the chemistry of the non-benzenoid hydrocarbons and contributing to the generally similar chemical behavior of this whole class of substances. Although not mentioned in Richter's "Lexikon" geranyl chloride was evidently first made by Jacobsen 163 and later by Tiemann 164 who pre- pared large quantities of it by the action of hydrogen chloride on geraniol. Dupont and Labaune 165 passed dry hydrogen chloride into a solution of geraniol or linalool in toluene at 100 and noted that both alcohols gave the same chloride, which they called linalyl chloride, and Kerschbaum 166 following Tiemann's first method, made it by treating geraniol with phosphorus trichloride. The first study of the chloride and its reactions was carried out by Forster and Cardwell, 167 who employed Darzen's method, dissolving the geraniol in pyridine and treating with thionyl chloride. The chloride was shown to be a derivative of geraniol rather than linalool by the preparation of geranyl acetone, by the action of geranyl chloride on the sodium derivative of acetoacetic ester, and hydrolysis of the geranyl acetoacetate by barium hydroxide. The constitution of geranyl acetone is shown by reference to the constitution of farnesol 168 and the work of Kerschbaum. 169 The chlorine atom in geranyl chloride is stabilized by the proximity of a double bond, (CH 3 ) C = CH . CH 2 CH 2 C = CH . CH 2 C1 but it reacts normally with sodium ethoxide to give the ethyl ether and with sodium acetoacetic ester and sodium malonic ester. From the latter substance 181 Schimmel & Co., Semi-Ann. Rep. 191S, II, 61. 1(82 Cf. Parry, "Essential Oils" Vol. II, Ed. II, 1919, 98. 183 Ann. 157. 236 (1871). Ber. 29. 921 (1896) ; SI, 832 (1898). 195 Roure-Bertrand Fits' Bull. 1909, II, 19. 166 Ber. 46, 1735 (1913). J. Chem. Soc. 103, 1338 (1913). 168 Harries & Haarman. Ber. -46, 1737 (1913), 189 Loc. cit. 194 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS geranyl acetic acid was made, C 10 H 17 . CH 2 CO 2 H. Sodium azide yields the azoimide and corresponding amine, geranyl amine. Prileshajev 170 has prepared the mono and dioxides of geraniol by direct oxidation by benzoyl peroxide. The dioxide is a mobile liquid boiling at 180-183 under 25 mm. CH 3 /\ /\ >C CH . CH 2 CH 2 C CH . CH 2 OH CH 3 | CH 3 geraniol dioxide (according to Prileshajev) Geraniol is markedly less stable than citronellol. On heating with phthalic anhydride to 200 geraniol is decomposed but citronellol forms the acid phthalic ester; concentrated formic acid also decomposes ge- raniol much more readily than citronellol. 171 Benzoyl chloride at 140- 160 also decomposes geraniol, 172 but not citronellol. Isogeraniol: Evidence of a shift in the position of one double bond in citral by the action of acetic anhydride is furnished by the isolation of an isomer of geraniol when the acetic ester of enol-citral is reduced by sodium amalgam in methyl alcohol acidified by acetic acid. 173 This alcohol, like geraniol, has a fine roselike odor and may be distinguished by means of its diphenylurethane melting at 73. Ac- cording to Semmler, the formation of isogeraniol may be represented as follows: CHjOH CH 3 170 J. Russ. Phys.-Cliem. Boc. 4%, 613 (1912) ; a trace of mineral acid hydrolyses one of the oxide groups, forming the glycol, Ci H 17 O. (OH) 3 .2H 2 O melting-point 94.5, also the anhydrous glycol CioH 17 O. (OH) 3 in two forms melting at 145 and 163. "iWalbaum & Stephen, Ber. 33, 2307 (1900). " 2 Barbier & Bouveault, Compt. rend. 122, 530 (1896). Semmler, Ber. kk, 991 (1911). ACYCLIC UNSATURATED HYDROCARBONS 195 Linalool: Linalool is isolated technically from oil of Central American linaloe wood. Its acetate is the principal constituent of oil of lavender and it is an important component of a great number of other essential oils, among which are ylang-ylang, cham- paca, rose, geranium, petit-grain, bergamot, neroli, jasmine and other oils. It is not easily isolated or purified since it yields no crystal- line addition products or derivatives from which linalool can easily be regenerated. Hydrogen chloride forms geranyl chloride, boiling-point 82-86 at 6 mm. 174 Mono linalyl phthalate may be prepared by forming the sodium compound of linalool, in ether and allowing this to stand several days with phthalic anhydride. 175 Linalool, being a ter- tiary alcohol, is partially decomposed when acetylation by acetic an- hydride is attempted, dipentene, terpinene, ct-terpinyl acetate and neryl acetate being formed. 176 Continued heating with acetic anhydride de- composes terpineol, also a tertiary alcohol, and maximum yields of ter- pinyl acetate, about 85 per cent, are obtained in 45 minutes. 177 When diluted with xylene, as proposed by Baulez, the maximum esterifica- tion, about 63 per cent, is obtained in 7 hours. 178 The conversion of linalool to terpinene and dipentene by heating with acids is believed to involve isomerization to geraniol. Anhydrous oxalic acid is much more energetic in its .action and yields a bicyclic diterpene, C 20 H 32 , isocamphorene. 179 Oxidation of linalool by benzoyl yields a mono oxide or dioxide depending upon the proportions of benzoyl peroxide employed 18 and m Forster & Cardwell, loc. cit. "'Charabot, Ann. chim. phys. (7), 21, 232 (1901). _ Ber. S9 t 1780 (1906). 178 Schimmei & Co''s Ber' im' I,' 127. 1 Semmler, Ber. 47, 2068 (1914).. j^s^^&^s^S&S^S^K^^ 196 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS H. Erdmann 181 employed linalyl acetate in studying the addition of sulfur to unsaturated substances to form what he terms thioozonides. These thioozonides decompose on heating, evolving hydrogen sulfide. A tertiary alcohol resembling linalool and containing two more hydrogen atoms (one less double bond) has been prepared by two well- known reactions, which have previously been discussed in connection with the synthesis of hydrocarbons, and the action of sulfuric acid upon olefines. Thus dihydromyrcene on treating with 85 per cent sulfuric acid 182 yields dihydrolinalool; the same tertiary alcohol is obtained by the action of ethyl magnesium bromide on methyl hep- tenone. (1) myrcene ^ CH 3 + H 2 ' > C = CH . CH 2 CH C = CH . CH 3 CH 3 "| CH 3 (2) CH 3 >C = CH.CH 2 CH 2 G = + C 2 H 5 MgBr H, A: > > C = CH . CH 2 CH 2 C < CH 3 "| OH CH 3 dihydrolinalool. Contrary to opinions previously held, Dupont and Labaune 183 find that the double bonds in linalool and geraniol react with sodium sul- fite. These alcohols are completely dissolved by continued shaking with aqueous sodium sulfite, the compounds C 10 H 1S . 2NaHS0 3 hav- ing been isolated. It has long been known that the ethylene bond in citral, in the group C = CH.CHO reacts readily with sodium sul- CH 8 fite, but this is the first instance of unsaturated alcohols reacting in this manner. Citronellol and Rhodinol: From the foregoing discussion of these adjacent carbon atoms. The oxide of linalyl acetate, made by Prileshajev's method, reacts with water readily to give the glycol Ci H 17 (OH)2.O 2 C2H3 which on saponifica- tion yields Ci H 17 (OH) 3 melting at 54-55. 181 Ann. 362, 137 (1908) 182 Myrcene is converted to cyclo dihydromyrcene by the action of sulfuric acid in acetic acid. 188 Rmire-Bertrand FiU> Bull. 1912, (3) 6 & 7; J. Chem. Soc. 1913, I, 746. ACYCLIC UNSATURATED HYDROCARBONS 197 two substances it is evident that these two alcohols occur together, and while recognizing the probable existence of rhodinol, the name citronellol will be retained and, following common usage, will be em- ployed for the alcohol C 10 H 20 0, containing one double bond, and hav- ing the following physical properties: PHYSICAL PROPERTIES. Observer Wallach 184 Tiemann ** Tiemann 18a Schimmel & Co. 187 Schimmel & Co. 188 Schimmel & Co. Boiling-Point 114 -115(12-13mm.) 117-118 (17mm.) 113-114 (15mm.) 225-226 109 225-226 c (7mm.) Density n D Method of Isolation 99 0.856^- 0.8565-1T 0.8612^5! 1.4561 1.4566 1.4578 Destroying geraniol at 250 reduction of citronellal by PCI, method 0.862 1.45611 Wallach's method j 0.8604 ( 0.8629 1.4565 ) 1.4579 f From Java citronella f 0.862 1 0.869 1.459 I 1.463 | Commercial preparation from oil of geranium Citronellol is considerably more stable than geraniol or linalool, to the action of alkalies, 10 per cent sulfuric acid, heating with formic acid or phthalic anhydride, phosphorus trichloride in the cold, heating with water as in Wallach's method of purifying citronellol. The for- mation of a cyclic hydrocarbon by loss of water from citronellol has not been observed. Citral: The constitution of citral and the nature of citral a and citral b have been discussed in the preceding general discussion. The following physical properties of citral have been noted: Observer Tiemann & Semmler" 8 Schimmel & Co. 190 Schimmel & Co. Boiling-Point 117-119(20mm.) Density n D Source 0.8972^1 * 1.4931 0.893*5! 1.4901 lemon- 92- 93 (5mm.) 0.8926 ~L 1.4885 lemon 1B4 Nachr. K. Ges. Wiss. Gottingen, 1896, 56. 185 Ber. 29, 906 (1896). 188 Ibid, 923. 187 Schimmel & Co.'s Ber. 1898, 62. 188 Ibid, 1902, I, 14. 189 Ber. 26, 2709 (1893). 190 Schimmel & Co. Rep. 1899, I, 72. 198 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS In addition to the chemical reactions of citral noted above, the fol- lowing may be noted. Potassium acid sulfate or moderately diluted sulfuric acid react on citral very energetically with ring closing, loss of water and the formation of p.cymene. The behavior of citral to sodium sulfite solutions has been the sub- ject of considerable investigation. In the presence of a very slight ex- cess of free sulfurous acid, in the cold, the normal, aldehyde addition OH product C 9 H 15 CH< is formed, separating as very fine, spar- OS0 2 Na ingly soluble, crystalline plates; regeneration of citral from this de- rivative is not quantitative. If this crystalline product is allowed to stand, and gently warmed, with an excess of bisulfite, it goes into solution as a labil dihydrodisulfonic acid derivative, from which citral can be regenerated by the action of caustic soda, but not by alkali carbonates. If the bisulfite solution of citral is strongly heated, the stabil dihydrodisulfonic acid derivative is formed and it is impossible to regenerate citral from this stabil combination. If the labil dihydro- disulfonic acid salt is treated with another molecular portion of citral, this goes into solution as a labil m C 9 H 15 CHO. (NaHS0 3 ) 2 + 2NaOH This reaction usually gives so much difficulty that Tiemann's 192 direc- tions may be given here. A solution of 350 g. sodium sulfite in one liter of water is made slightly alkaline to phenolphthalein, treated with 100 g. citral and gently shaken, keeping just slightly alkaline by the con- tinual addition of a calculated quantity of 20 per cent sulfuric acid (or acetic acid) . The solution should always be distinctly red by phenol- phthalein, since in slightly acid solution the sparingly soluble crystal- line compound will separate. The various addition products formed by sodium bisulfite and citral may be summarized thus, 193 where X represents the S0 3 Na group. Evidently, in the stable derivatives, car- bon and sulfur are directly combined as C S0 3 Na or true sulfonic acid salts. 191 Cf. Gildemeister, "Die Aetherischen Oele," Vol. I. Ed. II. 429 (1910). Ber. 31, 3317 (1898). 188 G. Romeo, Qazz. chim. Ital. 8, (1), 45 (1918). ACYCLIC UNSATURATED HYDROCARBONS 199 (1) normal aldehyde addition product. CH 3 OH > C = CH . CH 2 CH 2 C = CH . CH < labil. CH 3 OS0 2 Na CH 3 (2) stable dihydrodisulfonate, formed in warm acid solutions, prob- ably of the type C S0 3 Na. CH 3 (HX) (HX) >C CH . CH 2 CH 2 C CH . CHO stable. CH 3 I CH 3 (3) labil dihydrodisulfonate, formed in slightly alkalins solutions, probably of the type C OS0 2 Na. CH 3 (HX) (HX) > C CH . CH 2 CH 2 C CH . CHO labil. CH 3 I CH 3 (4) Citral mono sodium hydrosulfonate, formed by citral + labil citral dihydrodisulfonate C 9 H 16 CHO.S0 3 Na (constitution not known) . labil. (5) Citral trihydrosulfonate. CH 3 (HX) (HX) OH > C CH . CH 2 CH 2 C CH . CH < labil. CH 3 OS0 2 Na CH 3 (6) A stable form of (5). stable. The hydrogenation of citral is of considerable industrial interest on account of the availability of citral in oil of lemon grass and the possibility of its conversion into the more valuable rose like citronellol, or the hydrogenation of one double bond only, yielding citronellal which, as noted above, is quantitatively convertible into isopulegol and the latter substance being convertible by hydrogenation into the well-known article of commerce menthol, now derived entirely from oil of peppermint. Skita 194 found that on hydrogenating over nickel at 190-200, chiefly a decane was formed, and at a lower temperature, 140, and under pressure Ipatiev showed that a decanol was the chief product. Law 195 attempted to reduce citral by electrolytic reduction in m Chem. Zentr. 1911, I, 1209. 188 J. Chem. Soc. 101, 1024 (1912). 200 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS alcohol acidified by sulfuric acid, but he was evidently not familiar with the properties of citral and related substances and it is impos- sible to tell from his article just what the result was. 196 According to Paal m hydrogenation by colloidal paladium or platinum, or by cat- alytic masses consisting of a supporting material on which small pro- portions of one of these metals are deposited, converts citral first to inactive citronellal and then to the saturated aldehyde tetrahydro- citral ; geraniol is reduced to inactive citronellol and the saturated alco- hol tetrahydrogeraniol. (This alcohol has recently been further studied by Ishizaka, Ber. 41, 2483 (1914) , who also prepared it by PaaFs meth- od.) Skita states that citral yields both citronellal and citronellol to- gether with a dimolecular aldehyde C 20 H 34 2 when using colloidal palladium as a catalyst. Condensation of aliphatic aldehydes with (3-naphthylamine and pyruvic acid usually yields well crystalline products suitable for the purpose of identification. Citral condenses with these two substances to form citryl-p-naphthocinchoninic acid, melting at 199-200. Cit- ral oxime and the phenylhydrazone are liquid at ordinary tempera- tures. Cyanacetic and malonic acid condense readily yielding well crystalline products. CN CN C 9 H 15 CHO + H 2 C < -- > C 9 H 15 CH = C < C0 2 H C0 2 H C0 2 H. C0 2 H C 9 H 15 CHO + H 2 C< -- > C 9 H 15 CH = C< C0 2 H C0 2 H Citral also condenses readily with acetone in the presence of alka- lies, and this led Tiemann 198 to the discovery of the ionones. The for- mation of psewdo-ionone is an example of the well-known type of con- densation illustrated by the formation of croton aldehyde, and mesityl oxide. According to Tiemann's patent specifications the condensation is effected by means of barium hydroxide, but other condensing agents give better results, for example 5 per cent of sodium ethoxide in abso- lute alcohol. 199 The resulting mixture is distilled and the fraction boiling at 138-155 at 12 mm. is purified from unchanged citral and condensation products formed from acetone alone, by distilling with 19 * I ^ t ^ ie P I R ion of tfl e writer, Law's experiments are of considerable interest and would be_ worth repeating with the cooperation of a skilled organic chemist. ohem - 4t - 8 - 1019 "Slack, Pref. & Eaa. Oil Record. 7, 389 (1916). ACYCLIC UNSATURATED HYDROCARBONS 201 steam, these impurities being easily volatile. By a second vacuum dis- tillation of this product a very pure pseudo-ionone boiling at 143-145 is obtained. CH 2 CH 3 CH 3 > C = CH . CH 2 CH 2 C = CH . CHO + H 3 C CO CH, CH 3 >C = CH.CH 2 CH 2 C = CH.CH = CH.COCH 3 AH. pseudo-ionone. The physical properties of pseudo-ionone are as follows; specific gravity at 20 = 0.8980, refractive index-p = 1.53346, boiling-point at 12 mm. 143-145 When pseudo-ionone is heated with dilute sulfuric acid, about 1 per cent, for several hours, ring closing results, probably through the intermediate addition of water and subsequent decomposition, giving two isomeric ionones, designated as. a and (3. CH- ,H CH, CH oc-ionone VH=CHCO CH, r CH. CH, /3-ionone The odors of the two isomers are noticeably different, a-ionone having a sweeter odor more nearly resembling orris root and p-ionone, in very dilute solutions, about 1 : 10,000, resemblirig more closely the fresh wood violet. Commercial ionone is usually a mixture of the two isomers containing mostly a-ionone. The conditions under which pseudo-ionone is condensed affect the relative proportions of a and p-isomers, more concentrated sulfuric acid at low temperatures in- creasing the proportion of p-ionone while phosphoric, hydrobromic and hydrochloric acid yield chiefly a-ionone. Many methods have been proposed to separate the two isomers, of which two only will be men- tioned. Pure a-ionone was isolated by Tiemann by making the oxime 202 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS of a commercial ionone containing mostly a-ionone, recrystallizing the oxime from petroleum ether and regenerating the ketone by means of dilute sulfuric acid. One method of separation is based upon the in- solubility of the sodium hydrosulfonate of a-ionone in aqueous sodium chloride solutions. 200 If sodium chloride is added to a hot solution of the sodium hydrosulfonates, the a-ionone derivative separates, crystal- lizing in very small plates; the (3-ionone hydrosulfonate remains in so- lution. Gildemeister 201 notes the following physical properties for commercial ionone; boiling-point 104-109 (4 to 5 mm.), d 150 0.9350 20 to 0.9403, n jj- 1.5033 to 1.5051. Chuit 202 gives the following for the two isomers. a-Ionone P-Ionone Boiling-point 127.6(12mm.) 134.6 (12mm.) Density, 15 0.9338 0.9488 Refractive index 1.50001 1.52008 p-Bromophenylhydrazone, M. P 142-143 116-118 Semicarbazone 107M08 148-149 The ionones may be hydrogenated to the ketone tetrahydroionone by means of hydrogen and colloidal palladium 203 or the ketone group may be converted to >CH 2 without affecting the double bonds. 204 Irone: On account of its similarity to the ionones, this ketone, an isomer of the synthetic violet ketones, may be mentioned here. It was isolated from the volatile oil of orris root and studied by Tiemann. It has been made by the condensation of A 4 cyclocitral and acetone, 205 and its close similarity to the ionones is shown by the following struc- ture. H.CH, H.CH=CHCO CH 3 CH 3 N CH 3 The physical properties of irone are, boiling-point 144 (16 mm), 20 d 150 0.9391, n -=y 1.5017. Its characteristic derivatives are the oxime, 200 Chuit, Rev. Gen. Chim. 6, 432 (1903). n "Die Aetherischen Oele", Vol. I, 485. Ed. II (1910). 202 Loc. cit. z03 Skita, Ber. 45, 3312 (1912). 04 Kishner, J. Russ. Phys.-Chem. Soc. tf, 1398 (1912). 2M Merhng & Welde, Ann. 366, 119 (1909). ACYCLIC UNSATURATED HYDROCARBONS 203 melting point 121.5, p-bromophenylhydrazone melting at 174-175 and thiosemicarbazone melting at 181. Irone is not made syntheti- cally on an industrial scale, nor isolated as such from the volatile oil of violet root, or orris. In view of the commercial value of the ionones Merling and Welde 206 undertook a study of similarly constituted unsaturated ketones. Any slight change in the constitution of these ke- tones causes considerable difference in odor. While the group - CH = CH CO CH 3 is essential to odors of this kind, as is shown by the fact that on hydrogenating the double bonds, the fra- grance of the ionones disappears, the particular quality of the odor is influenced greatly by the relative positions of the other ethylene bond and the methyl groups. Condensation products with acetone were pre- pared from the following three aldehydes, isomeric with cyclocitral. HO CHO The product derived from I was almost odorless but the products from II and III had faint violet like odors. The most intense odor is obtained when the aldehyde group is situated between the methyl and dimethyl groups. The perfume character of such acetone conden- sation products disappears when the aldehyde group does not adjoin CH 3 a methyl group. Nevertheless the grouping C CH C< does CH 3 CHO CH 3 not yield a perfume when condensed with acetone, as is shown by the condensation product obtained from (3-isopropylbutaldehyde and ace- tone, but when these groups are present in the cyclogeraniolene ring, a perfume results. The importance of the tertiary butyl group C(CH 3 ) 3 , to the odor of musk, has been brought out by the work 206 Ann. 366, 119 (1909). 204 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS of Bauer on artificial musk. 207 Austerweil 208 has shown that the group = CH.CRR 1 appears to be necessary to produce geraniol like odors. The condensation of citral with ethyl acetoacetate has been studied by Knoevenagel 209 who isolated five isomeric ethylcitrylidene acetol acetates. Condensation is brought about by adding a very small quan- tity of piperidine to a mixture of ethyl acetoacetate and citral at 15 and allowing to stand two days. The structure of these con- densation products is not yet definitely known. Citral reacts normally with methyl or ethyl-magnesium bromide to give secondary alcohols of rose like odor. 210 Sesquicitronellene, C 15 H 24 . This so-called aliphatic sesquiterpene was discovered in Java citronella oil by Semmler and Spornitz. 211 It has four double bonds, three of which are probably conjugated. Mol. Ref. 74.53 Mol. Ref. calc. for C 15 H 24 /= 4 69.6 E M 4.9 Sodium and alcohol readily reduce it to C 15 H 26 (evidence of at least one pair of conjugated double bonds) and hydrogen in the presence of platinum black yields the saturated acyclic hydrocarbon C 15 H 32 . As is frequently observed among the sesquiterpenes ring closing is easily effected, being brought about in this case by concentrated formic acid. Sodium and alcohol do not reduce the cyclic hydrocarbon showing that ring formation has occurred through one of the conjugated double bonds. The original Sesquicitronellene is readily oxidized and poly- merized. Its physical properties are, boiling-point 138-140 (9 mm.) , d 20 0.8489, n D 1.53252. Spinacene. C 30 H 60 . This very remarkable unsaturated hydrocar- bon has recently been described by Chapman 212 and by Tsujimoto. 218 It has been found in the livers of several species of the Spinacidae, a family of the Selachoidei, or sharks, and Chapman has therefore named it spinacene. In the fresh liver oils of certain species this hydrocarbon " Ber. 4, 2832 (1891) ; 32, 3647 (1899). 108 Compt. rend. 151, 440 (1910) . . , . *J. prakt. chem. (2), 97, 288 (1918). 810 Bayer & Co., Chem. Zentr. 1904, II, 624, 1269. Ber. 46, 4025 (1913). * a J. Chem. 8oc. Ill, 56 (1917) ; 113, 458 (1918), Chem. Aba, n, 1004 (1918). ACYCLIC UNSATURATED HYDROCARBONS 205 constitutes about 90 per cent of the oil. Fish liver oils previously known, such as those of the haddock, skate, hake, cod, and tunny, contain only about 2 per cent of unsaponifiable matter which appears to be cholesterol. From the standpoint of physiological chemistry, the manner of formation, secretion and physiological utilization of such an oil is of great interest, and inasmuch as the sharks are found, fos- silized, in many strata, geologically very old, the probability that shark liver oils have contributed to the formation of petroleum is at once suggested. Chemically, spinacene is of more than ordinary interest. Dry hy- drogen chloride passed into a cooled ether solution of spinacene forms the crystalline hexahydrochloride, C 30 H 50 .6HC1, and bromine in dry ether yields the crystalline dodecabromide C 30 H 50 Br 12 . Like chlorine and bromine derivatives of petroleum hydrocarbons and the terptnes, these spinacene derivatives are unstable and readily decompose on heat- ing. The hydrocarbon accordingly contains six double bonds. A moder- ately stable crystalline trinitrosochloride can be prepared by the usual methods. By catalytic hydrogenaton by means of platinum black, Chap- man obtained the saturated hydrocarbon C 30 H 62 , which is liquid at 20 and therefore is not a normal paraffine. Exaltation of the re- fractive index and partial polymerization by metallic sodium indicate that probably two pairs of double bonds are in conjugated positions. On distilling over sodium, partial decomposition also occurs, forming a hydrocarbon C 10 H 18 , which appears to be a monocyclic hydrocarbon containing one double bond, boiling at 170-175, and much resem- bling cyclodihydromyrcene in its properties. Cholesterylene: The hydrocarbons resulting from the decomposi- tion of cholesterol or cholesteryl chloride have been repeatedly in- vestigated on account of the possible connection of this hydrocarbon with the optical activity of petroleum. The properties of "choles- terylene" vary considerably according to its method of preparation. When equal parts of cholesterol and infusorial earth are rapidly heated to 280-300 a solid cholesterylene is obtained, which is capable of adding four atoms of hydrogen to form the solid cholestane. A sim- ilar mixture slowly heated for about eight hours at 300 gives an oil, probably a mixture, boiling at 257-267 at 12 mm., D = 0.9572 and [a] D + 49.12. The product obtained by rapid heating is laevo ro- tatory. 214 Cholesteryl chloride 215 yields an oil having properties prac- tically identical with those noted above. *Steinkopf, J. prakt. chem. (2) 100, 65 (1919). " Mauthuer & Suida, J. Chem. 800. Aba. 1904, I, 49 ; 1909, I, 714. 206 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS I ? 2 3 eot^oo rH rH rH , C 6 TO CM. acteristic Deri d "8 3 s W s t d n^ CO o i oSo * 1 1 <- < o IH 3 ) 3 C.CH i. (CH3) 2 C ,),CBr.CH sptylpalmii d 1 1 K^ 1 - I T3 a > s~\ rv5 c^ ^ i i A o> o o d S, ^ 1111 j^J^ - c>- 1 s c* I O QJ "d "d ^a d jQ T> -d^ c3 ft 1 ^ d d S, a a S, 1 1111 = ^3 -*J *+^ > fl) N x s 1 1 1 T^T^ ^^ H-3 C- SS 4J O g ^01 Hexene-( 2-Methyl 2-Methyl 3-Methyl o a s CO o Q Q S i aaa| SS91 1 tN o> c^ -o 1 1 1 < Q ^3 2 O a 1 G 1 jfl 3 'I "S ">> 1 a CD || s 11 1 A f| < 55 o S iQ .2-^ a^ >4 P fi s Ss of > i m** W S o co c5, !zj 5 fi 208 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS I 3 C) S IQ ^D CO CO l> *O * t*^ I 1 i-( O5 a a a a a a O 10 C^ t" to l-H Cq l-l ACYCLIC UNSATURATED HYDROCARBONS 209 1 v. Braun, Ann. 382, 22 (1911); Zelinsky and Prshevalsky, J. Russ Phys - Chem. Soc. 89, 1, 1168 (1907). 2 Brooks and Humphrey, J. Am. Chem. Soc. 40, 833 (1918). 3 Jawein, Ann. 195, 255; Ipatiev, Chem. Zentr. 1899, II, 177. 4 Umnova, J. Russ. Phys.-Chem. Soc. 42, 1530 (1911). 5 Wislicenus, Ann. 219, 313; Jawein, Ann. 195, 255 (1879). 6 Fomin and Sochanski, Ber. 46, 244 (1913). 7 Delacre, Chem. Zentr. 1906 (1), 1233. 8 Henry, Compt. rend. 144, 553 (1907). 9 Kondakow, J. prakt. Chem. (2), 62, 174 (1900). 10 Welt, Ber. 30, 1495; Przewalski, Chem. Zentr. 1909 (2), 794. 11 Sabatier and Senderens, Compt. rend. 135, 88 (1902). 12 Schorlemmer and Thorpe, Ann. 217, 150 (1902). 13 Bjelouss, Ber. 45, 625 (1912). 14 Sayzew, /. prakt. Chem. (2) 57, 38 (1898). 15 Kaschirsky, Ber. 11, 985 (1878). 16 Pawlow, Ann. 173, 194 (1874). 17 Butlerow, /. Russ. Phys.-Chem. Soc. 7, 44 (1875). 18 Senderens, Compt. rend. 144, 1110. 19 Briihl, Ann. 235, 11; Eijkwan, Chem. Zentr. 1907 (2), 1210. 20 Muset, Chem. Zentr. 1907 (1), 1313. 21 Sokolow, J. prakt. Chem. (2), 39, 444 (1889) ; Clarke and Riegel, /. Am. Chem. Soc. 34, 679 (1912). 22 Grigorowitsch and Pawlow, /. Russ. Phys.-Chem. Soc. 23, 172 (1891). 23 Mannich, Ber. 35, 2145 (1902). 24 Freund, Ber. 24, 3359 (1891). 25 Bjelouss, Ber. 45, 625 (1912). 26 Grosjean, Ber. 25, 478 (1892). 27 Kishner, Chem. Zentr. 1900, II, 725. 28 Wallach, Ann. 408, 163 (1915). 29 Kishner, J. Russ. Phys.-Chem. Soc. 43, 951 (1911). 30 Wolff, Ann. 394, 86 (1912). 31 Bjelouss, Ber. 45, 625 (1912). 32 Grignard, Bull. Soc. chim. (3), 31, 753. 33 Kondakow, J. Russ. Phys.-Chem. Soc. 28, 808 (1896). 34 Thorns and Mannich, Ber. 36, 2546 (1903). 35 Ross and Leather, Chem. Zentr. 1906 (2), 1294. 36 Bjelouss, Ber. 45, 625 (1912). 37 Krafft, Ber. 16, 3020 (1883). 38 Grignard, Chem. Zentr. 1901 (2), 624. 39 Freylon, Ann. chim. 20, 58 (1910). 40 Klages, Ber. 36, 3586 (1903). Chapter VI. Polymerization of Hydrocarbons. The polymerization of unsaturated hydrocarbons is a phenomenon the mechanism of which is exceedingly obscure, in fact, no very plaus- ible theories have been advanced to explain this kind of condensation, although the process is accepted and utilized daily in the industries. When unsaturated petroleum hydrocarbons are polymerized by sul- furic acid it has been assumed that alkyl sulfuric acid esters are formed which may then condense with other molecules of the original olefine, with the liberation of sulfuric acid, 1 (1) However, polymerization of hydrocarbons is brought about by a great variety of substances, energetic reagents such as anhydrous aluminum chloride or bromide, zinc chloride, ferric chloride, sulfur chloride, and also such substances as fuller's earth, forms of energy such as light, heat, the silent electric discharge and also certain metals, for example, metallic sodium. It is quite probable, therefore, that we shall have to go much deeper than the drawing of graphic formulae for plausible theories of polymerization; in fact, the question really is one involv- ing the nature of valence. It is beyond the scope and purpose of the present volume to go far afield in reviewing the subject of valence, but there are a number of phenomena, such as polymerization and the mechanism of organic reactions, absorption of light and its alteration as in fluorescence, and the decomposition of substances under the in- fluence of heat which are undoubtedly very closely related and, with 1 Kondakow, J. prakt. cKem. 5k, 442 (1896). 210 POLYMERIZATION OF HYDROCARBONS 211 valence, belong fundamentally to the subject of the constitution of matter. The observations noted in the following discussion have been brought together on account of their interest to organic chemists, rather than for any light that may be thrown upon the mechanism of poly- merization. Ethylene, as noted elsewhere in these pages, is relatively stable, but, at temperatures within the range 400-450, condensation, in contact with iron or copper, is fairly rapid. 2 Many substituted ethylenes con- taining negative groups such as chlorine, or the phenyl group, poly- merize on standing at room temperature, for example, styrene C 6 H 5 CH = CH 2 , vinyl chloride CH 2 = CHC1 (polymerization is par- ticularly rapid in sunlight), 1, 1-dichloroethylene CH 2 = C.C1 2 , vinyl bromide in sunlight, 1-chloro-l-bromoethylene CH 2 = C.ClBr, and 1, 1-dibromoethylene CH 2 = CBr 2 . These substances are rapidly oxi- dized by air or oxygen. On the other hand, allyl chloride and bromide, CH 2 = CH.CH 2 X, trichloroethylene CHC1 = CC1 2 , and 1.2 dibromo- ethylene, CHBr = CHBr are not spontaneously polymerized and are not appreciably oxidized on standing in contact with air or oxygen. Vinyl bromide, CH 2 = CHBr, is polymerized on standing in sun- light to what Ostromuislenski 3 calls a-caouprene bromide. Polymeri- zation under these conditions is very greatly affected by other sub- stances, light low boiling hydrocarbons very greatly retarding the re- action. This polymer, a-caouprene bromide, dissolves very readily in carbon bisulfide and its chemical properties are of particular inter- est, for example, it is quite inert to inergetic oxidizing agents and to concentrated mineral acids. Under the influence of ultraviolet light the polymerization appears to proceed further, forming what have been named P- and y-caouprene bromides. The (3-caouprene bromide is solu- ble in carbon bisulfide but the y-product is quite insoluble but swells in this solvent. The y-product may be converted into the soluble (3-bromide by boiling with chlorobenzene and then precipitating with petroleum ether. The tetrabromide of butadiene-caoutchouc, de- scribed by Harries, also exists in three forms whose behavior is ap- parently identical with the polymerized vinyl bromides just de- scribed. Ostromuislenski regards the polymers of vinyl bromide as structurally arranged as follows, . CH 9 CHBr.CH 2 CHBr.CH 2 CHBr . > Ipatiev, J. Chem. Soc. A 6s. 1907, I, 5. J. Rust. Phys.-Chem. Soc. 44, 204 (1912). 212 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Butylene, amylenes and hexylenes are more easily polymerized than their higher homologues and when condensation of such mono- olefines occurs, ring formation does not take place. Hydrocarbons containing two or more conjugated ethylene bonds are more rapidly oxidized by oxygen and are very easily polymerized, as, for example, butadiene (also called erythrene and divinyl) isoprene, dimethylallene, piperylene, the so-called aliphatic terpenes, myrcene ocimene, cyclopentadiene and the fulvenes, cyclohexadiene and the like. The structure of the polymers of these substances is known in but few instances, but one instance of ring formation is well known, i. e., the condensation of isoprene to the cyclohexene derivative dipentene. Also, dimethyl and tetramethylallene yield cyclobutane derivatives on poly- merization. Dimethylallene, (CH 3 ) 2 C = C = CH 2 , is of iterest as an isomer of isoprene. This hydrocarbon may readily be converted to isopropyl acetylene, and vice versa, indicating that the internal stress in the two hydrocarbons is approximately of the same order, (CH 3 ) 2 C = C = CH 2 5 (CH 8 ) 2 CH C = CH. Tetramethylallene is also easily changed to an acetylene derivative. In the series beginning with allene and including methyl, dimethyl, tri- methyl and tetramethylallene, the stability diminishes with increasing substitution of methyl groups. 4 When dimethylallene condenses to the dimeric cyclobutane deriva- tive six isomeric hydrocarbons are possible but two have been iso- lated, i. e., CH 2 C = C(CH 3 ) 2 boiling-point 61-62, 9 mm. CH 2 C = C(CH 3 ) 2 CH 2 = C C(CH 3 ) 2 boiling-point 37-38, 9 mm. (CH 3 ) 2 C C = CH 2 Tetramethylallene condenses to the hydrocarbon. 5 * Mereshkowski, J. Russ. Phya.-chem. Soc. tf, 1940 (1913). Tetramethylallene was obtained pure for the first time by Mereshkowski, by treating (CH 3 ) 2 C=C CH(CH 8 ) a Br with alcoholic caustic potash in an autoclave at 130, illustrating the marked effect of the double bond on the reactivity of the bromine atom. 8 This hydrocarbon has the unusually high optical exaltation of 2.596, due doubt- less to conjugated linkings of semi-cycUc character and also perhaps to the presence of the cyclobutane ring. POLYMERIZATION OF HYDROCARBONS 213 - ) 2 C C = (CH 3 ) 2 C-C = C(CH 3 ) 2 Polymerization is a property which is probably common to all sub- stances containing ethylene linkings. 6 In a study of the polymerization of a, (5 unsaturated ketones, Ru- zicka [Helv. chim. Acta, 8, 781 (1920) ] showed that the point of at- tack was the ethylene bonds, not the CO groups. Conjugated Dienes and the Synthesis of Rubber. The preparation of conjugated dienes has become a matter of great interest on account of the property, which some of these unsaturated hydrocarbons possess of polymerizing to rubber-like substances. Many industrially important organic substances derived from natural sources can also be manufactured by synthetic methods but the competition Ethylene bonds undoubtedly play a very essential part in the polymerization of fatty oils, and the phenomenon is most pronounced in the case of highly unsaturated oils such as tung, linseed, walnut and certain fish oils. However, the glycerine and carboxyl groups also probably enter into the process of condensation. Kronstein (Ber. 49, 722 [1916] showed that olive and cottonseed oils contain considerable pro- portions of glycerides which gelatinize like tung oil if the non-polymerizing portions of these oils are first removed by distillation. Polymerization of these oils is accom- panied by a decrease in their iodine absorption values. Depolymerization takes place readily since Morell (J. Soc. Ctiem. Ind. 57, 181 [1918]) has shown that the methyl esters, derived from polymerized tung oil, are of normal molecular weight. Salway (J. Soc. Ch&m. Ind. 89, 324T, [1920]) shows that the introduction of free fatty acids accelerates the polymerization of such oils (whale oil), and when the free fatty acids are heated, decrease of the iodine value and refractive index occurs. When the natural glycerides are heated, Salway supposes, (1) splitting off of free fatty acid ; (2) con- densation of the free fatty acid with the unsaturated linkings of the fatty oil; (3) possible anhydride formation in which reaction the free alcoholic glyceryl radicles take part. (1) CH 2 O.OCR CH 2 OH. CH 2 O.OCR -^ CH 2 O.OCR. + RCO.OH. CH 2 O.OC (CH 2 ) 4 CH=CH.C 11 H 19 CH.O.OC (CH 2 ) 4 CH=< (2) CH 2 OH. CH 2 CH 2 -^CH 2 O.OCR C.... C.... CH 2 O.OC. (CH 2 ) 4 CH CHj.AiHw C . . . . C . . . . anhydride LOCR formation 214 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS * of the two methods is often very close, and while the future of syn- thetic rubber is a matter of opinion, all chemists interested in this problem should keep in mind the fact that plantation rubber from Hevea braziliensis can be produced at a cost of approximately twenty- five cents per pound. The present relative importance of synthetic and natural rubber is not a matter of opinion, but of record, and with the exception of a quantity produced in Germany during the war, no synthetic rubber has been produced on an industrial scale or at prices which threaten the rubber plantations. The production of rubber from Hevea plantations has been much greater than the pioneers of the in- dustry had anticipated on account of the "wound response" of the trees on tapping. 7 The synthesis of good gutta percha would seem to offer a better chance of commercial success on account of the slow growth of the trees yielding gutta and the apparent difficulties of solving this phase of the rubber business by plantation methods. Yet even in this case the struggle between synthetic and natural camphor is suggestive. Camphor trees are seldom felled for camphor distillation until they have reached the age of approximately fifty years, yet camphor plan- tations, distilling the leaves and twigs, have been undertaken on an ex- tensive scale and the cost of manufacturing synthetic camphor has increased with the higher cost and diminishing supply of turpentine, the necessary raw material. The history of the subject 8 of artificial rubber has been marred by polemical controversies which have arisen largely on account of definitions and the difficulty of determining just what rubber is struc- turally and the difficulty of proving the identity of such amorphous substances. As regards the question of the identity of polymerized iso- prene rubber and natural Hevea rubber, it now appears that the former, when made either by polymerization by metallic sodium or by per- oxides, is not homogenous, as is indicated by the fact that the ozonides yield succinic acid, acetonylacetone, laevulinic aldehyde and laevulinic acid, corresponding to the two dimeric isoprene complexes 1.5-di- methyl-A^-cyclo-octadiene and 1 .G-dimethyl-A^-cyclo-octadiene. 9 Natural Hevea rubber, on the other hand, appears to be a homogenous product, the ozonide decomposition products being referrable to the i ^ubber Cultivation in the Far East, Science Progress. I. Jan. 1910 ; II. April, 1910. According to Eaton, Chem. Trade J. 1921, 242 approximately 2,000,000 acres are under cultivation for rubber. ,in-.flF f -T, Po , n . d ' J ' Am " Chem - oc- 36, 165 (1914) ; Luff, J. Soc. Chem. Ind. 85, 983 ilo oU P SSP"k.liJS? 0> uhm - Ind - S1 > 616 (1912) ; Gottlob, Indiana, Rubber J. 58, 305, o4o, oyl, 4oo (1919). Steimmig, Ber. 47, 350 (1914). POLYMERIZATION OF HYDROCARBONS 215 1.5-dimethyl-A 1 - 5 -cyclo-octadiene complex. The first direct evidence obtained by Harries of an eight carbon ring complex was later shown to be incorrect, the ketone then considered to be cyclo-octane-1 .5-dione proving to be impure heptane-2 . 6-dione. 10 For practical purposes, however, Ostromuislenski is little concerned with chemical standards of comparison between natural rubber and synthetic colloids resembling them, and advocates X1 a classification based upon the temperatures at which the colloid acquires and loses its elastic properties, and the range between these temperature limits. When these agree closely with the values for natural caoutchouc he proposes that the colloid be classed as normal, regardless of its ozonide decomposition products. Considering the conflicting results of different experimenters with the ozone method, and the difficulties of such work, the proposed classifi- cation would probably be as consistent and also more useful. Harries has contended that the earlier investigators, who discov- ered the polymerization of isoprene, did not really produce caoutchouc, but the question seems a futile and purposeless one. That isoprene could be polymerized to an amorphous rubber-like substance was evi- dent from the early work of Greville Williams 12 who did not recog- nize his product as rubber, but whose description of the product, to- gether with the results of later repetition of his work, indicate that his product was in fact rubber, and Bouchardat 13 who, in 1875, treated isoprene with concentrated hydrochloric acid at 0, and Tilden 14 who obtained a similar product in 1882 and announced later, 1892, that isoprene polymerizes spontaneously on long standing in the light and in contact with air. 15 Wallach 16 showed that light causes the polymeri- zation of isoprene in a sealed tube, but the change is more rapid in contact with oxygen or air. The polymerization of isoprene and sim- ilar dienes is more fully discussed in a separate section on the prop- erties of unsaturated hydrocarbons. As regards the preparation of the dienes, it is possible to note processes which are of industrial promise and, processes which are not likely to become commercial on account of the cost of raw materials or operating difficulties, or both. 10 Harries, Ber. yt, 784 (1914). 11 Cf. J. Russ. Phys.-Chem. Soc. 47. 1928 (1915). "Phil Trans. 150, 254 (1860). a Compt. rend. 89, 361, 1117 (1879). l *Chem. News. L6, 220 (1882). ls Chem. Neics, 65. 265 (1892). Ann. 227, 295 (1885). 216 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The earlier workers prepared isoprene by the destructive distilla- tion of rubber, a typical distillation yielding the following products: Isoprene 6.2 per cent Dipentene 46.0 " " Higher boiling oils 43.8 ' Carbonized residue 1.9 ' Loss and mineral matter 1.9 ' The formation of isoprene by pyrolysis of turpentine was first noted by Hlasiwetz 17 who passed turpentine through a "red-hot" iron tube packed with broken porcelain. A large number of products were obtained and isoprene was not then actually identified in the low boil- ing fraction, but Tilden 18 later repeated this work and proved the formation of isoprene in this manner. The yields of isoprene obtained by the pyrolysis of turpentine and dipentene vary greatly and are gen- erally very low. Harries 19 obtained 10 per cent of isoprene by decom- posing commercial pinene by means of his isoprene lamp (wires heated electrically to low red heat) . Herty and Graham 20 reported yields of 5.5 to 8 per cent from turpentine fractions, and 12 per cent from limonene while Harries obtained yields of 30 to 50 per cent from the latter hydrocarbon. By decomposing under reduced pressure, 4 mm., yields as high as 60 per cent are said to be possible from limonene. 21 Schorger and Sayre 22 also report low yields from turpentine, the two pinenes, a and (3, giving substantially the same yields. Very little attention has been paid to the temperature required for optimum yields of isoprene but, in contact with glass or porcelain, this tem- perature appears to be 550 to 600 , 23 According to Ipatiev, the con- densation of isoprene to dipentene is fairly rapid at 300. Small yields of butadiene and isoprene can also be obtained by the pyrolysis of petroleum oils and have been identified in the low boiling fractions of the light oils obtained by compressing oil gas or Pintsch gas to 150 to 200 pounds pressure. However, the fact that they have been detected in these pyrolytic products is a tribute to the analytical skill of the chemists who investigated these hydrocarbon mixtures. 24 Nevertheless, the preparation of isoprene and butadiene by the pyroly- " Ber. 9, 1991 (1876). 18 J. Chem. Soc. 45, 410 (1884). 18 Ann. 383, 228 (1911). 20 7. Ind. & Eng. Chem. 6, 803 (1914). 21 Staudinger & Klever, Ber. 44, 2212 (1911). 22 J. Ind. & Eng. Chem. 11, 924 (1915). 2 Mahood, J. Ind. < Eng. Chem. 12, 1152 (1920); Heinemann, Brit. Pat. 14,040; 24,236 (1910); 1953 (1912); Stephen, U. S. Pat. 1,057,680; Ostromuislenski, French Pat. 442,980 (1912) ; Sobering, German Pat 260,934 (1913). 34 Armstrong & Miller, J. Chem. Soc. 49, 74 (1886). POLYMERIZATION OF HYDROCARBONS 217 sis of petroleum oils at about 700, particularly in vacua, 25 has re- cently been patented. This method presumably would give better re- sults with light petroleum oils containing cyclohexane, cyclopentane, and their simpler homologues, since it has been claimed that tetra- hydrobenzene yields a certain proportion of butadiene on decomposition under these conditions. 26 However, in the ten years which have elapsed since this work was done, there have been no industrial developments along this line and considering the small yield of the desired dienes, the value of petroleum oils for other uses, and the difficulty of purify- ing the desired hydrocarbons, it is very doubtful indeed if the direct pyrolysis of hydrocarbons will ever prove to be an economic method of producing these hydrocarbons. In this connection, it should be noted that Ostromuislenski 27 has shown that on polymerizing iso- prene containing amylene or similar defines, the resulting "rubber" is very sticky and soft. Petroleum pentane is mostly normal pentane but attempts have been made to utilize this hydrocarbon as a raw material for the manu- facture of isoprene. It may be said of all the chemical methods for the preparation of these unsaturated hydrocarbons that no really new methods or reactions have been developed; all of the known methods of producing unsaturated hydrocarbons have been applied to the prepa- ration of these dienes but the great majority involve the elimination of halogens, usually chlorine, or of hydroxyl groups in the form of water. The production of isoprene from normal pentane involves the change to the carbon structure of isopentane. This is accomplished by one patentee 28 by taking advantage of the isomerization of defines effected by heat, which has already been noted in the case of the butylenes. Thus pentane is chlorinated to a mixture of the monochlorides and these are converted to amylenes by pyrolysis in contact with barium chloride, lime or other methods, and the mixture of amylenes then passed over alumina at about 450. Partial rearrangement to tri- methylethylene occurs and on treating the resulting mixture of amyl- enes with hydrogen chloride this hydrocarbon reacts most readily, the chloride thus formed being separated by fractional distillation. The hydrocarbons thus separated are passed again over alumina at 450, and so on. The purified monochloroisopentane is converted to tri- methylethylene by the usual methods and this treated with chlorine to 25 Engler and Staudinger, Ber. 46, 2468 (1913) : German Pat. 265,172 (1912). 28 Farbenfabr, Elberfeld, German Pat. 241,895. 27 J. Russ. Phys.-Chem. Soc. 48, 1071 (1916) ; Chem. Abs. 11, 1768 (1917). 28 Badische, German Pat. 280,596 (1919). + HC1 218 CHEMISTRY Of THE! NON-BENZENOID HYDROCARBONS form the dichloride which then forms isoprene with the elimination of two molecules of hydrogen chloride. The reactions involved are as fol- lows: CH 3 CH 2 CH 2 CH 2 CH 3 - > monochlorides - > rCH 3 >C = CHCH, amylenes - > 4 CH 4 l_ and isomers CH 3 >CC1.CH 2 CH 3 - >CH 3 CH. >C == CHCH 3 CH 3 (pure) CH. > CC1 . CHC1 . CH 3 > CH 2 CH3 Y-CH = C CH 3 Petroleum pentane is one of the cheapest raw materials which have been suggested for this purpose but the process involves a large num- ber of operations, distillations, purification of intermediates and the losses are large, for example, if each operation indicated above gave a yield of 90 per cent the final net yield of isoprene would be about 7 per cent. Pentane can, in fact, be chlorinated to monochloropentanes with a yield of about 90 per cent, exclusive of vaporization losses, but the losses on isomerizing the amylenes are large and it is impossible to chlorinate trimethylethylene without partially chlorinating further to trichlorides and tetrachlorides. Several patented processes employ phenol and cresols as raw ma- terials. Phenol may be hydrogenated to cyclohexanol, with good yields, and this alcohol may then be dehydrated by heating in contact with alumina, thoria or kaolin, to give cyclohexene. Cyclohexene gives small yields of butadiene and ethylene by direct pyrolysis, CH 2 H 2 C CH | > CH 2 OH CH = CH 2 + C 2 H 4 H 2 C fcH CH, POLYMERIZATION OF HYDROCARBONS $1$ Chlorination of cyclohexene to the dichloride, and then decomposing this, yields the conjugated diene, cyclohexadiene, but this hydrocarbon polymerizes to a substance more nearly resembling resin than rubber. Benzene itself is readily hydrogenated to cyclohexane and this may be converted to cyclohexene through the monochloro derivative by the usual methods 29 but none of these materials yield final products of good quality. It is much easier to prepare isoprene and butadiene, and in much purer condition, by using butyl or isoamyl alcohol as the raw ma- terials. A new method for the manufacture of n. butyl alcohol has been developed based upon the fermentation process of Fernbach, 30 the two principal products being n. butyl alcohol and acetone. Al- though originally developed in connection with the synthetic rubber problem it was carried out on a large scale during the recent war, essentially as a process for the manufacture of acetone. At com- paratively high temperatures butyl alcohol is decomposed partially to butadiene 31 but, as in many pyrolytic processes, the yields are small. The alcohol may be converted to the corresponding chloride and the resulting butyl chloride then chlorinated to the dichlorides which may then be decomposed by methods already mentioned, to butadiene, 32 CH 3 CH 2 CH 2 OH > CH 3 CH 2 CH 2 C1 dichloride > butadiene By similar methods isoamyl alcohol, the chief constituent of fusel oil, may be converted by hydrogen chloride to isoamyl chloride, which on chlorination yields a mixture of dichlorides, (CH 3 ) 2 CH.CHC1.CH 2 C1 boiling-point 142 C. (CH 3 ) 2 CC1.CH 2 CH 2 C1 " 152 C. CH 2 C1 >CH.CH,.CH 2 C1 170 C. CH 3 Of these dichlorides the second is the principal product, but the crude mixture, boiling-point 140-180, is used for the production of isoprene, the yield, according to Perkin, 33 being 40 per cent of the theory. As pointed out by Perkin the total available quantity of ordinary fusel oil, about 3500 tons, is wholly inadequate as a raw material for rub- M Schmidt, Hochschwender & Eichler, TL S. Pat. 1,221,382. w Fernbach & Strange, Brit. Pat. 15,203; 15,209; 16,925 (1910). The butyl alcohol contained in ordinary fusel oil from the manufacture of alcohol is isobutyl alcohol and is only a minor constituent. 31 Perkin & Mathews, J. Soc. Chem. Ind 82, 884 (1913). 82 Cf. Badische, German Pat. 255,519 (1913); 264,008 (1911); Harries, German Pat. 243,075; 243,076 (1910) ; Brit. Pat. 18,653; 22,035 (1912). 33 J. Soc. Chem. Ind. 31, 616 (1912). 220 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS ber synthesis, and is relatively high priced on account of its many in- dustrial applications. The per cent of isoamyl alcohol in commercial fusel oil varies somewhat according to the distillation range over which it is collected, but the fraction distilling at 128-131 contains approximately 87 per cent isoamyl and 13 per cent active amyl alcohol C 2 H 5 CH(CH 3 ) .CH 2 OH. Two typical analyses of commercial fusel oils are as follows, 34 From fermentation Percent From fermentation Percent of potatoes by wt. of corn by wt. n. Butyl alcohol ............. 6.8 n-Propyl alcohol ............. 3.7 Isobutyl alcohol ............. 24.3 Isobutyl alcohol ............. 15.7 Amyl alcohols ............... 67.8 Amyl alcohols ............... 75.8 Fatty acids .................. .04 Hexyl alcohols .............. 0.2(7) Fatty acids ................... 56 It is probable, in view of the researches of Ehrlich, 35 that such varia- tions in the character of fusel oils are due to differences in. the yeasts employed for fermentation or proteins otherwise introduced rather than the materials fermented. A sample of fusel oil from corn, examined by Pringsheim, 36 contained isopropyl and normal butyl alcohols in ad- dition to the normal propyl and isobutyl alcohols which are normally present in fusel oil from this source. In view of the efforts which have been made to utilize cheap fer- mentable material and the resulting butyl and amyl alcohols, as raw material for rubber synthesis, this phase of the work is reviewed here. Ehrlich claims that in ordinary yeast fermentation the fusel oil alco- hols are derived from the decomposition of protein material, or rather the amino acids leucine, isoleucine and the like, NH 2 (CH 3 ) 2 CH.CH 2 CH< + H 2 leucine C0H 2 - > (CH 3 ) 2 CH.CH 2 CH 2 OH + CO 2 + NH 3 isoamyl alcohol. Ehrlich established the following relations, (1) Pure yeast and pure sugar yields no fusel oil. (2) " " " " " + leucine yields isoamyl alcohol. (3) " + isoleucine yields d. amyl alcohol. The addition of ammonium carbonate or asparagin to yeast fermen- tations decreases the yield of fusel oil and the addition of leucine, or 4 T he Nitrocellulose Industry", Worden. 86 Cf. Brit. Pat. 6,640 (1906) ; Ber. 40; 1027 (1907). "BiocJiem. Z. 16, 243 (1909). POLYMERIZATION OF HYDROCARBONS 221 protein rich in this complex, increases it. Only traces of fusel oil are formed by alcoholic fermentation by means of Buchner's cell-free pressed yeast juice. 37 However, normal butyl alcohol at least can be- come, under certain conditions, one of the principal products derived from the sugar undergoing fermentation. Realizing the inadequacy of the supply of commercial fusel oil for possible rubber synthesis, Per- kin and his associates undertook to develop a special process of fermen- tation which would yield larger proportions of butyl or amyl alcohols. Although the anaerobic Bacillus butylicus was discovered by Fitz 38 in 1878 in a study of glycerine fermentation, and Perdrix 39 had described an anaerobic bacterial fermentation which gave very high yields of fusel oil, it does not seem to have occurred to anyone else to utilize this possibility until it had been developed by Fernbach and Strange. 40 As has been previously noted both the major products of this fermenta- tion, acetone and n . butyl alcohol, are necessary raw materials required by several different processes for the production of butadiene and dimethyl butadiene. All of the known methods of decomposing alcohols to unsaturated hydrocarbons have been applied to the problem of producing these sim- ple conjugated dienes. Butyleneglycol yields butadiene when passed over heated kaolin, alumina, or aluminum phosphate. 41 The butylene- glycol, required by this process, can be made from acetaldehyde, the primary raw material therefore being ethyl alcohol or acetylene. Acet- aldehyde may be condensed by well-known methods to aldol, which upon reduction yields butylene glycol, Alcohol, or acetylene > acetaldehyde > CH 3 CH(OH) .CH 2 CHO -+ CH 3 CH (OH) . CH 2 CH 2 OH * CH 2 = CH . CH = CH 2 Butyraldehyde yields a certain amount of butadiene* when passed over kaolin at 500-600 under reduced pressure 42 and isovaleric aldehyde yields some isoprene under the same conditions. 43 Secondary butyl alcohol can be prepared by (1), reduction of the commercial solvent methyl ethyl ketone, derived from "acetone oil," or (2), treating oil gas with 80 per cent sulfuric acid and hydrolysing the butyl hydrogen 37 Buchner & Meisenheimer, Ber. 39, 3201 (1906). Ber. 11, 481, 878 (1878). 39 Z. Spiritusind. 14, 177 (1891). * French Pat. 488,364 (1913). "Mathews, Strange & Bliss, Brit. Pat. 3,873 (1912); Cf. Bayer, German Pat. 261,642 (1913). 2 U. S. Pat. 1,033,327. U. S. Pat. 1,033,180. 222 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS sulfate so formed. The secondary butyl alcohol may then be decom- posed catalytically, with very good yields, to butylene, which on chlorination or bromination and subsequent decomposition, yields bu- tadiene. As has already been noted, it is very difficult completely to remove halogens from such substances on account of the stabilizing in- fluence of an adjacent double bond, the second ( 1 ) CH 3 CHBr . CHBr . CH 3 - -- > CH 3 CHBrCH = CH 2 (2) CH 3 CHBr . CH = CH 2 - ~* CH 2 = CH CH = CH 2 reaction taking place with difficulty (higher temperatures) and when the temperatures are sufficiently high for the complete removal of halo- gen, loss of the desired diene occurs through secondary reactions. The condensation of acetaldehyde and ethyl alcohol by passing over heated copper, followed by decomposition of the condensation product by passing over heated alumina, has been noted by Ostromuislenski 44 as a possible method, and the chemical changes, which really involve five consecutive reactions, may be summarized as follows: 2C 2 H 5 OH -> CH 3 CHO + C 2 H 5 OH -> CH 2 = CH.CH =CH 2 + 2H 2 The yields of butadiene are poor and considering the number of other reactions which also occur in this process, it is not likely to become of industrial interest. That tertiary alcohols are much more easily decomposed to un- saturated hydrocarbons, than secondary and primary alcohols, is well known, and advantage is taken of this fact in the employment of pinacone as an intermediate product. Thus acetone may be reduced and condensed to pinacone under a wide range of conditions and the use of amalgams for this purpose is particularly promising. 45 Pinacone is smoothly decomposed by passing over alumina at about 400 giving good yields of dimethylbutadiene. 46 acetylene acetate of lime > acetic acid starches and sugars starches and sugars **J. Ruas. Phys.-Chem. 800. W, 1472, 1494 (1915) ; ,7. Chem. Soc. Abs. 1916, I, 4. "Holleman, Rec. trav. chim. 25, 206 (1906) ; Bull. soc. cMm. 1910, 454. German Pat. 250,086. POLYMERIZATION OF HYDROCARBONS 223 CH 3 CH 3 CH 2 CH 2 o-c -, \_c 7/ dimethylbutadiene The synthetic rubber manufactured in Germany during the recent war was made from pinacone and dimethylbutadiene, the latter material being polymerized in sealed iron drums during a period of several months. Pinacone chlorohydrin also yields dimethylbutadiene, when heated with bases, dimethylaniline being recommended for this purpose, 47 and Kondakow 48 claims that pinacone dichloride gives better yields of the diene than pinacone itself. Decomposition of the alcohol pentene-2, ol-4 by passing over alumina or kaolin at 400 has been employed for the preparation of piperylene. Under certain conditions acetalde- hyde condenses to crotonic aldehyde and on methylating this aldehyde pentene-2, ol-4 is formed. OH 2CH 3 CHO CH 3 CH = CH.CHO * CH 3 CH = CH.CH< CH 3 -* CH 3 CH = CH.CH = CH 2 Many methods have been described which make use of well-known syntheses, but which are interesting from a theoretical point of view. Kyriakides 49 has described an interesting synthesis starting with chlo- roacetone, which is ethylated, and the resulting chlorohydrine is then treated with caustic alkali to obtain the oxide, as indicated in the fol- lowing, CH 3 COCH 2 C1 - > CH 3 CH 3 >C CH 2 - > >C CH 2 Y "German Pat. 319,505 (1916). 48 J". prakt. Chem. 62, 169 (1900). W J. Am. Chem. Soc. 36, 663 (1914). CH 3 v 224 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The oxide is decomposed by heating in the presence of kaolin at 440-460. Dimethylallene may be partially converted into isoprene by re- arrangement 50 and Ipatiev 51 has prepared isoprene from this hydro- carbon by adding two molecules of hydrogen bromide, followed by de- composing the dibromide by well-known methods, CH 3 CH 3 > C = C = CH 2 + 2HBr > CBr . CH 2 CH 2 Br CH 3 CH 3 CH 2 /-i OTT C*TT CH 3 Among the reactions of theoretical interest which have been em- ployed in research in this field, may be mentioned Euler's 52 prepara- tion of isoprene by the exhaustive methylation of methylpyrollidine, as follows, CH 3 CH CH 2 >NH + 2CH 3 I + NaOH > H 2 CH 2 CH 3 CH CH 2 CH 3 C == CH 2 | >N(CH 3 ) 2 I > | CH 2 CH 2 CH 2 CH 2 N (CH 8 ) 2 CH 3 C == CH 2 nz: isoprene It will be recalled that this method has been frequently used by von Braun and others in the investigation of alkaloids, on account of the ease with which nitrogen can be removed from organic bases. Phenol may readily be hydrogenated to cyclohexanol, which on oxi- dation by nitric acid 53 yields adipic acid. Conversion of this acid to M Webel (U. S. Pat. 1,083,164), claims that as. dimethylallene rearranges to isoprene when passed over alumina at 300, and preferably under diminished pressure^ 81 J. prakt. Chem. 55, 4 (1897). 62 J. prakt. Chem. 57, 132 (1898). 83 Bouveault and Locquin, Bull. Soc. chim, 1908, 3 t 437. POLYMERIZATION OF HYDROCARBONS 225 the amide, followed by treatment with hypochlorite, yields tetramethyl- enediamine and the method of exhaustive methylation applied to this diamine yields butadiene; cresol, treated similarly, yields isoprene. Polymerization of Conjugated Dienes to Rubber-like Substances. As pointed out elsewhere in these pages the polymerization of iso- prene had been observed by Greville Williams, Bouchardat, Tilden and Wallach. But the first attempt to polymerize isoprene which had been prepared from sources other than rubber itself was Tilden's investiga- tion of isoprene made by the pyrolysis of turpentine, published in 1888. 54 Tilden states that, "The action of hydrochloric acid on iso- prene converts it partially into caoutchouc ; the latter seems to be ob- tained more easily starting with the oily polymeride resulting from the action of heat." Some 28 years later, Ostromuislenski 55 showed clearly that the character of synthetic isoprene rubber was markedly affected by the method of polymerization; that on heating isoprene to 80-90 it undergoes spontaneous polymerization to a dimeride, (3-myrcene, and this hydrocarbon then yields "normal" caoutchouc when polymerized by sodium, or barium peroxide. However, when isoprene itself is treated with these reagents the resulting rubber is not normal. 56 Til- den seems to have been aware all along that rubber might be formed by the polymerization of isoprene. The polymerization of the isomeric hydrocarbon piperylene, CH 3 CH = CH CH = CH 2 had been ob- served by Hofman 57 and by Schotten, 58 but their publications contain no suggestion that their product resembled rubber. In 1892 Tilden, 59 in a communication to the Philosophical Society of Birmingham, stated, "I was very much surprised to find -that the contents of the flasks containing isoprene, prepared from turpentine, had entirely al- tered in appearance. Instead of a colorless, limpid liquid, there was now a thick syrup, in which floated several pieces of a yellow solid material. On examining it more closely this was found to be caout- chouc." * * * "A solution of synthetic rubber leaves, on evapora- tion, a residue which completely resembles in all its characteristics a like preparation made with Para rubber." * * * "Artificial rub- ber combines with sulfur in the same way as natural rubber, giving an elastic, resistant mass." A little later Tilden's results were con- 54 J. Chem. Soc. 45, 411 (1888). 65 J. Buss. Phj/s.-Chem. Soc. 48, 1071 (1916). "7. Ruse. Phys.-Chem. Soc. 47, 1928 (1915). "Ber. Ik, 665 (1881). K Ber. 15, 425 (1882). * Chem. News. 65, 265 (1895). 226 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS firmed by Weber 60 who prepared about 200 grams of synthetic iso- prene rubber, and Pickles 61 also confirmed Tilden's statement that isoprene polymerizes on standing in contact with air. The marked influence of oxygen upon the formation of polymers is well shown by experiments reported by Engler, 62 in which duplicate samples were exposed to oxygen and carbon dioxide, at 80 C. The very rapid polymerization of styrene, and the conjugated dienes, iso- prene and myrcene, are particularly noteworthy. PER CENT POLYMERS FORMED ON STANDING. In contact with COa In contact with 0* Days exposed 12841284 Limonene 2% 4% 5% 8% 4% 6% 8% 9% Phellandrene 4 6 8 9 9 13 16 21 Pinene 1 22 3 3 4 4 5 Myrcene 8 13 18 22 20 30 40 50 Camphene 3 4 5 6 5 7 8 9 Isoprene 14 per cent in 10 hours; 35 per cent, 10 hours Styrene 22 per cent, 20 hours; 67 per cent, 20 hours The presence of moisture apparently has no effect upon the rate of polymerization of hydrocarbons, although the smallest trace of mois- ture acts catalytically upon the polymerization of the aldehyde, gly- oxal; 63 monomolecular succinic dialdehyde behaves in a similar man- ner. The polymerization of dimethyl butadiene, dimethyl 2-3 buta- diene 1-3, to a rubber-like substance was first effected by Kondakow, 64 who noted that it polymerized spontaneously and more rapidly than iso- prene or butadiene. His publications upon the polymerization of this dimethylbutadiene, which he prepared from pinacone, would seem to justify Kondakow's claims of priority, so far as the dimethylbutadiene process, later patented and used industrially in Germany, is concerned. It was noted also, and confirmed by others, 65 that when dimethyl butadiene polymerizes, either spontaneously or in the presence of alco- holic caustic potash, a dimeride and a trimeride are produced, in addi- tion to the rubber-like substance. It was important for the technicali- ties of later patent controversies that Kondakow had described his di- methylbutadiene rubber as insoluble in most organic solvents, although it is now generally recognized that this property varies considerably " J. Boc. Chem. Ind. 13, 11 (1894). n J. Chem. Soc. 97, 1085 (1910). 93 8th Int. Congr. Appl. Chem. 25, 661 (1912). "Harries, Ber. 40, 165 (1906) ; 41, 255 (1908). 64 J. prakt. Chem. 6//, 109 (1901). 6B Lebedew, J. Rues. Phys.-Chem. Soc. $1, 1818 (1909); Harries, Ann. S88, 210 POLYMERIZATION OF HYDROCARBONS 227 with all rubbers, depending upon the degree of polymerization; in fact, vulcanization is essentially a process of effecting higher degrees of polymerization. It is well known also that dimethylbutadiene poly- merizes more rapidly than other similar hydrocarbons. Perkin states, "The situation in 1906 might be summed up in this way; it had been recognized, in a more or less general way, that most compounds con- taining a system of conjugated double linkings, show a tendency to polymerize, more or less readily. The polymerides are either viscous, ill defined substances, or well characterized caoutchoucs; or, again, hard resinous solids, like polystyrene. Their properties vary accord- ing to their method of preparation, and according to the molecular weight of the hydrocarbon employed as a raw material." Like natural Para rubber, Kondakow's rubber can be depolymer- ized by heat, although more readily than Para rubber, the principal product being a dimeric dimethylbutadiene resembling dipentene and which Richard 66 and Kondakpw regard as having the structure, CH 3 or The same hydrocarbon is also formed by careful polymerization of 2.3-dimethylbutadiene-(1.3). In the polymerization of isoprene to synthetic isoprene rubber a dimeric isoprene is formed, in addition to the dimeride, dipentene. This second hydrocarbon, called di-isoprene or myrcene by earlier writers,- yields a liquid tetrabromide, in con- trast to the crystalline dipentene tetrabromide. According to Lebedew "Compt. rend. 153, 116 (1911) ; According to Lebedew and Mereshkowski (J. Buss. Phys.-Chem. Soc. 45, 1249 [1913]) this dimeride has the following properties; boiling- point 85 at 13 mm., 205 at 750 mm., D|- O 0.8741 n D 1.48074 ; dry HC1 yields a Me //C.Me.CH a / monohydrochloride MeC >C.Me.CCl , boiling at 122-124 under 17 mm. ; T \CH 2 CH 2 \ Me oxidation by benzoyl peroxide, according to Prileschajev (q.v.) yields a dioxide which is hydrolyzed by aqueous benzoic acid to a tetrahydric alcohol* 228 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS and Mereshkowski 67 this hydrocarbon is 1 . 3-dimethyl-3-ethenyl-A 6 - cyclohexene. 68 On hydrogenating in the presence of platinum the side chain ethenyl group is first saturated, following the general behavior of substances containing double bonds in both the ring and side chain. Piperylene similarly should yield two dimerides. Butadiene yields a dimeride, C 8 H 12 , boiling at 36 under 23 mm., 129.5-131 under 760 mm. pressure. Hydrogenation by PaaPs method yields ethyl cy- clohexane; bromine reacts to form a tejrabromide melting at 69.5- 70.5 and oxidation yields the acid from which facts, and reasoning by analogy from the relations between isoprene and dipentene, Lebedew 69 concludes that the hydrocarbon is 1 -ethenyl- A 4 -cy clohexene, CH CH, CH \ CH.CH^CH. CH 2 CH 2 These details are of first importance as the yield of synthetic rubber, by present methods of polymerization, is seriously diminished by the formation of these oily polymers. As to why or how sodium effects the polymerization of isoprene, m Loc. cit. Cf. Harries, Arm. 383, 157 (1911). 69 J. Ruas. Phys.-CTiem. 8oc. S, 1124 (1911). POLYMERIZATION OF HYDROCARBONS 229 no one has hazarded a theory. Perkin 70 relates that Weizmann and Mathews were induced to try the effect of permitting the hydrocarbon to stand in contact with the metal, by their having noted the conversion of dimethylallene to isopropylacetylene by metallic sodium, (CH 3 ) 2 C = C = CH 2 (CH 3 ) 2 CH C^ a reaction which had been recorded by Favorsky. 71 This discovery, the polymerization of isoprene by sodium, was, according to Perkin, made by Weizmann and Mathews in July and August, 1910, although it was first publicly described in the following year by Harries. 72 The same discovery had evidently been made by Harries in the "em! of (the year) 1910." The polymerization of hydrocarbons may, according to Lebedew and Mereshkowski 73 be grouped in several well defined classes, (1), the styrene type, peculiar to ethylene hydrocarbons with unsymmetri- cal substitution of the hydrogen atoms by phenyl, or other groups, and yielding amorphous polymers of very high molecular weight and whose structures are not yet known; (2) the stilbene type, shown by sub- stances having symmetrically substituted groups; (3) the acetylene type, whose characteristic is the formation of benzene or its deriva- tives; (4) the allene type, yielding cyclobutane derivatives; (5) the 1 .3-butadiene or isoprene type, which forms cyclohexane derivatives and also polymers of high molecular weight, usually amorphous, and including rubber-like substances. The structures of the polymers of the styrene and stilbene type, when ascertained, may show that these two classes are really of the same type of polymerization. With isoprene and 2.3-dimethylbutadiene-(1.3) it. has been shown that with increasing temperature the proportion of the dimeride in- creases and that of the rubber-like polymer decreases. Since the re- action is markedly affected by catalysts, it follows that, for maximum yields of "synthetic rubber," a catalyst and the lowest possible temper- ature should be employed. The search for raw materials for the prepa- ration of the simpler conjugated dienes, and the effort to discover effi- cient methods for the preparation of these hydrocarbons has involved a great deal of research. The finishing step in the process, polymeri- zation, is still without a theory sufficiently tangible or plausible to be of use as a guide for further work. There has been a very noticeable 70 Loc. cit. " J. Russ. Phys.-Chem. Soc. 19, 558 (1887). Ann. S83, 157 (1911). " J. Russ. Phys.-Chem. Soc. 45, 1249 (1913) ; J. Chem. Soc. Alts. 1913, I, 1285. 230 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS abatement of research on rubber synthesis since 1912 (the manufacture of synthetic rubber may have been, for Germany, a war preparedness measure) . It is certain that all methods of synthesis previously known have been applied to this problem, and synthetic rubber has not yet made a place for itself. New methods of synthesis or polymerization, or changed economic values with respect to raw materials for synthe- sis, or cost of plantation rubber, may affect the situation in ways which none can now foresee. POLYMERIZATION OF HYDROCARBONS 231 <* < 2 O .9 || .^ e T | o e-g o^ 00 'I oo "* " ' ' $ cu ^3 S CO I S3 ^ S ' 'a % fj ^H i i a I a 3 2 S5 ^ S w e^ -^ ^ 1314s -rfl o i sho sh O r-t ^ l-H 51^ o!o s 3 l>. 00 O5 O Tg o O 'O -*^ CO B I i v g? S c? $ W O -J..8 W O o ci-g o B i a 7? CU ?% "H 11 a 3 fl r-t (N CO ^ "3 CO 232 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS 1-4 jQQ 1 |.t i-H TM l a 1 1 & ^ | 1 \ ? j =3 <1 "^ i 1 ^ 1 P ^ 6 CO <1 1 2 T. 5 "S, ^ 1 ll !i & ) < 3-Methy] 4-Methyl 9 n 3-Methyl \ 3.5-Dime i s 4 T ^ i t 3.5-Dime 1 9 C4 0) a ! CO Chapter VII. Cyclic Non-benzenoid Hydrocarbons. General Methods of Synthesis of Cyclic Non- benzenoid Hydrocarbons. Many of the well-known condensation reactions of the paraffine series can take place with intramolecular condensation or ring forma- tion. Thus the type condensation of acetic ester to acetoacetic ester can take place with the diethyl esters of adipic, pimelic and suberic acids to form 5, 6 and 7 carbon rings, respectively, for example, CH 2 CH 2 . C0 2 C 2 H 5 CH 2 CH., CH 2 < CH 2 < ">CO CH 2 CH 2 . C0 2 C 2 H 5 CH 2 CH . C0 2 C 2 H 5 Glutaric and succinic esters do not condense in this manner to give cyclobutane and cyclopropane derivatives, illustrating the relative dif- ficulty with which ring structures of 3 or 4 carbon atoms are formed. The calcium salts of adipic, pimelic and suberic acids give, on heating, cyclopentanone, cyclohexanone and cycloheptanone respectively, but calcium succinate gives the cyclic diketone CH 2 CO CH 2 CH 2 CO CH 2 When the calcium salt of cyclohexane -1.3-dicarboxylic acid is decom- posed by heat the bicyclic ketone is formed which Stark x calls "deme- thylated pinone." melting*"' /nt 170' 157 ' 158 ' d ' 9322 ' semicarbazone 233 234 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Ring closing incidental to Grignard's synthesis of carboxylic acids has been observed, as in the case of 1 . 5-dibromopentane, which with mag- nesium forms the dimagnesium compound, and then on treating with carbon dioxide yields cydohexanone and pimelic acid. 2 Succinic ester and sodium condense to give a six carbon ring, suc- cinosuccinic ester. On hydrolyzing and heating with sulfuric acid the cyclic diketone is obtained which may be reduced to cyclohexane by converting it first into the alcohol, cyclohexanediol-1.4, then into the corresponding iodide and reducing this with zinc dust and acetic acid, exactly as in the case of aliphatic alcohols and iodides. Nc^ *. ^ H ^rY Hi "^ Cyclohexane The method of Wiirtz and Fittig, of treating alkyl halides with me- tallic sodium effecting condensation with formation of sodium halide, has been employed for ring formation. Freund made cyclopropane by treating trimethylene bromide with sodium. 3 CH 2 Br CH 2 CH 2 < + Na 2 > CH 2 < I + 2NaBr. CH 2 Br CH 2 Methyl cyclobutane was prepared by Perkin, Jr., in a similar way from 1.4 dibromopentane. 4 CH 2 - CHBr . CH 3 CH 2 CH CH 3 | +Na 2 > I I CH 2 -CH 2 Br CH 2 -CH 2 'Grignard & Vignon, Compt. rend. 1U 1358 (1907> KlvcoV^ow /%n 25 (1882K The origi al material for this synthesis, trimethylene Ilycerine dis^ill a tion. m n commercial P'odnct, being isolated from the forerunning in *J. Chem. 8oc. 5S, 201 (1888) ; 65, 599 (1894). CYCLIC NON-BENZENOID HYDROCARBONS 235 and cyclohexane has been made from 1 . 6-dibromohexane and sodium. Condensations to carbocyclic derivatives have also been made as indicated by the following synthesis ; the disodium compound of acetone dicarboxylic ester being treated with iodine 5 gives, C0 2 R C0 2 R C0 2 R C0 2 R CHNa + I 2 + NaHC CH CH C0< >CO >CO< >CO CHNa + I 2 + NaHC CH CH C0 2 R C0 2 R C0 2 R CO,R Instead of using free iodine or bromine, alkyl halides may react with sodium malonic ester or similar sodium compounds, as in the following syntheses carried out by W. H. Perkin, Jr. 6 CH 2 Br C0 2 R CH 2 CO 2 R + CH 2 < + 2CH 3 ONa I >C< CH 2 Br C0 2 R CH 2 C0 2 R from which cyclopropane monocarboxylic acid is readily made by loss of C0 2 from the dibasic acid. In the same way trimethylene bromide (1) and pentamethylene bromide (2) yield CH 2 ( 1 ) > CH 2 < > CH . C0 2 H cyclobutanecarboxylic acid CH 2 CH 2 CH 2 (2) >CH 2 < >CH.C0 2 H cyclohexanecarboxylic acid CH 2 CH 2 The above syntheses are capable of considerable variation and exten- sion as the following syntheses indicate: (1) CH 2 C1 CH 2 (C0 2 R) 2 CH 2 CH(C0 2 R) 2 + +2C 2 H 5 ONa^| CH 2 C1 CH 2 (C0 2 R) 2 CH 2 CH(C0 2 R) 2 CH 2 CNa (CO,R) 2 CH 2 CH (C0 2 R) 2 I + Br 2 -*I -> CH 2 CNa (C0 2 R) 2 CH 2 CH (C0 2 R) 2 CH, CH.C0 2 H CH 9 CH. C0 2 H v. Pechmann, Ber. SO, 2569 (1897). Ber. 35, 2091 (1902). 236 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS (2) CH 2 -C(C0 2 R) 2 CH 2 CNa(C0 2 R) 2 / \ CH 2 < + CH 2 I 2 -*CH 2 CH 2 CH a - CNa(C0 2 R) 2 \ / CH 2 C(C0 2 R) 2 CH 2 CH.C0 2 H ->CH 2 < >CH 2 CH 2 CH.CO.H Acetoacetic ester and ethylene bromide yield cyclopropyl methyl ke- tone in the following manner: CH CH 3 CH 2 Br CO CO II CH 2 | CH 2 CH 2 Br + H 2 C + 2C 2 H 5 ONa- I >C -* [ >CH.CO.CH 3 | CH 2 | CH 2 C0 2 R C0 2 R Polymerization of unsaturated substances sometimes results in ring formation, as in the condensation of isoprene to dipentene and isoprene- rubber. r J"; I" . v > en CH CHi ^N XH 2 L J :" *C*^ Vinylacrylic acid also polymerizes readily, in the following man- ner, 7 when heated with barium hydroxide. CH 2 = CH . CH = CH . C0 2 H . CH 2 . CH =CH . CH 2 CH 2 = CH.CH = CH.C0 2 H. ^CH 2 .CH =CH.CH 2 The Grignard reaction has also been employed to effect ring closing as in the preparation of 1-methyl-l-hydroxycyclopentane by Zelinsky and Moser. 8 T D8bner, Ber. S5, 2129 (1902). Ber. S5 t 2684 (1902). CYCLIC NON-BENZENOID HYDROCARBONS 237 CH C = I C = Mgl CH 2 CH 2 CH 2 CH 2 ^CH CH OTI CH >CH A w OH" c /X CH 2 ^H 2 L CH 3 OMgl \ / C CH CH CH In the same manner that acetone condenses to give mesityl oxide, diacetylbutane treated with sulfuric acid yields methylcyclopentene- methyl ketone. 9 CH 2 CH 2 CO CH 3 CH 2 C COCH 3 CH 2 < >CH 2 < || CH 9 CO CH, CH, CH 3 Diacetylpentane when similarly treated yields a methylcyclohex- enemethyl ketone. CH 2 CH \ C COCH 3 C CH CH Condensation of alkyl halides with benzenoid hydrocarbons, with elimination of halogen acid, takes place very rapidly in the presence of anhydrous aluminum chloride (the Friedel-Crafts synthesis). This reaction has been employed for ring closing, as, for example, phenyl- valeryl chloride being converted into \i~c^ benzo- cycle heptan one Kippin* & Perkin, J. CTiem. Soc. 57, 14, 24 (1890). 238 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Halogen acids are very readily eliminated from alkyl chlorides alone, when treated with aluminum chloride, as for example, chloro- pentanes and chlorohexanes, but the nature of the resulting products has apparently never been investigated. Condensation of the aldehyde citronellal to isopulegol, through the action of acetic anhydride, is not typical but illustrates the tendency, so frequently observed, to form rings of six carbon atoms. CH 3 CH 3 CH CH H 2 C CH 2 H 2 C>' CH 2 H 2 C CHO H 2 C C< \ \ / OH CH C ' . . CH 3 CH 3 CH 3 CH 3 citronellal isopulegol Methylheptenone is condensed by dehydrating agents to a mixture of m-xylene and 1 . 3-dimethyl-A 3 -cyclohexene. The initial reaction product apparently is 1.3-dimethyl-A 1 - 3 -cyclohexadiene, which it will be noted has the same arrangement of the double bonds as in a-ter- pinene and the ease with which terpinene is converted to cymene is well known. Apparently this cyclohexadiene derivative undergoes auto-reduction to give about equal parts of m-xylene and 1.3-di- methyl-A 8 -cyclohexene. 10 CH 3 CH 3 /BE. / \ H 2 C CH H 2 C CH H 2 C C CH 3 H 2 C C CH 3 \ / ' \ / C C H H The condensation of pseudo-ionone to a and |3-ionone by means of sulfuric acid is supposed to take place through the addition and subse- 10 Wallach, Ann. 395. 74 (1913). CYCLIC NON-BENZENOID HYDROCARBONS 239 quent loss of water. The discovery of this reaction (and the formation of pseudo-ionone from citral and acetone by the action of barium hy- drate or other alkalies) by Tiemann and Kruger, 11 in 1882 marked the beginning of the industrial manufacture of this now well-known "syn- thetic violet" perfume. The two ionones are cyclohexane derivatives (see p. 201). Pinacone condensation may take place intramolecularly to form carbocyclic structures, as for example, the formation of 1 . 2-dimethyl- 1 . 2-dihydroxycycloheptane from diacetylpentane. 12 CH 2 CH 2 COCH 3 CH 2 CH 2 COH . CH 3 CH 2 < - >CH,< CH 2 CH 2 COCH 3 CH 2 CH 2 COH . CH 3 A special synthesis, that of cyclopropane derivatives, has been ef- fected by means of diazomethane or diazoacetic ester by Buchner and Curtius. 13 Thus fumaric ester and diazomethane yield cyclopropane- dicarboxylic ester. CH 2 CH.C0 2 R CH 2 CH.C0 2 R CH.C0 2 R \ +11 -> I -^CH 2 <| = N CH.CO 2 R N CH.OXR CH.C0 2 R v Cyclopentanone has been made by applying the method of con- densing nitriles in the presence of sodium ethylate, a reaction discov- ered by Thorpe. 14 Thus 1 . 4-dicyano- valeric ester condenses to the imino compound. CH 2 CH 2 CN CH 2 CH 2 . CN 2 _CH-CN ( 2 R CH 2 CH = NH On hydrolyzing by means of sulfuric acid and heating the resulting acids the imino group is replaced by oxygen and two molecules of C0 2 are removed, resulting in cyclopentanone. "Ber. SI, 808 (1898). "Kipping & Perkin, J. Chem. Soc. 59, 214 '1891). "Ber. 18, 237 (1885). "J. Chem. Soc. 85, 1726 (1904) ; 91, 578, 1004 (1907). 240 CHEMISTRY OF CH 2 CH C0 2 H C = NH )H 2 CH NON-BENZENOID HYDROCARBONS CH 2 CH, \ CO. CH 2 CH 2 CO + 2 C0 2 H It has been shown by Thorpe 15 that ring closing to form rings of five carbon atoms takes place very rapidly and with approximately equal ease in both the following cases, CH 2 CN C=NH .CH 2 CN XH-CN CH 2 CH 2 CN CH 2 CH 2 CH 2 CH 2 CN / C = NH CH 2 CH CN Kon and Stevenson also find 16 that ring closing by elimination of wa- ter from the COOH group takes place readily forming products of the following type. u R There is no indication of the valency direction being different in any of these examples of ring closing. An instance of the ease with which substances containing a five- carbon ring are formed is the condensation of si/m.-dipropionylethane to l-methyl-5-ethyl-A 5 -cyclopentene-2-one by the action of 10 per cent aqueous caustic potash. 17 " J. Chem. Soc. 93, 165 (1908) ; 95, 1901 (1909). "J. Chem. Soc. 119, 87 (1921). "Blaise, Oompt. rend. 158, 708 (1914). CYCLIC NON-BENZENOID HYDROCARBONS 241 CH 3 CH 2 CO H 2 C CH 3 CH 3 CH 2 C = = C CH 3 \ ( CH 2 CO \ CH CO Acetonylacetone and acetonylacetophenone are unchanged under these conditions. Kishner 18 has discovered that when hydrazine reacts upon un- saturated ketones containing the group CH = CH . CO-pyrazoline bases are formed in many instances, which are readily decomposed to give cyclopropane derivatives. Thus pulegone yields carane: In a similar manner, isobutylidene acetone yields l-methyl-2-iso- propyl-cyclopropane, Pr Pr . CH CH, HC- CH 3 CH, CH, >CH.CH = CH.COCH 3 - HN< . N=C- CH \ +N 2 CH.CH 3 Cinnamic aldehyde yields phenylcyclopropane and phorone yields a dimethylisobutenylcyclopropane. This synthesis, discovered by Kish- ner, is another example, illustrating the remarkable reactivity of the group _ CH = CH CO . As noted above cyclopropane derivatives are formed by the reac- tion of diazoacetic ester and olefine bonds, a reaction employed by Buchner to throw light on the constitution of camphene. 19 J. Ruse. Phys.-Chem. Soc. 40, 987 (1913) ; J. Chem. Soc. Aba. 1913, I, 1163, 1165. Ber. 46, 759 (1913). 242 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS (CHa), H-C0 2 K The formation of a seven carbon ring from a cyclohexane derivative has been noted in the reaction of mesitylene and diazoacetic ester, the intermediate product being smoothly decomposed in the presence of copper powder at 105. This is another illustration of rearrangement which undoubtedly takes place through the intermediate formation of a cyclopropane derivative. 20 CH 3 C=:=CH C CH 3 CH.C0 2 C 2 H mesitylene / + CH = C - CH diazoacetic ester - CH 3 CH 3 C==CH CH CH 3 C.C0 2 C 2 H 5 C Clt CH 3 The formation of cyclic non-benzenoid hydrocarbons by hydrogena- tion of aromatic hydrocarbons is a useful method for the preparation of a limited number of substances and these could very properly be given the appellation hydroaromatic compounds. The hydrogenation of benzene at 180-200 over finely divided nickel was first carried out in 1901 by Sabatier and Senderens, 21 and cyclohexane made in this way has been employed to some extent as a motor fuel for aeroplanes. 22 On hydrogenating naphthalene, tetrahydronaphthalene is the principal product at 180-200, but at 250 and 120 atmospheres pressure deca- hydronaphthalene is formed. Tetrahydronaphthalene has recently become an industrial product, being recommended as a solvent or tur- 20 Buchner, Ber. 53, 865 (1920). "Compt. rend. 132, 210 (1901). 22 CJ. Brit. Pat. 133,288; 133,667 (1919). CYCLIC NON-BENZENOID HYDROCARBONS 243 pentine substitute. 23 The xylenes readily yield the corresponding di- methylcyclohexane, p-cymene is converted into para-menthane, and meta-menthane is easily obtained by the catalytic hydrogenation of sylvestrene. 2 * Indene at 250 and under pressure may be hydrogenated to octahydrindene or bicyclononane. 25 CH 2 CH 2 CH CH 2 [ 2 _CH 2 CH CH 2 Dibenzyl ketone, made from phenylacetic acid, yields dicyclohexyl- propane on hydrogenation by catalytic nickel and hydrogen. Cyclic Non-benzenoid Hydrocarbons. As pointed out in the preface, it is difficult to classify the non- benzenoid hydrocarbons in a way which will not unduly emphasize slight differences in chemical behavior or structure. As regards chemi- cal behavior we should certainly consider cyclopentane with normal pentane and cyclohexane with normal hexane. Also, the amount of information dealing with the derivatives of cyclohexane exceeds the sum total of that dealing with all the other cyclic non-benzenoid hydro- carbons. The reasons for this are the long standing interest in the chemistry of benzene, the conversion of benzene and a few of its de- rivatives to cyclohexane derivatives by hydrogenation, and the avail- ability of material for investigation, as in the case of the terpenes. Practically all of the other cyclic non-benzenoid hydrocarbons have been obtained only by synthesis, only a very few of the simplest of such hydrocarbons having been isolated from petroleum. Although the quantity of information regarding cyclohexane is so relatively large, the use of the term "hydroaromatic" for the cyclohexane series is very unfortunate and will be avoided in the following pages as much as possible. Some writers may consider that cyclopropane may properly be con- sidered together with ethylene and its derivatives but the so-called un- M Cf. Tetralin and Similar Hydrogenated Products, Frydlender, Rev. prod. chim. 23, 719 (1920). 2 Sabatier & Marat, Compt. rend. 156, 184 (1913). "Ipatiev, J. Chem. Soc. Abs. 1913, I, 1165; Osterberg & Kendall, J. Am. Chem. Soc. ifi, 2616 (1920), recommend Ipatiev's method for the preparation of cyclohexane from benzene. The method consists simply in placing the benzene and catalyst in a tight bomb, heating to 250 and passing in hydrogen at 1800 Ibs. pressure from a pressure cylinder. 244 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS saturation of cyclobutane places it close to cyclopropane and in a po- sition intermediate between cyclopentane and cyclopropane. How- ever, some concession must be made to the necessity of some sort of orderly arrangement of subject matter and the writer has elected to discuss the cyclic non-benzoid hydrocarbons in a series beginning with cyclopropane. There are two general classes of information regarding the cyclo- paramnes and particularly the cyclohexane series. On the one hand there is a relatively large amount of information obtained by the in- vestigation of pure substances, either synthesized or isolated from a natural product as is usually the case in the study of the terpenes ; this information is usually accurate and satisfactory, from a scientific point of view. The second type of information is much less definite and less reliable and has to do with very imperfectly known mixtures such as petroleum distillates, shale oils, rosin oils, and similar products whose literature is nevertheless considerable by reason of their com- mercial importance. In dealing with the chemistry of these substances the scientific and industrial works have usually been rigidly exclusive, each of the other class of information. However, the proportion of information of permanent scientific value contributed by the indus- tries is becoming greater than ever before and cannot be passed by, and in the following pages information from industrial sources will be included whenever it is of interest and appears to be of permanent scientific value. Cyclopropanes: Simple cyclopropane hydrocarbons have not been found in nature but the bicyclic terpenes sabinene and carene possess three carbon rings, as does also the ketone thujone. The similarity of the cyclopropane ring to the ethylene bond has repeatedly been pointed out. Its influence upon physical properties is less marked than in the case of the double bond, as has been reviewed in the sec- tion on physical properties. Carr and Burt 26 conclude, from a study of absorption spectra, that the cyclopropane ring is a "center of resid- ual affinity" similar in character but intermediate in quantity to that of the double bond, and as such can form a conjugated system with the carbonyl group. The relative stability of the derivatives of cyclo- propane varies within wide limits, with different substituent groups, as will be brought out in the following pages. Kohler and his students have shown, in a series of papers, that sub- stituents have exactly the same effect upon the mode of addition to a * J. Am. Ghem. Soc. 40 t 1590 (1918). CYCLIC NON-BENZENOID HYDROCARBONS 245 cyclopropane ring as to an ethylene linkage, even though the saturated open-chained compounds formed in the two cases are quite different in structure. 27 Thus, as pointed out by Kohler and Conant, the mode of addition of hydrobromic acid to cyclopropane hydrocarbons is deter- mined by the number and arrangement of the alkyl groups. The ring invariably opens between the carbon atoms that hold the largest and the smallest number of alkyJ groups and the principal product is al- ways one in which the halogen is combined with the carbon atom that holds the largest number of alkyl groups. In the case of cyclopropane carboxylic acids the C0 2 H groups may affect the ease with which addi- tion takes place, but the product is always either a y-bromo acid or the corresponding lactone. The few ketones that have been studied behave like the acids. In cases where a carbonyl group is next to the ring the halogen atom accordingly always goes to the ^-position in the ring, for example, CH 2 | > CH . C0 2 H + HBr > CH 2 Br . CH 2 CH 2 . C0 2 H CH 2 CH 2 > CH . CO . C 6 H 5 + HBr CH 2 Br . Kohler and his assistants find that derivatives of the type C 6 H 5 CH CH.COC 6 H 5 C(C0 2 R) 2 are quite stable to cold permanganate solution and to ozone but are hydrolyzed by water with "unusual rapidity," and, in the absence of water, alcoholates, ammonia and amines rapidly convert them into isomerJc unsaturated compounds. The cyclopropane ring may be broken in different ways depending upon the conditions and the reaction employed. The phenylanisoyl derivative studied by Miss Hahn 28 breaks down in the three ways indicated below, (1) With alkali alcoholates C 6 H 5 CH CH . COC 6 H 4 OCH 3 \ . / > C 6 H 5 CH = C COC H,OCH 8 C(C0 2 CH 3 ) 2 | "Cf. Kohler & Conant, J. Am. Chem. 8oc. 39, 1404 (1917), */. Am. Chem. 8oc. 88, 1-320 (lOlrfi. H(C0 2 CH 3 ) 246 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS (2) When the dibasic acid is heated C0 2 is evolved accompanied by rupture of the ring, RCH CH . COC 6 H 4 OCH 3 \ : / > RCH = C . CH 2 COC 6 H 4 OCH 3 C(C0 2 H) 2 | C0 2 H (3) When the ester in solution in acetic acid is reduced with zinc dust the reduced derivate is obtained RCH CH . COC 6 H 4 COCH 3 \ / RCH.CH 2 COC 6 H 4 OCH 2 C(C0 2 CH 3 ) 2 | CH(C0 2 CH 3 ) 2 A series of cyclopropane derivatives has been made by Bruylants 29 starting with the novel reaction, CH 2 Br CH 2 Br CH 2 < + 2C 2 H 5 MgBr > CH 2 < CH 2 CN CH 2 C C 2 H 5 N.MgBr CH 2 CH CH 2 CH 2 <| CH C C 2 H 5 CH C C 2 H 5 II II NMgBr O Halogen derivatives of the type CH 2 . CH 3 | >CHC< CH 2 | CH 3 are quite stable to boiling aqueous caustic alkali but boiling with alco- holic alkali gives a mixture of the ether and the unsaturated* hydro- carbon. When the unsaturated hydrocarbon is treated with bromine a tribromide is formed, the double bond taking up Br 2 and the tertiary hydrogen atom being replaced without rupture of the cyclopropane ring, CH 2 CH 2 CH 2 CH 2 Br \ // \ / CH C +2Br C C.Br /I \ CH 2 CH 3 CH 2 f CH 3 **Rec. trav. chim. 28, 180 (1909). CYCLIC NON-BENZENOID HYDROCARBONS 247 Kohler has made a number of nitro derivatives of cyclopropane by re- acting upon unsaturated substances with nitromethane, brominating and removing HBr, for example, 30 C 6 H 5 CH = CH . COC (CH S ) ,'+ CH 3 N0 2 ^C 6 H 5 CH CH 2 COC (CH 3 ) . CH 2 N0 2 + Br 2 > C 6 H 5 CH CHBr . COC (CH,) , + CH 3 C0 2 K. ;H 2 N0 2 C 6 H 5 CH CH . COC (CH 3 ) , CH.N0 2 Cyclopropane is reduced to propane by hydrogen and catalytic nickel slowly at 80 and rapidly at 120, but cyclobutane requires a temperature of approximately 180 for hydrogenation to butane. 81 Cyclopropane thus occupies a position intermediate in stability, to hydrogen and nickel, between cyclobutane and ethylene, the latter being reduced to ethane at temperatures as low- as 15. Cyclo- propane is readily reduced to propane by colloidal platinum in acetic acid but cyclopropane- 1 . 1-dicarboxylic acid is not reduced under these conditions. 32 Ethylene is reduced a little more rapidly than cyclopro- pane by this method (Fokin-Willstatter method) . In contact with iron conversion of cyclopropane to propylene 33 can be observed at 100, but in the presence of platinum black the reaction is slow at 200, although rapid at 315. Cyclopropane can be prepared 34 by the reduction of 1.3-dibromo- propane by zinc in alcohol (75 per cent) at temperatures not exceeding 60. It was first made by the action of sodium on this dibromide. It is thus evolved as a gas, easily condensed to a liquid boiling at 35 (749 mm.). Methyl Cyclopropane, 35 boiling-point 4 to 5, is formed when 1.3- dibromobutane is treated with zinc dust in alcohol, in the same manner in which Gustavson prepared cyclopropane from 1 . 3-dibromopropane. 1 .1-Dimethylcyclopropane, 56 boiling-point 21, like other deriva- 10 Kohler & Rao, J. Am. Chem. Soc. 41, 1697 (1919). 11 Willstatter & Bruce, Ber. W, 4459 (1907). 12 Boeseken and others, Rec. trav. chim. 35, 260 (1916). "Ipatiev, Ber. 35, 1057 (1902) ; 86, 2014 (1903). "Gustavson, J. prakt. Chem. (2) 76, 512 (1907). "Demjanoff, Ber. 28, 22 (1895). "Ipatiev & Huhn, Ber. 86, 2014 (1903). 248 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS tives of cyclopropane, reacts only very slowly with permanganate, and may thus be distinguished from the isomer trimethylethylene, the lat- ter hydrocarbon being formed when 1 . 1-dimethylcyclopropane is passed over alumina at 340-345. The 2.3-dicarboxylic acid derivative of 1 . 1-dimethylcyclopropane is of interest as having been produced by Baeyer and Ipatiev 37 by the oxidation of carone and later synthesized by W. H. Perkin, Jr., and Thorpe. 38 It exists in two physically isomeric forms known as cis and frans-caronic acids. It was synthesized by treating the ester of mono- bromo-|3(3-dimethylglutaric acid with alcoholic caustic potash. C(CH 3 ) 2 / \ C(CH 3 ) 2 / \ KOH / \ C 2 H 5 2 C.CHBr CH 2 .C0 2 C 2 H 5 - > KO 2 C . CH CH . CO 2 K Hydrobromic acid at 100 breaks the ring in the following manner: C(CH 3 ) 2 \/ \ Br.C(CH 3 ) 2 /\ \ +HBr | H0 2 C . CH CH . C0 2 H - -- H0 2 C . CH 2 CH C0 2 H The alkali salts of cis and trans-caronic acids are quite stable to aque- ous permanganate. 1 .@.-Dimethylcyclopropane has been made from 2 . 4-dibromopen- 20 tane by Gustavson's method. It boils at 32-33, d -^ 0.7025, d 10 0.6806, n^- 1.3823. 1 .2.3.-Trimethylcyclopropane was made by first synthesizing 3- methylpentanediol-(2.4). This was converted to the corresponding dibromide by heating with hydrobromic acid and the dibromide treated with zinc dust in 80 per cent alcohol, yielding the hydrocarbon, boiling- 22 22 point 65-66,d , 0.6921, n^jyl. 3942. It is quite stable to aqueous permanganate. 1 .1 .2.-Trimethylcyclopropane was made by Kishner 39 from mes- ityl oxide by his hydrazine method. The hydrocarbon boils at 52.8 20 d 0.6949, n D 1.3866. It is easily dissolved by nitric acid (1.52) and reacts with concentrated sulfuric acid giving a mixture of kero- Ber. 29, 2796 (1896). 88 J. Chem. 800. 75 t 48 (1899). 89 J. Ruas. Phys.-Chem. Soc. kk, 165 (1912* CYCLIC NON-BENZEN01D HYDROCARBONS 249 sene-like hydrocarbons boiling mostly within the range 170-360. These higher boiling hydrocarbons are probably formed by the inter- mediate formation of an oleftne followed by polymerization in accord- ance with the general behavior of olefines to concentrated sulfuric acid which has been discussed elsewhere in these pages. With nitric acid in glacial acetic acid hydration occurs, resulting chiefly in isopropyl di- methyl carbinol. It is reduced by Sabatier's method as follows: CH 3 CH. CH 3 CH \ /I I C CH 3 C .... - >CH 3 CH 2 C CH 3 CH Fuming hydroiodic acid and bromine break the cyclopropane ring. Methylisopropylcyclopropane is formed by heating methyl iso- propyl pyrazoline (from isobutylidene- acetone and hydrazine hydrate) to 230 with caustic potash. 40 CH 3 CH 3 >CH CH.NH.N - > >CH CH CH 3 I M CH 3 i \ CH 2 - C-CH 3 \ CH 2 CH.CH 3 20 20 The hydrocarbon boils at 80-81, d-^ 0.7120, n -^- 1.3927. l-Methyl-1.2.-Diethylcyclopropane has been made by Kishner's 20 hydrazine method. It boils at 108-109, d 0.7382, n D 1.4102. It is markedly more stable to permanganate solution than 1.1.2.-tri- methylcyclopropane and is also less reactive to bromine. 41 Methylisobutylcydopropane was made by Zelinsky 42 by hydrogen- ating. the dimethylbicyclohexane, shown below, in the presence of platinum or palladium black, CH, CH CH 2 CH CH 2 CH < CH 2 / 2 ^^f -I~O. ^^/ AJ. O ^-^ - 1 " 1 - ^X \ CH 3 /\ CH 3 C< CH CH CH 2 "C<| CH 3 CH 3 CH CH 2 C< CH 3 CH 2 - > > C <| I CH, CH 3 CH 2 Br CH 3 CH 3 - > > CH . CH 2 CHBrCH 2 C < CH 3 | CH 3 Br A tricarboxylic acid derivative of methyl dicyclobutane has been prepared by Beesley and Thorpe 44 and is mentioned because of its curious structure, -its method of preparation and the fact that it exists in three distinct modifications, in accord with accepted ideas of stereochemistry. When the dibromoethyl ester of the acid CH 3 C. (CH 2 C0 2 H) 3 is treated with pyridine a dilactone ester is formed which readily yields the free acid, CHBr.C0 2 C 2 H 5 CH.C0 2 H. CH 3 C CHBr . C0 2 C 2 H 5 -- > CH 3 C C . C0 2 H. X CH 2 C0 2 C 2 H 5 ^CH.CO.H. Three distinct modifications melting at 193, 165 and 154 were iso- lated, which evidently correspond to the three theoretically possible acids, CH 6 \ \ \, c - c - -c C0 2 H. C0 2 H H C0 2 H H H COJ 48 Kishner, J. Russ. Phys.-Chcm. Soc. 45. 957 (1913). 44 J. Chem. Soc. 117 3 601 (1920). CYCLIC NON-BENZENOID HYDROCARBONS 251 CH 3 c c /\ C0 2 H / \ C0 2 H H C0 2 H H These acids are remarkably stable and are not affected by prolonged boiling with aqueous acids or alkalies. Cydobutane: This hydrocarbon, boiling-point 11-12, D 4 0.7038, is readily made by hydrogenating cyclobutene in the presence of nickel at 100. Hydrogen in the presence of nickel, at 180, con- verts cyclobutane to normal butane. It is stable at ordinary tempera- tures to bromine and hydriodic acid. Its simple derivatives show a striking resemblance in physical and chemical properties to the deriva- CH 2 CHOH tives of n. butane. Thus cyclobutanol I and n. butyl CH 2 CH 2 alcohol are very similar in odor and boiling-point, 123 and 116.8 re- spectively. W. H. Perkin, Jr., 45 who prepared cyclobutanol, stated, "It shows the closest resemblance to the fatty alcohols containing the same number of carbon atoms; it might, indeed, be readily mistaken for nor- mal butyl alcohol." Perkin also found that cyclobutylcarboxylic acid behaves very much like valeric acid, the amide giving excellent yields of the amine, with bromine and caustic potash. CH 2 CH . CONH 2 CH 2 CHNH 2 CH 2 CH 2 CH 2 CH 2 The cyclobutane derivatives all have slightly higher boiling-points than the corresponding normal butane derivatives. Cyclobutyl Series Normal Butyl Substance B.-P. Substance B.-P. Difl. R.CO,H 195 RxCOiH 186 9* R.NH, 81 RxNH, 76 5 R.OH 123 RiOH 116 7 R.C1 85 RxCl 77 8 RBr 104 RiBr 100 4 RI 138 RJ 131 7* 48 J. Chem. 800. 65, 950 (1894). 252 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Willstatter 46 and his co-workers have applied the well-known method of exhaustive methylation and decomposition of the tertiary base, to the preparation of cyclobutene. CH 3 CH 2 CH NH 2 CH 2 CH N CH 3 OH CH 3 CH CH 2 CH 2 CH 2 CH 2 CH I || +N(CH 3 ) 3 + H 2 CH 2 CH Cyclobutene readily adds one molecule of bromine to form the com- paratively stable dibromide boiling-point 171-174. When heated with quinoline this dibromide decomposes with rupture of the ring, giv- ing butadiene but with caustic potash at 200 acetylene is formed. 47 = CH 1 HC == CH CH 2 CHBr /i + (KOH) FCH = | |_CH = CH 2 CHBr \ + quinoline - -* CH 2 = CH - - CH = CH 2 The following methods of rupturing the ring of cyclobutane or its simple derivatives have been observed. (1) CH 2 CH 2 | H 2 + Ni at 180 > CH 3 CH 2 CH 2 CH 3 CH 2 CH 2 (2) CH 2 CHC0 2 .iCa + Ca(OH) 2 CH 2 CH 2 > 2CH 2 = CH 2 + CaC0 3 . + H 2 heat (3) 1 . 2-dibromocy clobutane + quinoline > butadiene. (4) Cyclobutylamine phosphate + heat > butadiene. (5) 1.2-dibromocyclobutane + KOH > acetylene. Gustavson 48 prepared a hydrocarbon C 5 H 8 by the action of zinc in alcohol on C(CH 2 Br) 4 and from the manner of its formation and its physical properties and chemical behavior Gustavson's hydrocarbon has been considered to be spirocyclane, *" B&r. 38, 1992 (1905) ; 40. 3979 (1907). 47 The 1.3-diphenyl derivative of cyclobutadiene is a stable crystalline hydrocarbon melting at 130. [Gastaldi & Cherchi, Ooze. chim. Ital. M (1), 282.] 48 J. prakt. Chem. (2) 5^ 105 (1896) ; 56, 93 (1897). CYCLIC NON-BENZENOID HYDROCARBONS 253 BrCH 2 CH 2 Br CH 2 CH, \ / C > BrCH 2 CH 2 Br However, it has been shown that by careful fractional distillation Gus- tavson's product may be separated into two hydrocarbons, one boiling at 37.5 and the other at 42. Philipow 49 has demonstrated that both hydrocarbons are derivatives of cyclobutane, the lower boiling one yielding levulinic acid on oxidation, CH 3 CH 2 C CH 3 CH 2 C < CH 2 CO CH 3 > OH > CH 2 -CH CH CH.OH. CH 2 C0 2 H methylcyclobutene The hydrocarbon boiling at 42 proved to be methenecyclobutane. Both hydrocarbons yield the same hydroiodide and treatment of this iodide with moist silver oxide yields an alcohol boiling at 116-119. The same alcohol is obtained directly from both hydrocarbons by careful hydration by dilute sulfuric acid, CH 3 CH 2 C CH 3 CH 2 C< + HI I II H 2 CH CH 2 CH 2 \ \ \ CH 3 CH 2 C CH 2 \ /^ CH, C CH 2 CH 2 dil. H 2 S0 2 Previous results on the oxidation of the hydrocarbon boiling at 42 and the alcohol had led to no very definite results, but Philipow showed that the oxidation of the alcohol is strictly analogous to the oxidation of 1 -methy Icy clohexanol ( 1 ) , 49 J, prakt. Chem. (2) 93, 162 (1916). 254 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CH 2 C0 2 H principal CH 3 C0 2 H + CH 2 < / CH 2 C0 2 H reaction \ \ HC0 2 H + cyclohexanone. CH 3 CH 2 C< C0 2 H C0 2 H principal OH. / CH 3 C0 2 H + CH 2 < > | / C0 2 H CO 2 H reaction CH 2 CH 2 \ \ HCOoH + CH 2 CO CH 2 C0 2 H CH, CH 2 CH 2 C0 2 H Hydrogenation had yielded a hydrocarbon C 5 H 10 , supposed, on the basis of the spirocyclane structure, to be ethylcyclopropane. Philipow made ethylcyclopropane 50 by Kishner's admirable method, from acetylcyclopropane, CH 2 CH 2 I >CH.CO.CH 3 + H 2 N.NH 2 -> I >CH.CH 2 CH 3 + N 2 + H 2 O CH 2 CH 2 Perkin and Colman 51 had stated that methylcyclobutane was pro- duced by the action of sodium, in toluene, on 1 . 4-dibromopentane but on repeating their work Philipow obtained a similar product but showed that it was a mixture of hydrocarbons, in which he identified piperylene and n . pentene. Demj anow 52 had made methylcyclobutane by the ac- tion of zinc in acetic acid on cyclobutylmethyl iodide. Philipow made this hydrocarbon in two ways, from cyclobutylaldehyde C 4 H 7 .CHO by Kishner's method, and also by reduction of Gustavson's hydrocar- bons by colloidal palladium (Skita's method), and the hydrocarbon obtained by the three methods proves to be identical, i. e., methyl- cyclobutane, boiling-point 36 -36. 5 (755mm.),d^ 3 0.7118, MR 23.58, MR calc. 23.02. Methylcyclobutane reacts with hydrogen in the pres- ence of catalytic nickel at 205 to give isopentane. M Philipow's ethylcyclopropane boils at 36.5 (755mm.), d "4^0.7055, MR 23.61, MR calculated 23.02. "/. Chem. Soc. 53, 201 (1888). 82 J. Ruse. Phya.-Chem. Soc. &, 842 (1910). CYCLIC NON-BENZENOID HYDROCARBONS 255 18 Cyclobutanone, boiling-point 99-101, d-^^ 0.9344, has an odor lo resembling acetone. Oxidation by nitric acid yields succinic acid. It was made by Kishner from cyclobutanecarboxylic acid by treating with ammonia to form the amide, brominating and then treating with bromine and caustic potash, CH 2 CH 2 CH 2 CH . C0 2 H > CH 2 CHCONH 2 > CH 2 CH 2 CH 2 CH 2 / \ / \ CH 2 C CONH 2 CH 2 CO \ /I > \ / CH 2 Br CH 2 Cyclobutene is of interest on account of the series of bromine substitution products which can be prepared from it without rup- ture of the ring. Thus cyclobutene adds a molecule of bromine to form 1.2-dibromocyclobutane. On treating this with alkali one mole- cule of hydrogen bromide is removed and the resulting bromocy- clobutene, like aliphatic defines containing halogen, is relatively quite stable. It adds HBr to form 1 . 1-dibromocyclobutane which on hydrolyzing with aqueous lead oxide yields cyclobutanone. Bromocyclobutene adds bromine to form 1 . 1 . 2-tribromocyclobutane, which can be converted to CH 2 CBr by loss of HBr, and this in AH.-!!. Br turn adds bromine to give CH 2 CBr 2 melting-point 126 and this CH 2 CBr 2 can be further brominated without breaking the ring to form penta- bromocyclobutane, and hexabromocyclobutane melting at 186.5 . 53 Ethylcyclobutane has been prepared by a very roundabout method from the amide of cyclobutanecarboxylic acid, which was converted into cyclobutylmethyl ketone, this reduced to cyclobutylmethylcar- binol and the latter converted to the corresponding iodide and reduced by zinc dust and acetic acid. 54 The hydrocarbon boils at 72.2-72.5, 83 Willstatter & Bruce, Ber. 40, 3979 (1908) "Zelinsky & Gutt, Ber. 84, 2432 (1908). 256 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS 1ft 20 19 5 0.7540, d- 0.7450 and n -- 1.4080. Oxidation by nitric acid yields succinic acid. Cyclobutylmethyl ketone boils at 136-136.5 (semicarbazone melting at 148) and the cyclobutylmethylcarbinol C 4 H 7 .CHOH.CH 3 , boils at 144 (phenylurethane melting at 87.5- 88). Lebedev 55 has shown that when substituted allenes are polymer- ized, cyclobutane derivatives are formed. When unsymmetrical di- methyl allene is heated in sealed tubes the principal product is 1 .2-di- isopropylidenecyclobutane CH 2 C C (CH 3 ) 2 together with 1 . 1- CH 2 C = C(CH 3 ) 2 dimethyl-2-methylene-3-isopropylidenecyclobutane. The former hy- 20 1Q 7 drocarbon boils at 179-181, d 0.8422, n ^- 1.5008 from which the increment of the molecular refraction is shown to be 2.32. It is readily hydrogenated to 1 . 2-diisopropyl-cyclobutane, boiling at 157- 158.5, d _ 0.7901. The monoozonide yields isopropylidene-2-cyclo- butanone, CH 2 C = boiling-point 171. Isopropyl-2- CH 2 C = C(CH 3 ) 2 cyclobutanone obtained by reduction, boils at 148-150 (semicarba- zone melting at 183). The hydrocarbon 1 .l-dimethyl-8-methylene-8-isopropylidene boils 20 at 149-150, d 0.7982. The corresponding saturated hydrocarbon obtained by reduction, i. e., 1 . 1 . 2-tTimethy\-3-isopropylcyclobutane, 20 boils at 145-146, d 0.7598. The two unsaturated hydrocarbons have a sharp kerosene-like odor. The two saturated hydrocarbons are not attacked by aqueous permanganate. The stability of the satu- rated cyclobutanes to sulfuric acid has not been noted. Cyclobutane-1 .1-Dicarboxylic Acid, melting-point 155, is pre- pared by a general method discovered by Perkin, i. e., the reaction of 1 . 3-dibromopropane and sodium malonic acid ester or sodium cyan- acetic ester. CH 2 Br C0 2 R CH 2 C0 2 H CH 2 < + 2Na + H 2 C< - >CH 2 < >C< CH 2 Br CN CH 2 C0 2 H Decomposition of the dicarboxylic acid yields, 86 J. Ruse. Phys.-Chem. Soc. 48, 820 (1911). CYCLIC NON-BENZEN01D HYDROCARBONS 257 Cyclobutanecarboxylic Acid, boiling-point 194. The acids of this type resemble fatty acids very closely, this acid readily yielding a pleasant smelling ethyl ester boiling at 160, an anhydride boiling at 160, an amide melting at 130 and a nitrile boiling at 150. Hydri- odic acid at 200 breaks the ring forming n. valeric acid. 56 When the silver salt is treated with iodine, a peculiar condensation with forma- tion of the ester of cyclobutanol results, 57 2C 4 H 7 C0 2 Ag + I 2 - -> C 4 H 7 C0 2 . C 4 H 7 + C0 2 + 2AgI Cyclobutane 1 .1 .2.2.-Tetracarboxylic Acid, melting-point 145- 150 is formed by the reaction C0 2 R C0 2 R CH 2 CNa< CH 2 C< C0 2 R I C0 2 R Br 2 C0 2 R > C0 2 R 2 CNa< or I 2 CH 2 C< C0 2 R C0 2 R On heating the free acid it loses two molecules of carbon dioxide and forms cyclobutane-l-2-dicarboxylic acid, melting at 137, known in cis and trans forms. On brominating the 1 . 2-dibromide is formed. By the action of caustic alkalies one molecule of HBr is removed to form the bromocyclobutene carboxylic acid, CH 2 CBr C0 2 H . CH 2 CBr KOH + C0 2 + HBr H 2 CBr C0 2 H . CH 2 C . C0 2 H Silver oxide in water replaces both bromine atoms with hydroxyl, these reactions being quite analogous to the formation of bromofumaric acid and tartaric acid from isodibromosuccinic acid under the same condi- tions, CHBr.C0 2 H. CH C0 2 H CH(OH).C0 2 H > M and | CHBr . C0 2 H CBr C0 2 H CH (OH) . C0 2 H . Cyclobutane -1 .8-Dicarboxylic Acid is known in cis and trans forms, melting at 136 and 171, respectively. Simonsen 5S has shown that when the ethyl ester of p-methoxymethylmalonic acid is digested "Kishner, J. Russ. Phys.-Chem. Soc. 40, 673 (1908). "Demjanov, J. Russ. Phys.-Chem. Soc. JkS, 835 (1911). J. Chem. Soc. 93, 1778 (1908). 258 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS with hydrochloric acid it yields cis-cyclobutane-1.3-dicarboxylic acid. The cyclobutane ring exists in a-truxillic acid, which, according to De Jong, 59 is 1.3-diphenyl-cyclobutane-2.4-dicarboxylic acid. It is formed from cinnamic acid by the action of light and on heating breaks up again into cinnamic acid. Cydopentane: The relationship between cyclopentane and cyclo- hexane, or their derivatives, is exceedingly close, and the increasing number of instances known, in which change of the one ring system into the other occurs, makes evident the rationality and convenience of considering these two ring systems together, rather than isolating the cyclohexane derivatives as "hydroaromatic" compounds, as has usually been done heretofore and thus widely separating the subject matter dealing with these two ring systems. Examples of the con- version of these two ring systems, one into the other, have been noted in the section on Rearrangements. Thus one of the smoothest reactions of this kind is the nearly quantitative conversion of 1 -methyl- 1 -a-hy- droxyethylcyclopentane to l^-dimethyl-A^cyclohexene by zinc chlo- ride. 60 Cyclopentane is also formed when the bromide of cyclobutyl- carbinol is reduced by the zinc-palladium couple and hydrobromic acid. 61 Kishner showed that when benzene is reduced under high pres- sure at 280, according to Wreden, that the product is not cyclohexane but methylcyclopentane 62 and Markownikow 63 has shown several in- stances in which benzene hydrocarbons give cyclopentane derivatives on hydrogenation. Cyclohexanol yields chiefly methylcyclopentane on heating with concentrated hydriodic acid. The hydrocarbons them- selves are quite stable; only in reactions of their derivatives does re- arrangement of the ring structure occur easily. Thus Markownikow and Fortey 64 independently observed that cyclohexane could be heated with hydriodic acid (and red phosphorus) in sealed tubes to 240 with- out change. Methylcyclohexane is, however, partially rearranged by heating with hydriodic acid to 270 to dimethylcyclopentane, and this change is effected without the formation of higher boiling products, in other words, is not a thoroughgoing decomposition such as occurs in "cracking" processes. Methylcyclopentane is one of the products of the action of aluminum chloride on cyclohexane. 65 "Chem. Abs. 1918, 1385; Stoermer & Laage, Ber. 54, 77 (1921). Meerwein, Ann. 417, 255 (1918). Demjanow, Ber. 40, 4960 (1907). 62 J. Rues. Phys.-Chem. Soc. 29, 210 (1897). M Ber. SO, 1214 (1897). "Proc. Ghem. Soc. 1897, 161. 'Aschau, Ann. 32J, 12 (1902). CYCLIC NON-BENZENOID HYDROCARBONS 259 Cyclopentane was prepared by Wislicenus 66 from cyclopentanone, the latter being prepared by the well-known method of heating calcium adipate. Cyclopentanone is also a constituent of the oily residues re- covered in the rectification of wood alcohol. The ketone on reduction under the same conditions usually applied to ordinary aliphatic ke- tones, for example, reduction by means of sodium in moist ether, yields cyclopentanol. The alcohol has an odor resembling amyl alcohol, boils 21 5 at 139, d ' 0.9395. Cyclopentanol is converted into the corre- sponding iodide by saturating with hydrogen iodide and hydrogen bro- mide yields the bromide, without rupture of the ring. Reduction of the iodide under the usual conditions, zinc and hydrochloric acid in 20 5 dilute alcohol, yields cydopentane, boiling-point 50.5-50.7, d ' 0.7506. Cyclopentane is inert to bromine in the dark but in sunlight sub- stitution with evolution of HBr occurs, approximately with the same ease as in the case of normal pentane. On heating with bromine in a sealed tube the reaction is very slow at 100 but more rapid at 128- 130, the reaction then being accompanied by deposition of carbon. Cyclopentane has not been sulfonated, the hydrocarbon being quite stable to sulfuric acid. Borsche 6r has prepared Cyclopentane sulfonic acid by an indirect method involving the conversion of cyclopentanol to the bromide, reacting on the bromide with magnesium in ether and treating the magnesium complex C 5 H 9 MgBr with S0 2 and then oxidiz- ing with aqueous permanganate. The potassium cyclopentyl sulfonate was crystallized from absolute alcohol. Salts of methylcyclohexane-3- sulfonate were prepared in the same manner from 1 methyl-3-bromo- cyclohexane. Cyclopentene is readily formed on warming cyclopentyl iodide with alcoholic caustic potash, closely resembling amyl iodide and its con- version to amylene under the same conditions. Cyclopentene boils at 46. When cyclopentyl bromide is employed a small proportion of cyclopentyl ethyl ether is also formed, again paralleling the n.amyl derivatives. From Cyclopentene Meiser 68 prepared the dibromide, which he converted to the 1 . 2-gly col by hydroly zing with aqueous po- tassium carbonate; the glycol was converted to the chlorohydrin by "Ann. 275, 327 (1893). 87 Ber. W, 2220 (1907). Borsche prepared l-niethyl-cyclohexane-3-sulfone-chloride, which on reduction yields the 1-methyl cyclohexane-thiol (3), boiling at 172, the first of the cyclic mercaptans to be synthesized. M Ber. S2, 2050 (1899). 260 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS hydrochloric acid and the same product was made by the addition of hypochlorous acid to cyclopentene. In the cyclopentane series a very large number of derivatives are known, but the great majority of them have been synthesized by costly and usually very roundabout methods. The hydrocarbons them- selves have 'seldom been used for preparing derivatives; in fact, the hydrocarbons have been prepared from the derivatives. Cyclopentane or its alkyl derivatives have never been isolated in a pure state from petroleum or any other natural product. Cyclopentadiene, boiling-point 41, may be isolated from the fore- runnings when crude benzene is distilled, 69 and Etard and Lambert 70 found it among the products of the thermal decomposition of heavy paraffine oil. It polymerizes spontaneously to the dimeride C 10 H 12 on standing at ordinary temperatures and on distillation the dimeride is partially inverted to the original hydrocarbon. The dimeride (un- known constitution) boils at 170.- Stobbe 71 finds that the spon- taneous conversion to the dimeride is complete in about 30 days and when exposed to oxygen or air a diperoxide of the dimeride is formed, which Stobbe regards as having the following structure, / CH CH CH CH \ CH CH CH CH \/ \/ CH 2 CH 2 The dimeride is much more stable than the original hydrocarbon but may be further polymerized by heating to 160-180 in a sealed tube, a solid resin being formed. 72 The polymers of acyclic olefines and dienes are also more stable than the original hydrocarbons, the difference being marked in their behavior to concentrated sulfuric acid, hydrogen chloride, hydrogen bromide, etc. Hydrogen chloride com- bines with cyclopentadiene to form a monochlorocyclopentene, boiling- point 50 (40mm.), and this derivative, though not further acted upon by hydrogen chloride, combines readily with chlorine to form a trichlorocyclopentane boiling at 196. Addition of bromine to cyclo- pentadiene gives two stereoisomeric 1.4-dibromides, one a liquid and "Kraemer & Spilker, Ber. 29, 552 (1896). Compt. rend. 112, 945 (1891). 71 Ber. 52, 1436 (1919). "Kronstein, Ber. 35, 4150 (1902). CYCLIC NON-BENZEN01D HYDROCARBONS 261 one a crystalline solid ; the dibromides yield two stereo-isomeric aa-di- bromoglutaric acids on oxidation. Cyclopentadiene, like isoprene, com- bines with quinones to give stable crystalline compounds ; for example, with benzoquinone to form the product C^H^C^, melting-point 78. Cyclopentadiene reacts violently with concentrated sulfuric acid and dilute sulfuric acid resinifies it. Like cyclohexadiene, its polymers do not resemble caoutchouc but are resinous. Cyclopentadiene is of special interest on account of the reactivity of the CH 2 group. The hydrocarbon reacts with potassium with evolu- tion of hydrogen, forms C 5 H 5 MgI from CH 3 MgI with evolution of me- thane 73 and readily condenses with aldehydes and ketones under the influence of sodium ethylate. Thiele 7 * attributes this reactivity to the unsaturated character of the contiguous groups, its condensation with aldehydes and ketones paralleling the reactivity of substances contain- ing the group = C CH 2 C = O with these reagents under the same conditions. With acetone, acetophenone, and benzophenone the following intensely colored hydrocarbons are formed: CH = CH CH 3 > C = C < dimethy Ifulvene CH = CH CH 3 CH := CH CH 3 >C = C< methylpheny Ifulvene CH = CH C 6 H 5 CH = CH C 6 H 5 dipheny Ifulvene CH = CH C 6 H 5 Courtot 75 has pointed out the similarity in chemical behavior of the CH 2 in Cyclopentadiene with the corresponding group in fluorene and indene, which hydrocarbons are also colored. The fulvene derivatives, discovered by Thiele, polymerize on warming and absorb oxygen, by autoxidation much more rapidly than Cyclopentadiene. 76 Stobbe and Diinnhaupt 77 have shown that Cyclopentadiene polymerizes very slowly in the absence of oxygen and, unlike styrol, the polymerization is but very slightly affected by light. 4-Methyl-2-Ethylcyclopentadiene: When the ethyl ester of levu- Grignard & Courtot, Compt. rend. 158, 1763 (1914) ; Courtot. Ann. chim. \, 5> (1915) . Ber. S3, 666 (1900) ; 54, 68 (1901). "Loc. cit. "Engler & Frankenstein, Ber. 84, 2933 (1901). " Ber. 52, 1436 (1919). 262 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS linic acid in alcohol solution is condensed by sodium ethylate, an un- saturated cyclic dicarboxylic ester is formed which may have either of the two following structures, I. CH 3 C CH 2 C C0 2 H CH - - C CH 2 CH 2 . C0 2 H 4-methylcydopentadiene-l-carboxy--propionic acid II. CH 3 C CH 2 C C0 2 R R0 2 C CH 2 C C CH 3 2 . 4-dimethylcyclopentadiene-l -carboxy-3-acetic acid The discoverers of the condensation of levulinic ester favor I as being the structure of the reaction product. The free acid melts at 218, with evolution of carbon dioxide and formation of the hydrocarbon, the structure of which, if the above structure I proves to be correct, is 4-methyl-2-ethylcyclopentadiene. The hydrocarbon boils at 135, but on distilling at ordinary pressure about one-third is polymerized, the tendency to polymerize evidently being abnormally great. 78 Methylcyclopentane has been made synthetically by a number of methods and has been shown to be present in the light distillate from Russian petroleum. Methyl-cyclopentane-2-one is formed by heating the calcium salt of p-methyladipic acid. The ketone, boiling-point 143.5, may be purified by the sodium bisulfite compound, then reduced to methylcyclopentanol- (2) , boiling-point 150.5-151, and the latter reduced, by heating with concentrated hydriodic acid, to methylcyclopentane. 80 When made by reducing the iodide, l-methyl-2- iodo-cyclopentane, by the copper zinc couple the hydrocarbon showed the following physical properties, 81 boiling-point 71-72, d 0.7664. It has an odor like well-refined gasoline. A mixture of concentrated sulfuric and nitric acid has little effect on it but fuming nitric acid alone reacts rather violently, acetic acid, carbon dioxide and water being the chief reaction products. Nitric acid Sp. Gr. 1.075, at 115- 120 gives chiefly the tertiary nitro derivative. According to Namet- kin, 82 2-nitro-l-methylcyclopentane is also formed, boiling-point 98- 22 99 at 40 mm., d -JQ- 1.0381, and succinic and a-methylglutaric acids "Duden & Freydag, Ber. S6, 944 (1903). 80 Konowalow, J. Russ. Phya.-Chem. Soc. 8, 125. 81 Markownikow, Ber. SO, 1222 (1897). n J. Ruas. Phya.-Chem. Soc. 43, 1603 (1911). CYCLIC NON-BENZENOID HYDROCARBONS 263 are also formed. The tertiary nitro derivative can be isolated from the secondary nitro derivatives by dissolving the latter in aqueous alkali. According to Markownikow 83 tertiary nitromethylcyclopentane boils at 92 (40 mm.) or at 177 at atmospheric pressure, with considerable decomposition. Both nitro derivatives give good yields of the corre- sponding amines when reduced by tin and hydrochloric acid. The tertiary amine may be converted into the corresponding tertiary alco- hol by nitrous acid and after distilling, boiling-point 135-136, solidi- fies to crystals melting at 30. Chlorine reacts energetically with methylcyclopentane at ordinary temperatures in diffused daylight. 84 The tertiary chloride, prepared from the tertiary alcohol, is unstable, partially decomposing on distil- lation, boiling-point 123. By direct chlorination of methylcyclopen- tane, derived from petroleum, Markownikow obtained a mixture of chlorides from which he was unable to isolate any definite product, the presence of cyclohexane in the original methylcyclopentane adding to the difficulty. Methylcyclopentane has been made by means of the Grignard re- action, ring closing being brought about by treating 5-acetylbutyl- iodide with magnesium in ether, CH 2 CH 2 CO CH 3 CH 2 CH 2 CH 3 +Mg | >C< CH 2 I > CH 2 CH 2 OMgl CH CH 2 CH 3 I >C< CH 2 CH 2 OH the alcohol being converted into the iodide and the latter reduced by zinc dust and acetic acid. 85 Cyclopentanone. This ketone has usually been prepared by ring closing of the ethyl ester of adipic acid by means of sodium. The resulting ester may be regarded as a carbocyclic derivative of aceto- acetic ester and by the general method of decomposing such esters to ketones, this cyclic ester yields cyclopentanone, CH 2 CH 2 C0 2 R +Na CH 2 CH C0 2 R | I >CO > CH 2 CH 2 C0 2 R CH 2 CH 2 83 Ann. 307, 355 (1899). 84 Markownikow, loc. cit. 85 Zelinsky & Moser, Ber. 35, 2684 (1902). 264 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CH 2 CH 2 > | >CO. CH 2 CH 2 Thorpe and Best 88 have described a series of derivatives of cyclo- pentanone which are quite stable to acids but are decomposed by alkali with rupture of the ring, the ring being stable to acids as long .as a CN or C0 2 C 2 H 5 group is present adjacent to the ketone group. The cor- responding imino derivatives are exceedingly stable to alkaline hydro- lyzing agents. The derivative 2-cyanocyclopentane-l-one is a com- pound which resembles ethyl cyanoacetate in many of its properties; thus when treated with alcoholic sodium ethoxide it yields a sodium derivative which on treating with methyl iodide yields 2-cyano-2- methyl-cyclopentane-1-one. These derivatives will not be described in any detail but are mentioned since they show the great similarity in the chemistry of the open chain and cyclopentane series, and also since Best and Thorpe showed that the CN group could readily be removed by heating with dilute sulfuric acid yielding derivatives of cyclopen- tanone. CN CN CN CH 2 - CH CH 2 C . Na CH 2 C CO + NaOC 2 H 5 CO or H 9 CH, CONa / H 2 CH 2 CH 2 CH 2 CH 2 - CH CN + RI CH 2 C R CH 2 CH \ CO + dil. acid 2 CH 2 alkyl cyclopentanones. Thorpe and Best also made 2.5-dimethylcyclopentane-l-one, boiling- point 149, and 2-ethyl-cyclopentanone, boiling-point 149, and 2-methylcyclopentanone by similar methods. Cyclopentanone is formed during the carbonization of wood. It boils at 129, d 20 0.948, n 1.4366. 87 Acetic anhydride enolizes it D M J. Chem. Soc. 95, 690 (1909). T Wallach, Ann. 353, 330. CYCLIC NON-BENZENOID HYDROCARBONS 265 to form cyclopentenol acetate. The semicarbazone melts at 205-207 when slowly heated, or at 212-213 when heated rapidly. It con- denses with formic acid ester to form oxymethylenecyclopentanone C 5 H 6 : CH(OH), melting-point 72-73. A cyclopentanonesulfonal is known melting at 127-128. CH 2 CH 2 \ C(S0 2 C 2 H 5 ) 2 CH 2 CH 2 It condenses readily with aldehydes to form derivatives of the general type 88 HC.R CH 2 C CO 2 C v, The condensation product of the above type formed with benzaldehyde melts at 191, with anisaldehyde 215, cinnamic aldehyde 222.5, pipe- ronal 257 and cuminol 145.5. Acetone condenses with cyclopenta- none 89 to form the isopropylidene derivative, CH 3 CH 2 C = C ' CO CH 8 a liquid very soluble in water, boiling at 195-199. Another re- action which has been useful in the synthesis of hydrocarbons derived from cyclopentanone is the condensation with bromoacetic ester and zinc, according to Reformatsky's method, to give cyclopentanolacetic ester, the free oxy acid decomposing on heating to give methenecyclo- pentane, boiling-point 78-81. "Vorlander, Ber. 89, 1838 (1896) ; Hobohm & Menzel, Ber. S6, 1499 (1903) ; Wal- lach, Goett. Nachr. 1907, 404. 88 Wallach, Ann. 894, 368. Wallach, Ann. S47 t 325. 266 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS GEL CH CH 2 CH 2 CH 2 CH 2 OH >CO > | >C< CH 2 CH 2 CH 2 C0 2 H CH 2 CH 2 | >C = CH 2 + H 2 + C0 2 CR CR, Methenecyclopentane has a penetrating rather disagreeable odor, yields the glycol C 6 H 10 (OH) 2 on oxidation by permanganate and, fol- lowing the general behavior of glycols, this glycol is converted by dilute acids to cyclopentane aldehyde. In the same manner a-bromopro- pionic acid yields the oxy acid from which ethylidenecyclopentane may be prepared. This hydrocarbon, boiling-point 114, behaves like the terpenes in many of the characteristic reactions. Thus it yields a ni- troso chloride, which on treating with alkali loses HC1 and on hydrolyz- ing the resulting oxime, A 1 -acetylcyclopentane, is formed, 91 CH 2 CH 2 CH, GIL CR, CR CR CH. >C-C-CH 8 CH 2 CH 2 | .OH CH 2 CH 2 CR CH C C CH 3 N.OH \ C _ C CH 3 II CH 2 CH Cyclopentanone reacts normally with the Grignard reagent, giving l-methylcyclopentanol(l) with methyl-magnesium iodide, 92 or the ethyl derivative with ethyl-magnesium iodide. 93 These tertiary alco- hols readily decompose to give alkyl cyclopentenes >C< OH CH 2 CH 2 I CH CH boiling-point 135 melting-point 35-37' CR CH CH, CR \ ( / C.CH, 91 Wallach, Ann. 365, 274. B2 Zelinsky & Namjetkin, Ber. S5, 2683 (1902). M Wallach, Ann. 365, 276. CYCLIC NON-BENZENOID HYDROCARBONS 267 CH 2 CH 2 OH I >c< CH 2 CH 2 C 2 boiling-point 155-157' d 21 . 5 0.916 CH, CH \ C.C 2 H 5 CH 2 CH 2 boiling-point 108 d 20 0.7915 By condensing with bromoisobutyric acid and decomposition of the re- sulting oxy acid, isopropylidenecyclopentane is formed, which, like ter- pinolene and other hydrocarbons having a semicyclic double bond of this nature, is converted to isopropylcyclopentene by alcoholic sul- furic acid, the double bond shifting to the ring, CH 9 CH, CH 3 \ CH 2 CH CH 3 \-c/ /H 2 CH 2 CH 2 CH 3 boiling-point 136-137 d 20 0.817 The reactivity of cyclopentanone is also shown by its condensation on treating with sodium ethylate or with hydrogen chloride to dicyclo- pentenepentanone, which may be reduced first by hydrogen and palla- dium and then by sodium to the saturated alcohol. Heating the alco- hol with zinc chloride yields cyclopentylcyclopentene, boiling-point 197-198. CH CH CH /u, >- CH < ] CH, -CH, & H H. 268 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS By the condensation of three molecules of cyclopentanone a ketone C 15 H 20 is formed, which on hydrogenation, as indicated above, yields the saturated ketone dicyclopentylcyclopentanone, 94 C 15 H 20 0, the cor- responding alcohol readily yielding dicyclopentylcyclopentene, !H 2 boiling-point 290 at 760 mm., d 20 0.939 This hydrocarbon may be regarded as a tricyclic "sesquiterpene." 1 .2-Methylcyclopentanone CH 8 / \ H 2 C C = boiling-point 140-141 ( H 9 C C d 20 0.917 H, This ketone can be prepared from camphor-phorone, or from a-methyl adipic acid. It does not condense with aldehydes in the presence of caustic soda. 1 .S-Methylcyclop&ntanone, boiling-point 144-145, d 22 0.913, can be prepared from (3-methyladipic acid. It condenses readily with alde- hydes in the presence of caustic soda, the benzaldehyde compound melting at 149-151 (inactive form). When prepared from optically active (3-methyladipic the 1.3-methylcyclopentanone is also active. 95 [a] -f 124-133. When reduced to the alcohol and then converted D to the iodide, the latter yields l-methyl-A 2 -cyclopentene when treated with alcoholic caustic potash. 96 "Wallach, Ann. 389, 182. M Wallach, Ann. SS2, 349 ; S9L, 371. "Zelinsky, Ber. S5, 2488 (1902). CYCLIC NON-BENZENOID HYDROCARBONS CH, 269 HC H 9 C CH CH boiling-point 69 0.7663 a] D + 59.07 l-methyl-A 2 -cyclopentene 1 . 3-Methy Icy clopentanone reacts with methyl-magnesium iodide to give 1.3-dimethylcyclopentanol(3), boiling-point 143-145. Accord- ing to Zelinsky this tertiary alcohol is decomposed by oxalic acid mainly to l-methyl-3-methenecyclopentane, 97 CH 3 CH CH 2 CH CH boiling-point 93 ( 1Q d-- 0.7734 Potassium permanganate solution oxidizes l-methyl-3-methene- cyclopentane to the glycol and 1 . 3-methylcyclopentanone. 1. 3-Methy Icyclopentanone condenses with acetone 98 to form 4-iso- propylidene-l-methylcyclopentane-3-one, which can be employed for the synthesis of a series of hydrocarbons containing the methyl and isopropyl groups in the 1.4 position. CH S CH / \ CH 3 CH >co + CH, H = CH, CH, >C ii 2 C CH 2 = c c = o " Ber. Sit, 3950 (1901). M Wallach, Ann. 394, 372. boiling-point 203-205' d 0.9315 270 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The ring may be broken by the rearrangement of the oxime, by sul- furic acid, to the isoxime and hydrolyzing the latter by boiling with hydrochloric acid to amidocapronic acid," a reaction which is very generally applicable to the cyclic ketones. CH, H 2 C CH 2 1 .1-Dimethylcyclopentane: Kishner 10 finds that all of the re- actions of dimethylcyclobutylcarbinol, which were studied by him, are abnormal in that the cyclobutane ring is changed to the cyclopentane ring. The carbinol may be prepared by the Grignard reaction ap- plied to the ester of cyclobutanecarboxylic acid, CH 2 +2CH 3 MgI CH, CH 3 CH 2 < >CH.C0 2 R >CH 2 < ">CH C< CH 2 CH 2 I CH< AH When this carbinol is treated with hydrogen bromide the product formed is 2-bromo-l . 1-dimethylcyclopentane, which on treating with alcoholic caustic potash yields the A 2 unsaturated hydrocarbon. CH 2 / \ CH 3 CH 2 CH C< o,l CH ' HBr CH 2 < CHBr CH 9 - CH 8 C< I CH 3 CH 9 Wallach, Ann. 312, 184 J. Rues. Phys.-Chem. Soc. 1,0, 994 (1908). CYCLIC NON-BENZENOID HYDROCARBONS 271 CH, CH C< // I CH 3 > CH boiling-point 78-78.5 \ 9O CH 2 -6H 2 d lF- 0.7580 20 n ^ 1.4190 It has an odor resembling naphthalene. On oxidation by nitric acid it yields aa-dimethylglutaric acid. According to Kishner, 101 both hydrobromic and hydriodic acids reacting on the above carbinol yield halogen derivatives, which on treating with alcoholic caustic potash yield 1.1. -dimethy l-A 2 -cyclopentene together with the isomeric hydro- carbon 1.2-dimethyl-A 1 -cyclopentene. When the bromide, obtained from the carbinol, is reduced by the copper-zinc couple, a saturated hydrocarbon is formed which Kishner 102 regards as 1 . 1 Dimethylcy- 20 clopentane, boiling-point 88.3-88.5, d -^- 0.7553. 1 .2-Dimethylcyclopentane, obtained by reduction of 1 . 2-dimethyl- 90 A 2 -cyclopentene by Sabatier's method, boils at 92.7-93, d^ 90 0.7534, n^- 1.4126. 1 .^-Dimethyl-^-Cyclopentene, one of the products obtained by the decomposition of dimethylcyclobutylcarbinol, as described above, is identical with the hydrocarbon made by Maquenne 103 from per- seitol. On reduction by means of concentrated hydriodic acid it is converted to a hydrocarbon C-H 17 , which Aschan 104 regarded as 1.3- dimethylcyclopentane, but Kishner 105 states that more probably it is 1.2-dimethylcyclopentane together with a little methylcyclohexane. The olefine reacts with hydrobromic acid to form an unstable bromide, yields a nitrosochloride melting at 73-75, and on oxidation yields y-acetobutyric acid. CH 2 C CH 3 CH 2 COCH 3 CH, / j CH / \ CH 2 C CH, CH 9 CO,H 3 vyxj. 2 ^W 2 101 J. Ruse. PJiys.-CJiem. Soc. W, 994 (1908). 102 J. Russ. Phys.-Chem. Soc. 97, 509 (1905). 108 J. Chem. Soc. A&*. 1S9S (1), 635. 104 Chemie d. Alicyclischen Verb, p. 473. 108 J. Ruse. Phys.-Chem. Soc. 40, 994 (1908). 272 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS 1 .8-Dimethylcyclopentane was synthesized in an optically active form by Zelinksy 106 from d-1.3-dimethylcyclopentanol(3) by con- verting the alcohol to the iodide and reducing the latter by the well- known method of treating with zinc dust and acetic acid. The hydro- 16 carbon boils at 90.5-91, d ^0.7497, [1~~ ~\ tene | / CHs 108. 0.782 1.431* 1 ^.S-trimethyl-A^yclopen- tene *Auwers, Ann. 415, 110 (1918). > CH, 119. 0.796 1.442* 20 ** The refractive indices marked with an asterisk are for n a. 274 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS C 2 H 6 - 1 l-diethyl-A 2 -cyclopentene C 2 H 5 ' 144. 0.808 1.446 C.H, 1 ^-diethyl-A'-cyclopentene 151.5 0.812 1.452 cyclopentene CH 8 140. 0.803 1.447* l ' 2 tfn t e riethyl " Al " CyCl Pen ' ' ^-C 2 H 5 181.5 0.814 1.451* Naphthenic Acids: These acids probably occur in more petroleums than is commonly supposed, although they are usually associated with Russian petroleum. The naphthenic acids occurring in Russian petro- leum include relatively simple low-boiling acids so that the mixture of acids, as they are obtained by treating with aqueous alkali and precipitating with acid, possesses a marked, very persistent and rather disagreeable odor. The Gulf Coast petroleums also contain organic acids but they are of high boiling-point and, when separated from lu- bricating oil, are practically odorless and are easily salted out of their solutions in aqueous solutions. The acids in the Gulf Coast oils have never been studied to the extent of determining their nature, but their alkali solutions have pronounced emulsifying and foam producing power and may accordingly find employment in compounding cutting oils or emulsions but can hardly be employed in soaps on account of the ease with which their alkali salts, or soaps, are salted out. Markownikow 108 fractioned the methyl esters of the Russian naph- thenic acids. The fraction distilling at 160-165 was essentially CTHjACHg. The purified free acid distilled at 213-214 and the 20 purified methyl ester boiled at 164-166, d -^- 0.90509. The amide Ann. Sff! t 369 (1899). CYCLIC NON-BENZEN01D HYDROCARBONS 275 melts at 121-123.5 and by converting the amide to the amine, a secondary amine was formed which is perhaps identical with the sec- ondary amine resulting from the reduction of secondary nitromethyl- cyclopentane. The acid methylcyclopentane-2-carboxylic acid CH 2 CH.CH 3 CH 2 CH 2 CH.C0 2 H isolated by Perkin and Freer, 109 boils at 219-219.5 and an isomeric acid, probably the 1.3 acid, prepared by Euler, 110 boils at 220. The aldehyde derivatives of methylcyclopentane have been observed by Markownikow and the two synthetic acids has not been satisfactorily explained. The starting point in the researches of Perkin and Freer lxl was the condensation of sodio-malonic ester with 1.4-dibromopentane, CH 2 CH CH 3 C0 2 R / \ / CH, Br + 2Na + CH 2 CH 2 Br C0 2 R C0 2 R On heating the free dicarboxylic acid a few degrees above its melting- point it decomposes to carbon dioxide and l-methylcyclopentane-2-car- boxylic acid, boiling-point 219.5-220.5. These carboxylic acid deri- vatives of cyclopentane are of interest since the simpler naphthenic acids of Russian petroleum are evidently derivatives of cyclopentane. The acid named above has a most disagreeable odor, somewhat resem- bling valeric acid. It is not acted upon by bromine at ordinary tem- peratures but at 100 rapid substitution occurs with evolution of hy- drogen bromide. Zelinsky 112 has applied the Grignard reaction to the preparation of naphthenic acids but the yields are usually very poor. From 1-methyl- 109 J. Chem. Soc. 53, 199 (1888). 110 Ber. 28, 2952. 111 J. Chem. Soc. 53, 195 (1888). 112 Ber. 35, 2687 (1902). 276 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS 3-bromo-cyclopentane the acid l-methylcyclopentane-3-carboxylic acid was prepared by passing carbon dioxide into the ethereal solution of the magnesium derivative. The acid distills at 115-116 (15 mm.), ooo d 3S- 1.006; the amide melts at 149-150. This acid is possibly iden- tical with the methylcyclopentanecarboxylic acid described by Euler, referred to above. Rearrangements of the iodohydrins of the methylcyclohexenes to aldehyde derivatives of methylcyclopentane have- been lobserved by Tiffeneau. 113 Thus the iodohydrin of A 3 -methylcyclohexene, on treat- ing with silver nitrate is converted into the aldehyde, which on oxi- dation yields the corresponding acid, previously obtained by Zelinsky, 9 H 3 H OH CHO cqn The iodohydrin of cyclopentene does not rearrange but gives the 1 . 2- oxide. In the case of the phenylcyclohexane derivative 114 or substi- tuted phenyl derivatives, rearrangement to the cyclopentane ring does not take place but the phenyl group migrates to the a-position with the formation of a cyclohexenol, which is converted to the isomeric ketone. Cydopentane-1 .2-Dicarboxylic Acid has been prepared from 1.3- dibromopropane and sodium-malonic ester and also by the action of iodine on the disodium derivative of the ester, CH / \ CH 2 C . Na (C0 2 R) CH 2 C.Na(C0 2 R) / \ CH 2 C (C0 2 R) CH 2 C(C0 2 R) and decomposing the tetracarboxylic acid by heating, in the usual manner, 118 Compt. rend. 159, 771 (1914). 114 Le Brazidec, Compt. rend. 159, 774 (1914). CYCLIC NON-BENZENOID HYDROCARBONS C0H. 277 CH CH 2 C< / I C0 2 H CH \ C0 2 H C0 2 H / \ CH 2 CH C0 2 H CH CH C0 2 H It is known in cis and trans forms, the cis form readily forming an anhydride. l-Methylcyclopentane-2.3-Dicarboxylic Acid, melting-point 99- 104, has also been prepared by one of Perkin's methods, i. e., condens- ing 1 . 3-dibromobutane with the disodium derivative of the ethyl ester of ethane tetracarboxylic acid, followed by decomposition of the tetra- carboxylic acid in the usual manner. 115 The cis and trans modifications of cyclopentane 1 . 2 . 4-tricarboxy- lic acid are known, and are best prepared by the reaction of ethyl a(3-di- bromo-propionate on the disodium derivative of ethyl propane-ct a y y~ tetracarboxylate. 116 C(Na)< CH 2 \ C(Na)< C0 2 R C0 2 R C0 2 R C0R BrCH 2 rCH. B C(C0 2 R) 2 CH 2 - CH 2 < | C0 2 R C(C0 2 R) 2 CHCO^R CCLH ie CH, CH. / \ CH CH.CO 2 H . 2 H. vyj.0. CO, The trans form, melting-point 129-130, yields the anhydride of the cis form when heated with acetic anhydride and the anhydride then may be hydrolyzed to the pure cis form, melting at 146-148. ""Fargher, J. Ohem. Soc. U7, 1355 (1920). "Perkin & Goldsworthy, J. Chem. Soc. 105, 2666 (1914). Chapter VIII. The Cyclic Non-ben- zenoid Hydrocarbons. The Cyclohexane Series. The conception of cyclohexane and its derivatives as "hydroaro- matic" compounds has served a useful purpose in connection with the study of the constitution of benzene. Reduction of ortho, meta and para derivatives of benzene yield the corresponding derivatives of cyclohexane which would not be expected from Ladenburg's prism formula. It has also been shown that tetrahydrobenzene and dihydro- benzene do not have the bridged structures shown below, H H 2 C C / \ / \ H 2 C CH 2 HC CH Hr~\ OTT TTr^ OTT Z \U Orlo J3.VX L/Jl \/ . v C H H and Baeyer has shown that the tetrahydroterephthalic acids do not possess bridged ring structures but contain double bonds, the addition products indicating that the two isomeric acids have double bonds in the two positions shown below, and 278 THE CYCLOHEXANE SERIES 279 It should be borne in mind that derivatives of hydrocarbons such as cyclohexane are capable of exhibiting stereoisomerism when two or more substituents are present. Thus if the six carbon atoms are conceived to lie in one plane, as in the plane of the paper, then six of the hydrogen atoms in cyclohexane will lie above the plane and six below the plane. Cyclohexanecarboxylic acid can obviously exist in only one form but a cyclohexane dicarboxylic acid can exist in two forms. Thus the 1.4-dicarboxylic acid derived from terephthalic acid, studied by Baeyer, can exist in two stereo isomeric forms, H HO^C 2 H 'H Baeyer likened these stereoisomers to fumaric and male'ic acids, con- sidering the double bond and the ring structure as preventing free rotation in much the same manner, H C0 2 H o H C0 2 H male'ic acid H0 2 C H \ / C /\ H maleinoid cyclohexane-1 .4- dicarboxylic acid H C \ C0 2 H H fumaric acid fumdroid cyclohexane-1 .4- dicarboxylic acid 280 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Cyclohexane and its alkyl derivatives probably occur in more petro- leums than is generally known but this hydrocarbon and its mono- methyl and dimethyl derivatives were long ago recognized as impor- tant constituents of the light naphtha from Russian petroleum, hence the term "naphthenes" suggested by Markownikow. Careful examina- tion of the lighter distillates of the petroleums from southern Cali- fornia, Mexico and the southern Texas and Louisiana fields will un- doubtedly show the presence of cyclohexane and its simpler alkyl de- rivatives. * Quite a variety of methods have been employed for the preparation of cyclohexane and its derivatives but the methods of cat- alytic hydrogenation of Sabatier and Senderens Ipatiev and Skita are so far superior to most of the others that the latter are generally only of historical interest. A number of syntheses of cyclohexane and its derivative will be briefly mentioned, as follows: (1) Treatment of 1 . 6-dibromohexane with sodium. 1 (2) Reaction of 1 . 5-dibromopentane, malonic ester and sodium ethylate, yielding cyclohexane-1 . 1-dicarboxylic acid, which in turn yields cyclohexanecarboxylic acid by decomposition. (3) Heating the calcium salt of pentane-1 . 5-dicarboxylic acid, forming cyclohexanone. 2 (4) Condensation of the ethyl ester of pentane-1. 5-dicarboxylic acid to form the ester of cyclohexanone-2-carboxylic acid. (5) Condensation of two molecules of succinic ester 3 to the 2.5- dicarboxylic acid derivative of the 1.4-cyclohexane-dione; this readily loses C0 2 to give the 1.4-diketone, which can be reduced by the usual methods to 1 . 4-cy clohexanediol from which cyclohexane may be pre- pared by reducing with hydriodic acid or the iodide converted into A 1 - 4 -cyclohexadiene. (6) Condensation of 8-ketonic acids to 1 . 3-cy clohexanediones. (7) Addition of sodium malonic ester to a (3 unsaturated ketones to form derivatives of 1.3-cyclohexanedione. (8) Condensation of two molecules of acetoacetic ester with alde- hydes to form open chain diketonic acids which condense further to cyclic unsaturated ketonic esters which readily lose C0 2 on saponifi- cation to give cyclohexenone derivatives. Similar products are ob- tained by the condensation of methylene iodide and two molecules of sodium acetoacetic ester. 4 a W. H. Perkin, Jr., Ber. 27, 216 (1894). 2 Markownikow, Compt. rend. 110, 466 (1890) ; 115, 462 (1892). 'Baeyer, Ann. Htf, 106 (1888) ; Ber. 23 f 1276 (1890). Hagemann, Ber. 26, 876 (1893). THE CYCLOHEXANE SERIES 281 (9) Condensation of aliphatic aldehydes and ketones, for ex- ample, methyl heptenone to a mixture of meta-xylene and dimethyl- cyclohexene; citronellal to isopulegol, etc. (10) Addition of chlorine to benzenoid hydrocarbons, for example, the addition of chlorine to benzene to form hexachlorocyclohexane. (11) The indirect reduction of unsaturated substances by first adding bromine or hydrobromic acid and then replacing the bromine by hydrogen by treating with acetic acid and zinc, for example the conversion of dihydro and tetrahydroterephthalic acids to cyclohexane- 1.4-dicarboxylic acid. (12) The hydrogenation or reduction of benzenoid hydrocarbons. As mentioned above, these methods, particularly the well-known method of Sabatier and Senderens, have practically superseded all the older methods and promise to become of industrial importance for the hydrogenation of benzene to cyclohexane, phenol to cyclohexanol and cyclohexanone, both the latter products being of value as com- mercial solvents (see below). The ease of reduction or hydrogenation varies considerably with the number and character of the substituent groups. Thus terephthalic acid and the dihydro and tetrahydro-tereph- thalic acids are reduced with difficulty, but mellitic acid is easily re- duced by reducing agents to cyclohexanehexacarboxylic acid. HOC 3 :o 2 H With the phenols the ease of reduction increases with the number of hydroxyl groups, resorcin being easily reduced to cyclohexane-1.3- dione. Benzene is reduced to cyclohexane in the presence of catalytic nickel at 180-250, but the refinements of the process as earned out industrially are not generally known. Cyclohexane was manufactured in this way in the United States and in Germany during the recent war, the cyclohexane being used to some extent as a motor fuel 5 for aero- planes. In the case of the alkyl derivatives of benzene some decompo- 8 Dayton Metal Products Co. Brit. Pat. 133,288; 133,667 (1919). 2 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS sition also takes place, for example, in the hydrogenation of para- cymene at 170-180 to para-menthane, small proportions of methyl and ethylcyclohexane are also formed. At about 300 the mixture in equilibrium in the presence of nickel consists chiefly of benzene and at this temperature dehydrogenation of cyclohexane to benzene can be effected. The relative ease with which this change is brought about makes possible the detection of small proportions of cyclohexane in the presence of normal hexane, and other saturated hydrocarbons, the fraction distilling at 75-85 being passed over nickel at 300 and the distillate treated with concentrated sulfuric acid to remove and poly- merize defines and then treated with nitrating acid mixture to remove the benzene which may be identified as crystalline dinitrobenzene. 6 As regards the hydrogenation of benzene to cyclohexane by hydrogen in the presence of platinum black Willstatter and Hatt 7 show that the reaction proceeds quantitatively at atmospheric pressures in about six hours in glacial acetic acid solution, using about . 1 part of platinum black. The hydrogenation is distinctly slower when glacial acetic acid is not used as a solvent. The catalyst is exceedingly sensitive to traces of thiophene, less than 0.01 mg. of thiophene per gram of benzene com- pletely preventing the hydrogenation. Toluene is reduced to methyl- cyclohexane under the same conditions much more readily than in the case of benzene, i. e., in about 3% hours. The hydrogenation of benzene derivatives to the corresponding de- rivatives of cyclohexane may conveniently be considered here. As- chan 8 reduced sodium benzoate by sodium amalgam, neutralizing the caustic soda by carbon dioxide, as fast as formed, thus preventing the precipitation of the sodium benzoate by the concentrated caustic soda. When the hydrogenation is incomplete a considerable pro- portion of A 2 -cyclohexene carboxylic acid is formed. Ipatiev 9 ob- tained yields of 40 to 50 per cent of cyclohexane carboxylic acid by his method of reducing at 300-320 and hydrogen at about 210 atmos- pheres in the presence of nickel oxide. Phthalic acid under the same conditions is more readily reduced to the corresponding cyclohexane- 1.2-dicarboxylic acid and this method is probably the best method of Tausz, Chem. Ztg. 37, 334 (1914). According to Zelinsky (Cf. Wieland, Ber. tf, 484 [1912]), hydrogen is dissociated from cyclohexane, with the formation of benzene, at temperatures below 300 and in the presence of nickel ; under the same conditions cyclopentane and cycloheptane are stated to be practically unchanged. In the absence of a catalyst cyclohexane yields considerable benzene at 490 ; normal hexane at sightly higher temperatures yields methane, amylene and other hydrocarbons. (Jones, .7. Chem. Hoc. 107. 1582 [1915].) 7 Ber. tf f 1471 (1912). 8 Ber. 24, 1864 (1891). Ber. 41, 1005 (1908). THE CYCLOHEXANE SERIES 283 preparing this acid. It is noteworthy that at the same temperatures and with the same catalyst but without the use of pressure no hydro- genation of phthalic acid could be detected. The older method of re- duction by means of sodium and amyl alcohol was used successfully for the reduction of anthranilic acid to 2-amidocyclohexanecarboxylic acid, and also para-amidobenzoic acid to 4-amidocyclohexanecarboxy- lic acid. 10 Osterberg and Kendall 1X used sodium and alcohol for the reduction of the oxime of cyclohexanone to cyclohexylamine and also report that the method of Sabatier and Senderens gives good yields of the amine from the oxime, but state that the method of reducing aniline to cyclohexylamine, according to Ipatiev, did not give satisfactory re- sults. Ipatiev reported yields of 40 to 50 per cent of the amine by reducing aniline with hydrogen and nickel oxide, employing a hydro- gen pressure of about 120 atmospheres at 220-230. Quinoline could not be hydrogenated by Padoa and Carughi, 12 using the method of Sa- batier and Senderens, but Ipatiev succeeded in reducing it to deca- hydroquinoline by his high pressure method. Diphenylamine also can- not be reduced to dicyclohexylamine by the Sabatier and Senderens method, other products being formed, but the Ipatiev method, at 225- 230, gives a good yield of dicyclohexylamine. 13 Paals' method on aniline is reported to give a yield of about 10 per cent of cyclohexyl- amine. Osterberg and Kendall recommend Ipatiev 's method for the preparation of cyclohexane and cyclohexanol. Ipatiev finds that nickel oxide hydrogenates benzene and its derivatives several times faster than reduced nickel in the presence of hydrogen under pressure at about 255. The cyclohexane so produced is practically pure and the reduction is complete in about iy 2 hours when using 2 g. nickel oxide to 25 g. benzene. Decomposition with the formation of methane and the separation of carbon begins to be noticeable at about 290. Un- der the same conditions Ipatiev reduced phenol to cyclohexanol, di- phenyl to dicyclohexyl, naphthalene (in two successive operations) to decahydronaphthalene, dibenzyl to dicyclohexylethane, (3-naphthol to p-hydroxydecahydronaphthalene and a-naphthol to a-hydroxydecahy- dronaphthalene. Anthracene was reduced by Godchot 14 by reduced 10 Einhorn & Meyenburg, Ber. 27, 2466 (1894). In the case of anthranilic acid the reaction proceeds in two ways, pimelic acid also being formed, probably through the intermediate formation of salicylic acid. NH 2 OH CH 2 CH 2 CO 2 H C 8 H 4 < > C a H 4 < + H.,0 + 4H | C0 2 H C0 2 H CH 2 CH 2 CH 2 C0 2 H 11 J. Am. Chem. Soc. 42, 2616 (1920). Atti. Accad. Lincei (5) 15, 113 (1907). "Ipatiev, Ber. 41, 991 (1908) ; 40, 1281 (1907). "Compt. rend. 139, 605 (1904). 284 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS nickel at atmospheric pressure to tetra and octohydroanthracene, but Ipatiev succeeded in completely reducing it to perhydroanthracene, C 14 H 24 , by his high pressure method, using nickel oxide as a catalyst. By one operation tetrahydroanthracene, melting-point 103-105, is the chief product; a second operation yields mainly decahydroanthra- cene, C 14 H 20 , melting-point 73-74, and a third operation using the decahydroanthracene yields the completely reduced hydrocarbon, C 14 H 24 , melting-point 88-89. Slight carbonization and formation of methane occurs at the temperatures employed, i. e., 260-270. Phenanthrene was also reduced in steps, the completely reduced hydro- carbon being finally obtained. At lower temperatures the hydrocar- bon was not completely reduced, the temperatures required being con- siderably higher than for the reduction of benzene, At 320 phenanthrene > chiefly C 14 H 12 and C 14 H 14 At 360 " > " C 14 H 18 At 370 " > " C 14 H 24 The completely reduced phenanthrene is a liquid boiling at 270-276, does not crystallize at 15 and is inert in the cold to nitrating acid mixture, bromine and aqueous permanganate. Phenyl ether is decom- posed under the conditions of Ipatiey's method, yielding a mixture con- sisting of cyclohexane, cyclohexanol and cyclohexyl ether (temperature employed 230). As regards the practicability of developing Ipatiev's method into an industrial process, it may be pointed out that the pressures employed are at least no higher than the lowest pressures employed for the syn- thesis of ammonia from nitrogen and hydrogen, and the temperatures required are very much lower. The technique of operating at such pressures on an industrial scale has been improved to a degree which should make Ipatiev's process entirely feasible industrially. In con- nection with the hydrogenation of complex benzenoid hydrocarbons it should be noted that attempts have recently been made to hydrogenate coal under high pressures to oily hydrocarbon mixtures from which oily hydrocarbons, or polynaphthenes having lubricating value, may be obtained. This work, the details of which are not yet available, are undoubtedly based upon the earlier findings of Bergius 15 that, under 15 According to U. S. Pat. 1,342,790, issued to F. Bergius, pulverized coal is mixed with a mineral oil boiling above 200 and introduced as a thick paste into a reaction vessel, where it is heated to about 400 and subjected to the action of hydrogen, with- out introducing any catalytic substance, under a pressure of 100 atmospheres. Partial hydrogenation is claimed, a heavy oil boiling about 300-400 being formed from the coal substance. Bergius claims that with soft coals as much 85% of the coal may thus be converted into oily liquid or oil soluble products. THE CYCLOHEXANE SERIES 285 very high pressures, hydrogenation can be effected without & catalyst. For the hydrogenation of such an impure material as coal, it is obvious that either a high pressure method of the Bergius type, or the em- ployment of a catalyst not poised by sulfur, will be required. Skita has shown that the benzene ring may be reduced at ordinary temperatures by colloidal platinum. Thus cinnamic aldehyde is con- verted into cyclohexyl propyl alcohol by reduction in this way using very slight pressures, i. e., about one atmosphere. Cyclohexanone is also very rapidly reduced to cyclohexanol in the same manner. Acetic acid is usually employed as a solvent. The unsaturated ketone pu- legone also yields the saturated alcohol menthol under the same con- ditions. 16 For research and laboratory preparations, Skita's method is usually to be preferred although the method does not appear to have been applied to many reductions of the benzene nucleus. Aqueous alcohol may also be employed as a solvent. In the reduction of pulegone to menthol 5 grams, in 40 c.c. acetic acid, and with colloidal platinum and a little gum arabic as a protective colloid to retard pre- cipitation of the metal, the reduction is complete in sixty minutes. Cyclohexane: The presence of cyclohexane in Russian petroleum was shown by Markownikow 17 and Young found it also in a specimen of gasoline from an American petroleum, but the exact origin of the oil examined by Young is not known. Its presence in the fraction boiling at 80-81 may be indicated by the physical constants of the fraction and the isolation of adipic acid among the products of oxidation by nitric acid. The boiling-point is usually given as 80.8, but the physi- cal properties of a specimen of the hydrocarbon carefully purified by several treatments with slightly fuming sulfuric acid are given by 11 2 Auwers 18 as follows, boiling-point 80.0-80.2 at 749 mm., d -^- 0.7869, n D 1.42910, M D 27.66 (calculated 27.71). Cyclohexane is slightly more stable to heat than normal hexane, but the decomposition of both is noticeable at 500 and under pressure. According to Ipatiev 19 it is decomposed at 500-510 and in the pres- ence of alumina (110 atmospheres pressure) to a complex mixture of decomposition products among which methyl cyclopentane was identi- fied by means of the easily formed tertiary nitro derivative 16 Ber. tf, 1496 (1915). 11 Ber. SO, 974 (1897). 18 Ann. 410, 262 (1915). J. Ruse. Phys.-Chem. Soc. tf, 1431 (1912). 286 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CH 2 -CH 2 N0 2 A>C< Cyclohexane is practically inert to bro- H 2 CH 2 CH 3 mine in the cold and is only very slowly reacted upon at its boiling- point and in diffused daylight, but bromination is rapid in direct sun- light. In the presence of anhydrous aluminum bromide a mixture of high boiling products is formed, but when dibromo cyclohexane or cyclohexene is similarly treated it is possible to identify hexabromo- benzene among the reaction products. 20 Considerable has been written about the so-called "jormolite" re- action which, according to Nastjukow, the cyclohexenes and other non- benzenoid cyclic hydrocarbons undergo when treated with formalde- hyde in the presence of concentrated sulfuric acid or aluminum chlo- ride. According to Nastjukow 21 a mixture of cyclohexane, anhydrous aluminum chlorine and trioxymethylenes react forming a mixture of condensation products but no definite reaction product was isolated. The reaction is usually carried out using sulfuric acid as the condens- ing reagent and the saturated acyclic hydrocarbons are supposed not to give the "formolite" reaction. It does not appear that any definite reaction products have ever been isolated and the character of the reaction therefore is at present a matter of speculation; also it does not appear that the reaction has been carried out with a sufficient num- ber of pure hydrocarbons to warrant the proposal that it be employed in the study of petroleum fractions to determine what types of hydro- carbons are present. As carried out according to Nastjukow one vol- ume of the hydrocarbon mixture is treated with one volume of concen- trated sulfuric acid and then one-half volume of concentrated (40%) formaldehyde is gradually added, with agitation. A precipitate is formed which after washing with gasoline, water and ammonia, may be dried and powdered, the product resembling a brown re'sin. The higher boiling oils generally yield larger proportions of "formolite" resin than the lighter oils, but the yield of resin appears to vary evi- dently with the conditions of the operation. The product prepared according to Nastjukow's directions contains considerable sulfur, a spindle oil giving a resin containing 6.98 per cent sulfur and 6.66 per cent oxygen. When the mixture is kept cold the condensation product is liquid and does not contain sulfur. 22 According to Gurwitsch 23 only 20 Bodroux & Taboury, Bull. soc. chim. 9, 592 (1911). 21 J. Russ. Phys.-Chem. Soc. J7, 46 (1915). J. Russ. Phus.-CJiem. Soc. W, 1596 (1910). 28 Wissensch, Grundlagen d. Erdolbearb., 46. THE CYCLOHEXANE SERIES 287 certain defines such as "partially reduced aromatic compounds" and aromatic hydrocarbons react to give formolite resins. Terpenes are said to give "formolite" resins, but they are very energetically poly- merized, oxidized and esterified by sulfuric acid alone and it is en- tirely obscure what the function of the formaldehyde is supposed to be. Most books on petroleum testing describe the "formolite" reaction and often use the phrase "formolite number," but this test and these phrases are meaningless until the reaction is studied in many cases, not only of cyclohexane, the cyclohexenes, and commercial oils of known chemical character, but also applied to a series of definite pure hydro- carbons of various types. Cyclohexenes should give "formolite" resins since they polymerize readily with sulfuric acid alone and cyclohexa- diene reacts violently with sulfuric acid. Nastjukow may have dis- covered something, but, if so, no one has been able to determine what it is. A series of metallo derivatives of cyclohexane has been prepared by Griittner, 24 who has prepared cyclohexyl derivatives of lead, tin and bismuth. The starting point in all cases was the reaction of cyclohexyl magnesium bromide on the chloride or bromide of the other metal. Bromocyclohexane reacts with magnesium in ether very much like n . hexyl bromide, the secondary reaction RMgBr + R.Br- MgBr 2 + R.R., taking place in both cases. The cyclohexyl derivatives show slight differences from the simple alkyl derivatives of lead. Tetracyclohexyl lead reacts with hydrogen chloride or hydrogen bromide to give X 2 Pb and the simple alkyl derivatives give Pb X . R 3 under (CAi), the same conditions. Cyclohexyl-magnesium bromide and lead chlo- ride in ether react very smoothly. Tetracyclohexyl tin crystallizes in fine microscopic aggregate melting at 248 and is easily soluble in ben- zene, chloroform and carbon bisulfide; bromine reacts with it to give SnBr 2 (CgHnJz, long well formed needles melting at 58. Tiffeneau and Gannage 25 prepared dicyclohexyl-mercury by the action of sodium amalgam on bromocy clohexane ; the mercury derivative forms needles of a camphor-like odor, melting at 139. Mercury derivatives of "Ber. 47, 3257 (1914). " J. Chem. Soc. At s. 120 (1), 472 (1921). 288 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS methylcyclohexane were similarly prepared from 4-bromomethylcyclo- hexane. Derivatives of the type RHgCl were made from dicyclohexyl- mercury by the action of benzoyl chloride or arsenic trichloride. Some Simple Derivatives of Cyclohexane: An extensive review of the derivatives of cyclohexane is beyond the scope and purpose of the present volume but a brief description of some of the more important derivatives indicating the close parallelism in the chemistry of cyclo- hexane and normal hexane, and other examples of chemical behavior which are likely to prove of interest in connection with the chemical investigation of petroleum, are given. Cyclohexane is readily acted upon by dry chlorine, direct sunlight not being required. The monochloride, boiling-point 141.6-142.6, 22 d -pr^- 0.9976, is also readily prepared by the action of concentrated hydrochloric acid or PC1 3 on cyclohexanol. On treating with alkalies the chloride forms cyclohexene and when alcoholic caustic alkalies are employed a small proportion of cyclohexylethyl ether is formed; in fact, the behavior of the chloride closely parallels the behavior of the mono- chloron. hexanes. Like the monochloropentanes and monochloro- hexanes the cyclohexyl derivative is decomposed by passing over anhy- drous barium chloride or alumina at 350-450, cyclohexene being formed almost quantitatively. 26 Another process for converting chloro- cyclohexane to cyclohexene describes passing the chloride over lime at 350-450 or over barium chloride at 300-400. 27 Fortey 28 decom- posed the chloride by heating with quinoline and described the result- 4 ing cyclohexene as boiling at 82.3 d 0.8244, but Auwers 29 gives the 15 6 following physical constants, boiling-point 83-83.5 (760 mm.), d ' 0.8143, n D 1.44921, M D 27.03 (calculated M D 27.24). Cyclohexene cannot be made satisfactorily from cyclohexanol by heating with an- hydrous oxalic acid, the principal product being dicyclohexyl oxalate, but heating with potassium acid sulfate gives an 80 per cent yield of cyclohexene. 30 A small proportion of cyclohexyl ether, boiling-point 239-240, is also formed. The bromine and iodine derivatives are naturally more easily decomposed than the chlorides, but a double bond adjacent to the halogen stabilizes the substance as in the aliphatic se- 2a Badische, Anilin n. Soda Fabr., J. C'hem. Soc. Abs. 1913 (1), 349. 27 Schmidt, Hochschwender & Eichler, Chem. Abs. 1917, 1885. 28 J. Chem. Soc. 73, 941 (1898). 29 Ann. JklO. 257 (1915). 80 Willstatter & Hatt, Ber. tf, 1464 (1912). THE CYCLOHEXANE SERIES 289 ries. Usually the halogen derivatives have not been prepared from the hydrocarbon but from the alcohols. Thus cyclohexanol and concen- trated hydriodic acid yield cyclohexyl iodide and quinite yields the corresponding 1.4 dihalogen derivatives. When cyclohexane is chlorinated in the cold a mixture of chlorides is obtained. Two dichlorocyclohexanes are obtained, one boiling at 105.4-106.4 (50 mm.) and the other distilling at 112.4-113.4 (50 mm.) ; the former on prolonged boiling with alcoholic caustic pot- ash yields a chlorocyclohexene. On distilling at atmospheric pressure the dichlorides decompose markedly. Continued chlorination yields tetrachlorocyclohexane, crystallizing from chloroform in long prisms melting at 173 . 81 Cyclohexane is practically unacted upon by the usual nitrating mixture of nitric and sulfuric acids, but may be nitrated by heating in a sealed tube with dilute nitric acid according to the method discovered by Konowalow. 32 Its behavior in this respect is practically identical with that of n.hexane and the properties of the resulting nitrocyclo- hexane are quite different from the properties of nitrobenzene. On reduction with tin and hydrochloric acid, the corresponding amine is not formed but cyclohexanone or its condensation products are formed, evidently through the intermediate formation of the oxime of cyclo- hexanone, \NQ, i50 form oxime clohexanone Dinitro and trinitro derivatives of cyclohexane cannot be prepared by the nitration method noted above. Alkyl derivatives of cyclohexane, such as methyl or dimethylcyclohexane, containing a tertiary hydro- gen atom are much more easily nitrated by Konowalow's method, the nitro group replacing the tertiary hydrogen atom. Like primary and secondary nitro derivatives of the aliphatic series, nitrocyclohexane is soluble in alkalies, evidently forming salts of the iso form whose struc- ture is noted above. Nitro cyclohexane boils at 205.5, d20 1.0616. "Sabatier & Mailhe, Compt. rend. 137, 240 (1903). "Compt. rend. 121, 652 (1895). 290 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Nametkin 83 states that the yield of nitrocyclohexane is increased by nitrating the hydrocarbon with about three parts by weight of alu- minum nitrate. Cyclohexane yields about 56 per cent of nitro- cyclohexane together with cyclohexanone and dinitrodicyclohexyl, C 12 H 20 (N0 2 ) 2 melting at 216.5. Aminocyclohexane and other amino derivatives of the cyclohexanes differ markedly from the amines of the benzene hydrocarbons, par- ticularly in their behavior when treated with nitrous acid. As noted above aminocyclohexane is best prepared by reduction of the oxime of cyclohexanone in alkaline solution or by catalytic hydrogenation by the Sabatier and Senderens method. Heating cyclohexanone or similar ketones with ammonium formate and reduction of the resulting formyl derivative to the amine has also occasionally been employed, 34 but re- duction of the oxime generally gives much better yields. In the case of tertiary nitro compounds, such as 1 -nitro- 1-methy Icy clohexane, re- duction of the nitro group gives satisfactory yields since the tertiary nitro derivatives cannot rearrange to the iso forms with the resulting formation of oximes and ketones. The aminocyclohexanes yield comparatively stable nitrites when treated with nitrous acid and on heating their aqueous solutions de- composition takes place with difficulty yielding the corresponding alco- hol (yields usually very poor), and decomposition also proceeds in another manner with the formation of ammonia and unsaturated hy- drocarbons. The latter reaction can be modified so as to serve admir- ably for the preparation of unsaturated hydrocarbons, particularly in cases where the resulting unsaturated hydrocarbon is easily rearranged or polymerized. For this purpose the amine is subjected to exhaustive methylation and the resulting alkylated ammonium hydroxide decom- posed by gentle heating, a method mentioned in connection with cyclo- butene and which warrants more extensive applications in research. Decomposition of the phosphates of amines of this type by heating has also been employed for converting the amines to unsaturated hydro- carbons. 85 The method of reducing the oximes to amines has been employed for the preparation of the 1.3-diamine and 1.4-diamine, the oximes being prepared from the corresponding ketones. The 1 . 2-diamine has been prepared from anthranilic acid which can be best reduced by the U J. Russ. Phvs.-Chem. Soc. 42, 581 (1910). "Leuchart & Bach, Ber. 20, 104 (1887). "Harries, Ber. S^ 300 (1901). THE CYCLOHEXANE SERIES 291 method of Ipatiev or by sodium and amyl alcohol. 36 The amide of the reduced acid may then be converted to the diamine in the usual manner by bromine and alkali. 37 The cyclohexadienes have been of considerable interest on account of their close relation to benzene. A cyclohexadiene boiling at 84-86 was first made by Baeyer 38 and the same laborious methods of prepa- ration were later employed by Crossley. 39 A product evidently identi- cal with Baeyer's and having the same boiling-point was made by Markownikow by decomposing chlorinated cyclohexane isolated from Russian petroleum. 40 Fortey reported a cyclohexadiene boiling at 81- 82 41 and Harries and Antoni 42 obtained a product of the same boiling- point, 81.5, by the decomposition of the phosphate of 1.4-diamino- cyclohexane. From 1 . 2-dibromocyclohexane Crossley 43 also obtained the low-boiling product and, from its method of preparation and the fact that oxidation by nitric acid yielded oxalic and succinic acids, con- cluded that the low-boiling product was 1 . 3-cyclohexadiene. CH, CIL CO,H. C0 2 H C0 2 H C0 2 H. Zelinsky and Gorsky 44 obtained the high-boiling hydrocarbon from 1.4 dibromocyclohexane and the low-boiling one from 1.2 dibromo- cyclohexane. Both hydrocarbons form different and characteristic di- bromides and tetrabromides. A 1 - 4 cyclohexadiene B.-P. 85-86 "Einhorn & Meyerburg, Ber. 27, 2466 (1894). "Einhorn & Bull, Ber. 29, 964 (1896) ; Ann. 295, 187. "Ber. 25, 1840 (1892). "/. Chem. Soc. 85, 1410 (1904). 40 Arm. 302, 30 (1898). 41 J. Chem. Soc. 73, 945 (1898). 42 Ann. 328, 93, 105 (1903). 48 Z,oc. cit. "Ber. 41, 2479 (1908). 292 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS A 1 - 3 cyclohexadiene B.-P. 81.5 Although the evidence of the existence of the two isomeric hexa- dienes is quite clear some doubt as to their constitution has been ex- pressed on account of the fact that neither of the hexadienes shows the exaltation of the molecular refraction which two conjugated double bonds were supposed always to show. However, Auwers has shown that cyclopentadiene, which must contain conjugated double bonds, and cycloheptadiene, containing conjugated double bonds, do not show any exaltation and the conjugated cyclic trienes show only very slight ex- altation. 45 The agreement with the calculated value of A 1 ^-cyclo- hexadiene is in fact within the experimental error, if we accept the more recent determinations of Harries 46 and of Willstatter and Hatt. 4T EMa EM D Harries 0.05 0.09 Willstatter and Hatt 0.00 0.02 A similar discrepancy between the observed refractivity and the expected exalted value due to conjugation of double bonds confused for a time the question of the constitution of the substituted cyclo- hexadiene, a-terpinene (q. v.). The determination of the constitution of such hydrocarbons has been particularly difficult on account of the ease with which the double bonds shift their positions. Thus the prepa- ration of pure a- or y-terpinene, a- or (3-phellandrene, and terpino- lene is practically impossible. The cyclohexadienes show a very marked tendency to oxidize to benzene (or its homologues), for example, the oxidation of the ter- pinenes to cymene. Also cyclohexadiene (probably a mixture of the two isomers) is converted to benzene by dehydrogenation in the pres- ence of nickel at the remarkably low temperature of ISO . 48 Dilute acids very frequently cause shifting of double bonds when a more stable substance can result, and a double bond in a side chain fre- quently shifts to the ring. Thus 2-pheny 1 and 2-propy 1-A 2 5 8 (9) -men- 46 For a fuller discussion of the refractivity of cyclic and acyclic hydrocarbons see the chapter on physical properties. "Ber. 45, 809 (1912). 47 Ber. 45, 1647 (1912). "Boeseken, Rec. trav. chim. 37, 255 (1918). THE CYCLOHEXANE SERIES 293 thatriene are quickly converted to the isomeric benzene deriratives by warming with 3 per cent hydrochloric acid. 49 Conjugated dienes react with concentrated sulfuric acid with al- most explosive violence, with tar ^formation and reduction of the acid, a behavior frequently noted on refining crude benzene containing cyclo- pentadiene and cyclohexadiene and this energetic action is particu- larly marked when the crude benzene has been manufactured from oil, as in Pintsch gas "hydrocarbon" or carburetted water gas tar. Un- der the same conditions that amylene and such simple defines give good yields of. the alcohols (i. e., by treating with ordinary sulfuric acid in the cold and diluting with water) the conjugated diolefines yield only tar. Cyclohexanol: This alcohol promises to become a common com- mercial product 50 as a result of the development of methods of cat- alytic hydrogenation, being readily prepared from phenol. Cyclohex- 37 anol has a camphor-like odor, boils at 160.9, melts at 23, d -j^- 0.9397, nj) 1.46055. 51 It is sparingly soluble in water but is hygro- scopic, a little water lowering the freezing-point, a eutectic point being noted at 47.4, the liquid containing 4.97 per cent of water at that point. 52 The acetate resembles amyl acetate and while it has no very marked physiological action, the narcotic action of the vapors is about three times greater than the same property of amyl acetate. 53 The naphthylurethane 54 melts at 139-140. When phenol is reduced with hydrogen over active nickel at 160- 170, the nickel having been reduced from the oxide at 300, the prod- uct is chiefly cyclohexanol together with a little unchanged phenol and a little cyclohexanone. Holleman 55 removed the cyclohexanone by condensing it with benzaldehyde in the presence of alkali. When cyclo- hexanol is passed over copper, with a little air, at 280 cyclohexanone Klages, Ber. 40, 2360 (1907). 60 The use of cyclohexanol in soap is said to enable one to incorporate solvents such as benzene, tetraline. chlorinated solvents and the like in the soap and also facilitates the manufacture of soaps containing phenolic insecticides. Its use as a solvent for rubber in reclaiming rubber is mentioned in German Patent 366,146. Like fusel oil and amyl acetate, cyclohexanol and its acetate are of value as solvents for nitro cellulose, such solutions being capable of considerable dilution with the common hydrocarbon solvents, gasoline, benzene, etc. The use of cyclohexanol and cyclo- hexanone in the manufacture of celluloid has been patented by Raschig, German Pat. lT4,yi4 (190o), 51 Auwers, Ann. 410, 257 (1915). "Forcrand, CAnpt. rend. 155, 118 (1912). "Lehmann, Chem. Abs. 1913, 2432. "Neuberg, Bioch. Z. 27, 339. Rec. trav. chim. 24, 19 (1905) ; Brochet [J. Soc. Chem. Ind. SS, 1031 (1913)] used nickel and hydrogen at 120-180 and 10 to 15 kilograms per sq. cm. pressure. 294 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS is produced in good yields. Sabatier and Senderens 56 reduced phenol at a higher temperature and obtained a mixture of cyclohexanone and cyclohexanol, from which they prepared cyclohexanol by passing again over the catalyst with hydrogen at a lower temperature, 140-150, and prepared nearly pure cyclohexanone by passing the mixture over copper at 330. Cyclohexanol is readily oxidized to cyclohexanone by chromic acid, under the same conditions that aliphatic secondary alco- hols are oxidized to the corresponding ketones. Cydohexane-1 .2-diol, melting-point 99-100, is formed when cyclohexene is oxidized by cold dilute permanganate in the usual man- ner. Cyclohexane-1. 3- diol, 57 melting-point 65, is produced by the catalytic hydrogenation of resorcinol in the presence of nickel at 130. It is easily soluble in water and alcohol, does not reduce Feh- lings' solution or give a color with ferric chloride. Cyclohexane-1 .2.3- triol was also obtained by the catalytic hydrogenation of pyrogallol^ the triol forming very hygroscopic crystals melting at 67. Cyclo- hexane-1.3.5-diol, made by reducing phloroglucine with sodium amal- gam, melts at 184. Cyclohexane-1 .4- diol, also called quinite can be obtained by catalytic hydrogenation of hydroquinone and was also prepared by Baeyer by reducing cyclohexane-1.4-dione with sodium amalgam. It was named quinite on account of its relation to benzo- quinone, which it yields when oxidized by chromic acid. Two other hydroxyl derivatives of cyclohexane may be mentioned on account of their interest to biochemistry, i. e., quercite, cyclohexane-1.2.3.4.5- pentol, and inosite, cyclohexane-1.2.3.4.5.6-hexol. Quercite is known in two forms [ CS 101. 0.771 1.423 y v yCH 1.1-dimethylcyclohexane { }\ 120. 0.781 1.430 1.2-dimethylcyclohexane < V-CH, 123. 0.779 1.429 rated* mor^^d^^rom^rch^the?. 6 ^^ ^ Te&SG8 ** ^ methyl groups are & Ssorokina, Chem. Ztg. 37, 725 (1913). " J. RUBS. Phys.-Chem. Soc. 45, 613 (1913). "Ann. 4^0, 88 (1919). 302 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CH, 1.3-dimethylcyclohexane " 1.4-dimethylcyclohexane 1 . 1 .3-trimethylcyclohexane 1.2.4-trimethylcyclohexane C 1.2.3-trimethylcyclohexane 80 1 .3.5-trimethylcyclohexane 1 .2.4.5-tetramethylcyclohex- ane 119. 120. 0.771 0.769 1.425 1.424 138. 0.790 1.436 140. 0.778 1.429 149.6-150 138. 0.772 1.429 161. 0.785 1.434 The boiling-points and densities of other alkyl derivatives of cyclohexane are given in the following table. ALKYL DERIVATIVES OF CYCLOHEXANE II. Name D D r> 20 Ref B.-P. Density f -^ 1.2-Methylethyl cyclohexane 151 0.784 i n.Propylcyclohexane 156 0.7865 a Tertiarybutylcyclohexane . 166-167 1/>o 08305^^- 4 1.3-Methylethylcyclohexane 145-146 0.8320 6 1.3-Methylpropylcyclohexane 164-165 5 1.2-Methylisopropylcyclohexane [o-menthane] ... 171 21 0.8135-^- 6 1.3-Methylisopropylcyclohexane [m-menthane] . . 166M67 94 0.7965^- 6 "Zelinsky, Ber. 35, 2677 (1902), gives the following, boiling-point 119.5-120, d~ 0.7661. Zelinsky prepared the hydrocarbon by converting 1.3-dimethyl cyclohexanol (1) to the iodide and reducing it with zinc in acetic acid. oTreppmann & Krollpfeiffer, Ber. 48, 1226 (1915). THE C YCLOHEXANE SERIES 303 ' 25 1.4-Methylisopropylcyclohexane [p-menthane] ... 167-168 0.8028-^r 6 2-Cyclohexyl-2-methylbutane CeHu.C(CH3) 2 C 2 H5 191M92 0.8226^- 4 2-Cyclohexyl-2-methylpentane 206-207 0.8372^ 4 3-Cycloliexyl-3-methylpentane 207-208 0.8310^- 4 3-Cyclohexyl-3-ethylpentane 222-223 05388^ 4 in* 2-Cyclohexyl-2.4-dimethylpentane 220-221 0.8304-^- 4 1 rO 3-Cyclohexyl-3-methylhexane 224-226 0.8406 -^- 4 17 l-Methyl-2-isoamylcyclohexane 204 0.812-^5- 7 1 Sabatier and Senderens, Compt. rend. 132, 210, 556 (1901). 2 Murat, Ann. chim. phys. (8) 16, 108 (1909). 3 Kursanoff, Ber. 34, 2035. 4 Halse, J. prakt. Chem. (2) 92, 40 (1915). The hydrocarbons described by Halse were made by Willstatter's method of catalytic hydrogenation. 5 Mailhe and Murat, Bull. soc. chim. 7, 1083 (1910). Zelinsky, Ber. 35, 2677 (1902), gives the boiling-point 148-149, d 1 ^ 0.7896. 6 Sabatier and Murat, Compt. rend. 156, 184. 7 Murat, Ann. chim. (8) 16, 108 (1909). Of the hydrocarbons noted in the above tables, cyclohexane, methyl- cyclohexane, 1 . 3-dimethylcy clohexane and 1 . 3 . 4-trimethy Icy clo- hexane have been reported in the lighter fractions of Russian petro- leum, and the methyl, propyl, 1 . 3-dimethyl and 1.4-dimethyl derivatives have been reported in rosin oil. The method of identify- ing alkyl cyclohexanes by brominating in the presence of aluminum bromide to benzene derivatives which are supposed to retain the alkyl groups in the same relative positions as they occurred in the original cyclohexane hydrocarbon, is open to the objection that profound alteration of the carbon structure of the hydrocarbon has frequently been observed in the presence of aluminum bromide; thus 1.1-dime- thy Icy clohexane gives a bromide derivative of para-xylene. The same objection could be made to an attempt to convert the alkyl cyclohexane to the corresponding alkyl benzenes by dehydrogenation over nickel at about 300. In view of the extreme difficulty of separating such hydrocarbons from petroleum by fractional distillation, to which diffi- culty Young has called attention, and the equally great difficulties and uncertainties of identifying them by chemical means (conversion to benzene derivatives or oxidation to known acids, etc.), it is quite 304 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS probable that many of the cyclic hydrocarbons reported to have been identified in Russian petroleum have been so reported on faulty or insufficient evidence. Although a fair number of alkyl derivatives of cyclohexane are known, very little is known of their chemical behavior; for example, to concentrated sulfuric acid, nitric acid, chromic acid and the like. Even in the case of the menthanes, it is not known in what positions bromine enters on bromination and whether or not the tertiary hydro- gen atoms are reactive to sulfuric acid or are easily oxidized. In view of the very large losses which result on treating petroleum distillates with sulfuric acid, it would be desirable to know whether the different types of substituted cyclohexanes, bicyclic and polycyclic hydrocar- bons of different types, saturated in the sense that no double bonds are present, are resistant to air oxidation, resinification, destruction by concentrated sulfuric acid, etc. The Substituted Cyclohexenes follow very generally the chemical behavior noted in the so-called terpene series. Only in comparatively recent years has it been realized that the chemistry of these hydro- carbons occurring in nature cannot be dissociated in any way from the chemistry of the simple derivatives of cyclopentane, cyclohexane and cycloheptane. The boiling points, densities and refractive indices of a number of unsaturated hydrocarbons of this series are given in the following table. THE CYCLOHEXANE SERIES 305 e*-* 3 o o r ll" 5 co -H ^ co t>: O O .-i O -i ^ p3 w 1 v^ /S/S O uOO = o S\ g M 5S I 1 I $ X I 1 -*J -*^ 1 306 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS e* o o o o . i-H OS i I - 2 " g- i Ji I I ! . I | r}j .Jl - I ||1I 13 i^llail f-id COTP OOt>. 13 11 THE CYCLOHEXANE SERIES 307 Wallach 81 tabulates the physical properties of a series -of cyclo- hexane derivatives each of which contains a semicyclic double bond, as follows, Boiling-point . .. 102. d ................... 0.8025 M D .............. ... 32.15 MD (calc.) ......... 31.83 <( ^= Boiling-point .... 137.M38. d ............... 0.823 M D ............ 36.82 (calc.) ..... 36.43 123.-124. 0.798 36.95 36.43 122.M23 0.7925 36.93 36.43 CH, 153. 0.813 41.65 41.03 156. 0.8125 41.65 41.03 CH, Boiling- point d ..... _ , _ , >=CH.C 2 H. < >=CH.C 2 H 6 CH^ V=CH.C a H, Mr>(calc 157.M58. 0.821 41.60 ) 41.03 CH, 170.-173. 0.814 46.35 45.64 CH, 172.-174. 0.815 46.28 45.64 Boiling- point 160.M61. d ..... 0.836 MD .. 41.56 MD (calc.) 41.03 CH, 173.-175. 0.825 46.25 45.64 172.-174. 0.831 45.88 45.64 1 .2-Dimethyl-^-Cyclohexene is of special interest since Meer- wein 82 discovered that it is smoothly formed by the dehydration of the cyclopentane derivative 1-methyl-l-a-hydroxyethylcyclopentane. CH 2 CH 2 CH 3 CH 2 CH 2 C CH 3 I >C< - > | II CH 2 CH 2 CH (OH) . CH 3 CH 2 CH 2 C CH 8 The hydrocarbon is also formed by the dehydration of 1 . 2-dimethyl- cyclohexanol(l). It is therefore readily prepared from methylcyclo- hexane-2-one by treating with methyl-magnesium iodide and dehy- drating the resulting alcohol. The nitrosochloride is bluish in color, easily volatile with steam and melts at 58-60. It yields a dibro- mide C 8 H 14 Br 2 melting at 154-156 and by oxidation, the glycol melting at 38-39. 83 Ann. 360, 34. "Ann. 417, 255 (1918). Wallach, Ann. S96, 278 (1913). 308 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Ethylidenecyclohexane has been synthesized by the Reformatsky synthesis, condensing cyclohexanone and H, CH CH, CH "Wallach, Ann. 589, 189 (1912). "Wallach, Ann. 360, 56. "Wallach, Ann. 896, 264 (1913). 88 J. Am. Chem. 8oc. 41, 1678 (1919), THE CYCLOHEXANE SERIES 309 The hydrocarbon I, 1.4-di-isopropenyl-A 1>4 -cyclohexadiene, melts at 117-117.5, yields an oily tetrabromide and also a crystalline tetra- bromide melting at 107-109. When it was attempted to add more bromine, substitution and evolution of hydrobromic acid occurred similar to the behavior of A 3 - 8(9) -p-menthadiene noted by Perkin, which adds smoothly only two atoms of bromine supposedly on ac- count of the fact that the two double bonds are in the conjugated position. Bogert and Harris regard their liquid and crystalline tetra- bromides as cis and trans isomers of the substance CH d C CH,Br The refractive index of the hydrocarbon was determined in benzene and in chloroform solutions, using Eisenlohr's values, and an exalta- tion of the molecular refraction, due to the two conjugated double bond systems of 3.776, was found. Bogert and Harris note that almost the same exaltation of the molecular refraction was noted in the case of 1.4-disopropenylbenzene, i.e., 3.841, which they believe points to a structure analogous to Dewar's structure for styrene, 310 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CH C=CH Z CH=C-CH 3 The values obtained for the magnetic rotatory power of p-di-isopro- penyl-benzene, however, point to a Kekule, not a Dewar structure. Cantharene, a hydrocarbon obtained by the decomposition of cantharidene, has been synthesized by Haworth 87 from 1-methyl-A 1 - cyclohexene in the following manner. The nitrosochloride of the methylcyclohexene was heated with sodium acetate in acetic acid, removing hydrogen chloride in the usual manner and by hydrolyzing the resulting unsaturated oxime l-methyl-A 6 -cyclohexene-2-one was obtained. A methyl group was introduced by means of methyl- magnesium iodide and the resulting tertiary alcohol was decomposed by heating with 8 per cent oxalic acid. Cantharenol Cantharene Monocyclic Sesquiterpenes : The name "sesquiterpene" has been employed for a number of hydrocarbons of the formula C 15 H 24 occur- ring in essential oils. Semmler has recently made several hydrocar- bons of this empirical formula by condensing isoprene with various n J. Chem. Boo. 10S, 1242 (1918). THE CYCLOHEXANE SERIES 311 terpenes by heating them together in sealed tubes. Very little is known regarding the constitutions of the sesquiterpenes beyond the fact that some are acyclic and have four double bonds, some are monocyclic and have three double bonds, some are bicyclic having two double bonds and others are tricyclic and have only one double bond. It will readily be understood that the possible number of isomeric hydrocarbons is very great and it now appears that most of the hydrocarbons, described in the literature of twenty years ago as definite hydrocarbons, are in reality mixtures and that the separation of pure individual hydrocarbons from such mixtures is a difficult task indeed. Also it was usually assumed in the literature that the hydro- carbons regenerated from crystalline derivatives, such as the dihydro- chlorides, were identical with the original hydrocarbons, whereas many instances are known in which the structure of the regenerated hydro- carbon is quite different from the original. The monocyclic sesquiterpenes are probably derivatives of cyclo- hexane and are accordingly so classified. The physical data are often very helpful in showing whether the sesquiterpenes are monocyclic, bicyclic or tricyclic. As noted by Parry 88 the following constants are typical of these several groups. M ol. Refraction Specific Gravity (Calculated) Monocyclic sesquiterpenes 0.875 to 0.890 67.76 Bicyclic " 0.900 " 0.920 66.15 Tricyclic 0.930 " 0.940 64.45 Catalytic hydrogenation by Paal's, Skita's or Willstatter's methods and the reactions with hydrogen chloride or hydrogen bromide also indicate the number of double bonds in the hydrocarbon, hydrogena- tion being more certain since conjugated linkings frequently do not add the maximum number of molecules of halogen acid. Zingiberene and Zingiberol: This sesquiterpene occurs in ginger oil. According to Semmler and Becker 89 it is monocyclic and con- tains three double bonds, one of which is in the ring and two in the side chain. The molecular refraction indicates that two of these double bonds are in conjugated positions; MR = 68.37, calculated for Ci 5 H 24 /- 3 is 67.86. This optical evidence is also supported by its chemical behavior, forming a dihydrochloride, melting at 169-170. Catalytic hydrogenation in the presence of platinum gives hexahydro- "The Chemistry of Essential Oils," Ed. Ill, Vol. I, 71. *>Ber. tf t 1914 (1913). 312 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS zingiberene C 15 H 30 , but reduction by sodium and alcohol yields the monocyclic dihydrozingiberene, C 15 H 26 , which also is good evidence of the existence of two conjugated double bonds, and since this reduction takes place very readily Semmler concludes that these conjugated double bonds are in the side chain. As with other substances contain- ing conjugated double bonds, zingiberene resinifies and polymerizes very readily on standing in the air or warming with sodium. Semmler and Becker state that when zingiberene is treated in acetic acid solu- tion with a little sulfuric acid, that it is condensed to a bicyclic isomer which they have named isozingiberene. They have proposed the fol- lowing constitution for these two hydrocarbons , CH, X. :-CH, CH^N Zingiberene -3 ^ Isozingiberene Zingiberene forms a nitrosochloride melting at 96-97 and a nitrosite melting at 97-98( by treating zingiberene in cold petroleum ether with acetic acid and sodium nitrite). A nitrosate melting at 86 (with decomposition) is formed by treating the hydrocarbon, dis- solved in cold glacial acetic acid, with ethyl nitrite and slowly adding nitric acid. Zingiberene is associated in the essential oil of ginger, with a sesquiterpene alcohol, zingiberol, 90 which yields zingiberene on decom- position by warming gently with potassium acid sulfate. The alcohol is partially decomposed on treating with acetic anhydride and does not readily yield a phenylurethane, and on treating the alcohol with hydrogen chloride or hydrogen bromide in acetic acid, the dihydro- chloride and dihydrobromide respectively, of zingiberene are formed. The alcohol is evidently a tertiary alcohol and from its relations to zingiberene the hydroxyl group must be situated at positions (8) or 80 Brooks, J. Am. Chem. Soc. 38, 430 (1916). THE CYCLOHEXANE SERIES 313 (12) in the above figures. The alcohol has a persistent -aroma of ginger oil but does not have the sharp taste of the "gingerol" discov- ered by Garnett and Greier 91 and recently shown by Nelson, 92 and others 93 to be a phenol derivative. [Cf. particularly Lapworth, Pear- son and Royle.] Zingiberol distills at 154-157 (14.5 mm.). The physical properties of zingiberene, and isozingiberene e^ as follows, Zingiberene Iso-zingiberene BoUing-point . 128-129 (9 mm.) 118-122(7mm.) d 0.8684 0.9118 20 n .............................. 1.4956 1.5062 D Mol. Ref. calc. for CaH*/^ ...... 67.86 " found 68.37 66.50 " " calc. for CuH*/^ ...... 66.15 According to Semmler and Becker the dihydrochloride noted above is really a derivative of isozingiberene since the latter hydrocarbon is formed from the dihydrochloride by digesting with alcoholic caustic potash. Hydrogenation by platinum black in acetic acid yields hexa- hydrozingiberene, C 15 H 30 , boiling-point 128-129 (11 mm.) d20 0.8264, nj) 1.4560. Isozingiberene adds only four atoms of hydrogen to form the saturated bicyclic hydrocarbon. Like myrcene and other conjugated dienes, zingiberene is readily condensed by heating at about 215 to a bicyclic isomer, and to a dimeride, C 30 H 48 , boiling-point 260-280 (11 mm.), d 20 o 0.9287. A synthetic monocyclic sesquiterpene has been made by Roenisch 94 in a manner which leaves little doubt as to its constitution and may properly be named isoamyl- 0.846 to 0.850, n_ 1.47459, 5 [ menthadiene, M = 13.06, the double bonds in the latter hydrocarbon being in the conjugated posi- tions. Limonene and dipentene are most conveniently identified by means of their tetra-bromides, 12 which are best prepared by adding the cal- culated amount of bromine to the hydrocarbon dissolved in about four volumes of acetic acid, keeping the mixture chilled during the gradual addition of the bromine. Remarkably high yields of the nitroso- chlorides of limonene and pinene can be obtained by following the method recently described by Rupe. 13 Concentrated sulfuric acid and concentrated sodium nitrite solution are separately dropped into a flask containing a thin paste of common salt and concentrated hydrochloric acid. The evolved gases are cooled, dried by passing through calcium chloride and then passed into a solution of limonene in one volume of ether and one half volume of glacial acetic acid, cooled in ice and salt. Heating limonene or dipentene nitrosochlorides with alcoholic caustic alkali yields carvoxime. 14 Both d and -carvoxime melt at 72 but the racemic carvoxime, derived from dipentene, melts at 93. When limonene nitrosochloride reacts with sodium azide the chlorine 8 Bcr. tf, 2374 (1910). J. Chem Soc. 93, 1815 (1908); .97, 1249 (1910). Evidence of racemic menthyl mandelates, cf. Findlay, J. Chem. Soc. 91, 905 (1907). 10 Hantzsch, Ber. 45, 553 (1912). 11 J. Chem. Soc. 89, 854 (1906). "Power & Kleber, Arch. Pharm. 232, 646 (1894). 13 Helv. chim. Acta. 4, 149 (1921). "Deussen & Hahn, Chem. Zentr. 1910 (1), 1142. 318 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS atom is replaced by the azide group N 3 ; the resulting nitroso-azide also yields carvoxime on decomposition. 15 Oxidation of limonene in the presence of water yields carvone, carveol (q.v.) and a resin. 16 Oxidation of limonene by chromyl chloride yields chiefly cymene which is then further oxidized to a-p-tolylpropaldehyde and p-tolylmethyl ketone, 17 resembling terpi- nene in this respect, but in the case of limonene much more resin is formed. Condensation of limonene with formaldehyde brought about by the use of para-formaldehyde in glacial acetic acid with the addi- tion of a little sulfuric acid, yields an unsaturated alcohol homo- 19 limonenol, boiling-point 122-126 at 13 mm., d 0.9720. Accord- ing to Prins 18 addition of formaldehyde to a double bond occurs to give an oxide ring which may then hydrolyze to a glycol, which in turn decomposes to give an unsaturated alcohol, as in the case of limonene, or many yield a methylene ether, as indicated in the fol- lowing, R RCH = CHR > RCH CHR > RCH CH< H,i- -4 CH 2 OH RC = CHR R RCH CH< H,OH or CH 2 CH 2 This reaction with formaldehyde has also been studied by Prins in the cases of pinene, camphene, cedrene, etc. When dipentene dihydro- chloride, a by-product in the manufacture of artificial camphor, is treated with chlorine to form a trichloromenthane and this product decomposed, cymene is formed. 19 Carvomenthene: (A 1 -p-menthene?). When limonene is hydrogenated in the presence of platinum black, the reduction proceeds in two stages, the first product being carvo- menthene and the final product paramenthane. 20 Carvomenthene "Forster & Gelderen, J. Chem. Soc. 99, 2061 (1911). "Blumann & Zeitschel, Ber. p, 2623 (1914). "Henderson & Cameron, J. Chem. Soc. 95, 972 (1909). u Chem. Abs. H, 1662 (1920). "Brit. Pat. 142,738 (1919). "Vavon, Butt. Soc. chim. (4) 15, 282 (1914). THE PARAMENTHANE SERIES 319 made in this manner is optically active, [a^yg +118. .Bacon 21 prepared carvomenthene from limonene monohydrochloride (by HC1 in cold carbon bisulfide solution) by making the Grignard complex, magnesium limonene hydrochloride, and decomposing this with water. However, racemization accompanies the formation of the hydro- chloride. Bacon also made the hydrochloride of carvomenthene and reduced it to para-menthane in the same manner by means of the Grignard reaction. Carvomenthene boils at 175-177, its hydro- chloride boils at 85-86 (13 mm.) and the nitrosochloride melts at 95. Para-menthane: By the catalytic reduction of limonene by hydro- gen in the presence of platinum black, 22 by the hydrogenation of para- cymene in the presence of catalytic nickel 23 and by heating the semi- carbazone or hydrazone 24 of menthone with sodium ethylate at 15 160-170, para-menthane is produced, boiling-point 169, d _ 0.803. 15 In the latter process heating the semicarbazone at 160 first forms the hydrazone which subsequently decomposes to the hydrocarbon, >C = N.NH.CONH 2 + H 2 --- >>C = N.NH 2 + C0 2 + NH 3 The Constitution of Limonene: The structure of limonene is inti- mately related to the structure of terpineol, terpin and carvone. Til.den 25 and Wallach 26 had, at an early date, shown that when terpin is digested with dilute acids, it yields terpineol and that by more energetic dehydration terpineol also decomposes further, forming water and dipentene. C 10 H 18 (OH) 2 - -> C 10 H 17 .OH - -* C 10 H 16 terpin terpineol dipentene. Terpineol and terpin are converted by hydrogen chloride to a crystal- line dichloride 27 C 10 H 18 C1 2 melting at 50, which is identical with the dihydrochloride made from dipentene. 26 The position of the double bond in terpineol was suggested by Wallach 28 and confirmed by later researches of Baeyer 29 and others, particularly on the ground of the 21 Philippine J. Sci. 1908, 52. 22 Vavon, Compt. rend, Ufl, 997 (1909). 23 Sabatier & Senderens, Compt. rend. 156, 184 (1913). 2 Wolff, Ann. 39*, 86 (1912). "Bcr. 12, 848 (1879) ; J. Chem. Soc. S5, 287 (1879). 28 Ann. 230, 258 (1885). 2T List, Ann. 67, 367 (1848). 28 Ann. 277, 105 (1893). "Ber. 26, 2558 (1893). 320 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS relations between terpineol and carvone (see below). Wallach's pro- posed constitution of terpineol was CH, This proposed structure was open to the objection that such a sub- stance could not decompose with loss of water to give a hydrocarbon containing an asymmetric carbon atom, whereas Wallach himself had shown that dipentene was a mixture of the two active d and Mimonenes. A little later, 1894, Wagner 30 published his well-known paper "On the oxidation of cyclic compounds," in which he modified Wallach's terpineol structure to H C OH which is abundantly supported by other evidence published since, of which probably the most convincing is W. H. Perkin, Jr.'s synthesis of terpin, terpineol and a series of related substances. 81 *Ber. 27, 1636 (1894). 81 8th Int. Cong. Appl. Ctiem. VI, 224 (1912). THE PARAMENTHANE SERIES 321 CH 3 The group R C OH is readily synthesized by the action CH 3 of zinc methyl, and particularly easily by the action of magnesium methyl iodide on acid chlorides, esters or methyl ketones and Perkin accordingly carried out his synthesis as follows: (1) Pentane-1, 3, 5-tricarboxylic acid was heated with acetic anhy- dride when C0 2 and water were eliminated and cyclohexanone-4-car- boxylic acid was formed. H0 2 C.CH 2 CH 2 CH 2 CH 2 >CH.C0 2 H > OC< >CH.C0 2 H H0 2 C.CH 2 CH 2 CH 2 CH 2 (2) The ester of this acid was then treated with methyl-magnesium iodide in the usual manner when the ketone group reacts much more readily than the C0 2 R group ; the resulting hydroxy acid was converted to the corresponding bromide by heating with hydrobromic acid and the resulting tertiary bromide digested with sodium carbonate, remov- ing HBr to give l-methyl-A 1 -cyclohexene-4-carboxylic acid. CH 2 CH 2 HO CH 2 CH 2 OC< >CH.C0 2 R > >C< >CH.C0 2 R CH 2 CH 2 CH 3 CH 2 CH 2 Br CH 2 CH 2 \ /CH.CH 2 -> >C< CH.C0 2 H-CH 3 C >CH.C0 2 H CH 3 CH 2 CH 2 / \CH 2 CH 2 1 (3) On treating the ester of this acid with methyl-magnesium iodide an almost quantitative yield of a-terpineol was obtained. CHCH 2 CHCH 2 CH 3 CH 3 C ' CH.C0 2 R -* CH 3 C CH C OH CH 2 CH 2 CH 2 CH 2 CH 3 a-Terpineol is characterized by its fine odor of lilacs and is manu- factured in comparatively large quantities by decomposing terpin hydrate or terpin (made from pinene) by means of phosphoric acid. Unless specially purified the commercial product is liquid at ordinary temperatures and contains a little p-terpineol 82 melting when pure at "Stephan & Helle, Ber. 55, 2147 (1902). 322 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS 32, and the liquid terpinenol-1. 33 In nature, only a-terpineol ap- pears to be formed. Terpineol is the major constituent of commercial long leaf pine oil 34 made by distilling the wood with steam. A good commercial pine oil will show 75 per cent distilling between 211 and 218. It has proven particularly valuable for the concentration of low-grade copper ores by the flotation process. The stability of ter- pineol in the presence of alkali renders it valuable in the perfuming of soaps. Commercial a-terpineol melts at 35, boils at 217-218 at 760 mm., at 104-105 and 10 mm., has a density of 0.935 to 0.940 at 20 15 and a refractive index n 1.4808. 35 An exceptionally pure speci- men of a-terpineol, made by Wallach 36 by the action of dilute sulfuric acid on homonopinol, showed melting-point 37-38, boiling-point 218-219, and fa] D 106 (in 16.34 per cent solution in ether). The highest optical activity observed for natural terpineol is [a] D + 95 9' (from bitter orange peel oil) and [a] D 27 20' shown by a specimen of Z-terpineol from linaloe oil. A specimen of synthetic /-terpineol 37 showed [a]Q 117.5. Commercial terpineol is soluble in 9 volumes of 50 per cent, in 3 volumes of 60 per cent and in about 2 volumes of 70 per cent alcohol. When free from water it is miscible in petroleum ether. The nitrosochloride of d or Z-terpineol melts at 107-108, that of i-terpineol at 112-113; the corresponding nitrol- piperidine compounds melt at 151-152 and 159-160 respectively. By shaking terpineol with an excess of concentrated hydriodic acid the dihydroiodide C 10 H 18 I 2 is formed, melting at 77-78. Terpineol, being a tertiary alcohol, is very easily decomposed with loss of water when heated with potassium acid sulfate or oxalic acid; even acetic anhydride partially decomposes it, on heating, forming dipentene. Phenylisocyanate yields a phenylurethane, 38 the inactive form melting at 113. The a-naphthylurethane 39 melts at 147-148. As with most tertiary alcohols, the phenyl and naphthylurethanes are difficult to prepare, partial decomposition of the alcohol, with the formation of water, causing the conversion of phenyl isocyanate to diphenyl urea. In preparing the isocyanate it is advisable to separate the crystals of 83 Wallach, Ann. 362, 269 (1908). "Teeple, J. Am. Chem. Soc. 30, 412 (1908) ; Met. & Chem. Eng. 11, 247 (1913). " 8 Gildemeister, "Aetheriache Oele," Ed. 2, Vol. I, 394. *Ann. 360, 89 (1908). 87 Ertschikowsky, Bull. aoc. chim. (3) 16, 1584 (1896). 88 Wallach, Ann. 275, 104 (1893). "Schimmel & Co. Semi-Ann. Rep. 1906 (2), 33. THE PARAMENTHANE SERIES 323 diphenyl urea, which first form, by taking up the liquid portion in a little perfectly dry ether. The mixture should be permitted to stand three or four days protected from the moisture of the air. Good yields of terpinyl hydrogen phthalate and succinate can be obtained by allowing an excess of the alcohol to stand with the acid anhydride at temperatures below 100 . 40 The d-glucosides of both a and p-terpi- neol have been made by treating |3-tetra-acetylbromoglucose in ethyl ether with an excess of the terpene alcohol in the presence of silver carbonate. The acetyl groups are removed from the product by means of barium hydroxide. The resulting glucosides are rapidly hydrolyzed by hot dilute acids but are very slowly split by emulsin. 41 Glucosides of citronellol and of dihydrocarveol were prepared in the same manner. It is practically certain that glucosides of many terpene alcohols exist in nature, in addition to the few, such as coniferin, which are known to occur in nature. Tertiary alcohols appear to be capable of forming addition products with chromic acid, when the alcohols are dissolved in an inert solvent and shaken with concentrated chromic acid or the solid crystals. In the case of a and p-terpineols the addition products are liquid and unstable. 42 The hydration of a-terpineol to terpin hydrate can be beautifully demonstrated, as for a lecture experiment, by dissolving terpineol in 5 parts of 80 per cent phosphoric acid at 30, allowing to stand a few minutes and then diluting about six times with cold water, when within a few minutes a bulky matted mass of crystals of terpin hydrate will form. 43 The reaction is less complete with 60 per cent sulfuric acid and Aschan 44 has shown that 45 per cent sulfuric acid, shaken with pinene for 16 hours at + 1 gives a yield of 53 per cent terpin. The ease of making terpin hydrate from commercial long leaf pine oil has been pointed out by Teeple. 45 The synthetic a-terpineol made by Perkin was converted into terpin hydrate by agitating with dilute sulfuric acid; by heating with potas- sium hydrogen sulfate dipentene was obtained, thus completing the synthesis of these three important substances. Additional proof of the constitution of terpin was furnished by Perkin and Kay, who showed "Pickard, Lewcock & Yates, Proc. chem. 8oc. 29, 127 (1914) 41 Haemaelaeinen, Biochem. Z. 49, 398 (1913) '' 42 Wienhaus, Ber. 47, 322 (1914). 43 Prins, Chem. Abs. 1917, 2773. 44 Chem. Aba. 1919. 2759. 45 Loc. cit. 324 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS that when ethyl cyclohexanone-4-carboxylate was treated with large excess of methyl-magnesium iodide, terpin is formed. l.S-terpin. This synthesis proves conclusively the position of the two hydroxyl groups in terpin. Terpin Hydrate is not known to occur' in essential oils. It melts at 116-117 and readily loses a molecule of water on heating or on standing over sulfuric acid to form terpin, melting-point 104, whose structure is shown above. Terpin exists in two stereo-isomeric forms of the cis and trans type. 46 Terpin derived from terpin hydrate by dehydration is the cis form; the trans form, melting-point 156-158, is made from frans-dipentene dihydrobromide and silver acetate, jfrans-terpin does not crystallize with water of crystallization. Other evidence for the structure of limonene, a-terpineol and terpin had already shown their constitution with reasonable certainty. Wagner had proposed his now accepted constitution of a-terpineol largely to overcome the objection made against Wallach's constitution, that the latter could not give an optically active hydrocarbon, limo- nine, on dehydration. By oxidizing a-terpineol first with perman- ganate and then with chromic acid, Wallach 47 obtained a series of oxidation products finally resulting in homoterpenylic and terpenylic "Baeyer, Ber. 26, 2865 (1893) ; 29, 5 (1896). Van't Hoff, in 1874, had predicted that cyclic compounds of this type would be found to exist in two stereo isomeric forms. 227, 110 (1893). THE PARAMENTHANE SERIES 325 acid and he showed that these changes could readily be interpreted by Wagner's constitution for a-terpineol, i.e., H, '-OH ^CH, a-terpineol VMV\r\CL Homoterpenylic acid Terpenylic acid Although the constitution of these important acids was worked out with reasonable certainty 48 their synthesis by Lawrence and Simon- sen * 9 removes all question as to their structure. The constitution of these acids also has a very direct bearing on the constitution of pinene and Simonsen's synthesis is therefore mentioned in outline as follows, Wallach (Ann. 259, 322 [1890]), had suggested the above constitution for ter- penylic acid and its correctness has been confirmed by the work of Fittig (Ann. t88, 176 [1896]), Mahla and Tiemann (Ber. 29, 928 [1896]), and Schryver (J. Ohem. Soc. 63, 1338 [1893]). * 9 J. Chtm. Soc. 75, 527 (1899) ; 91, 184 (1907). 326 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CCLR (a) \ \/ CH io CH, C0 2 R HC CH CH 3 MgI C0 2 R H 2 C CH C-OH /\ C0 2 R H 2 CH, CH, C==0 CH, ^CH 3 When the latter ester is hydrolyzed with hydrochloric acid, C0 2 is eliminated; the ester of the resulting p-acetyladipic acid yields homo- terpenylic acid (ester) when treated with magnesium-methyl iodide, as in (a). The above work, together with Perkin's synthesis, conclusively proves the position of the double bond in a-terpineol. The position of the other double bond in limonene was shown by reference to the con- stitution of carvone and dihydrocarveol. The nitrosolimonene of THE PARAMENTHANE SERIES 327 Tilden and Shenstone 50 proved to be identical with carvoxkne, 51 from which it follows that at least one and perhaps both double bonds in limonene and in carvone are similarly situated. On reduction of carvone one double bond is saturated yielding dihydrocarveol, which on oxidation first by permanganate, followed by chromic acid and sodium hypobromite, finally yields 2-hydroxy-para-toluic acid, which, when the intermediate products are also considered, indicates that the double bond in dihydrocarveol is in the side chain. 52 Dihydrocarveol /H CH, NaOB OH 2-hydroxy- para-toluic acid The above facts make clear the relations between limonene, a-ter- pineol and terpin, and the dihydrohalogen derivatives obtained from all three, i.e., 80 J. Chem. Soc. SI, 554 (1877). 81 Goldschmidt & Ziirrer, Ber. 18, 2220 (1885). "Wallach, Ann. 275, 110 (1893) ; Tiemann & Semmler, Ber. 28, 2141 (1895). 328 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS , W by ^5 acetate and hydrolysis The Constitution of Carvone. The conversion of carvone to carvacrol, and carvoxime to carva- crylamine were early observed and, though not understood, served to call attention to the probability that carvone, carvoxime and limonene were para-menthane (l-methyl-4-iso propylcyclohexane) derivatives and that the oxygen atom in carvone occupied position (2) . Wagner 53 with his usual perspicacity, proposed a constitution for carvone in 1894, which has proven correct. He based his deductions upon the results obtained by Best 54 and by Wallach 55 on oxidizing carvone, although the constitutions of the oxidation products they obtained were not then definitely known. Terpenylic acid can be obtained from carvone in the following manner, the ring being broken at two points to give acetic acid as one of the oxidation products. CH a CO,H 4- o = c CO 2 H / CH 2 CH 2 \ / X i H Ber. 7, 2270 (1894). CH 8 CH 2 OH "Ber. *7, 1218 (1894). Ber. 27, 1496 (1894). THE PARAMENTHANE SERIES 329 C0 2 H. H 2 CH 2 C\ A H (by replacing OH by Br and reducing) CH 3 ' CH 3 The relation between carvone and limonene is very well shown by Wallach's 56 conversion of terpineol to carvone by removing hydro- gen chloride from terpineol nitrosochloride by means of caustic alkali and then boiling the resulting oxime with acids, thus hydrolyzing the oxime to the ketone and simultaneously removing the original hydroxyl group. :H, -Cl .^ >N.OH -p-menthadiene was made in the following manner. Fara-toluic acid was reduced by sodium and alcohol to l-methylcyclohexane-4-carboxylic acid which on bromination, fol- lowed by removal of HBr by sodium carbonate or quinoline in the usual manner, yielded l-methyl-A 3 -cyclohexene-4-carboxylic acid the ester of which yields A 3 -p-menthenol (8) when treated with mag- 77, 120 (1893). 330 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS nesium methyl iodide. On digesting this menthenol with potassium acid sulfate A 8>8(9) -p-menthadiene is formed. CH 3 h 3 -p-menthenol(8) Like many substances having two double linkings in the conjugated position, this menthadiene reacts with bromine to form a dibromide CHBr CH 2 Br in which the double bond has shifted to the > C = C < CH 2 CH 8 position. Also, as contrasted with limonene, this menthadiene is capa- ble of combining with only one molecule of HC1 or HBr, these products being liquid. The same behavior toward bromine and HBr and HC1 is shown by the ortho and mea-menthadiene derivatives containing conjugated double bonds, i.e., CH 5 CH } THE PARAMENTHANE SERIES 331 Another synthesis of A 3>8(9) -p-menthadiene was developed by Perkin in collaboration with Wallach. 57 In this synthesis 1-methyl- cyclohexane-4-one is condensed, by the Reformatsky reaction (zinc and a-bromopropionic ester) , to the oxy acid. The oxy acid loses water, when digested with acetic anhydride, and the resulting unsaturated acid decomposes further when distilled, losing C0 2 . The semicyclic hydrocarbon was then converted into its nitroso- chloride and this by eliminating hydrogen chloride by alkali yields an oxime which was hydrolyzed in the usual manner to the ketone. CH 3 H, h*-p-menthenol(8) The ease with which this teriary alcohol is decomposed with loss of water to form the A 3<8(9 >-p-menthadiene is worthy of note; shaking "Ann. 374, 198 (1910). 332 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS with 1 per cent sulfuric acid at warm temperature effects this change. The products obtained by these methods are, of course, optically inactive and therefore, to obtain A 3 -p-menthenol (8) of high optical activity, Perkin selected natural pulegone as his original material. As is well known, pulegone decomposes on heating with alkali to give l-methyl-cyclohexane-3-one, which in this case showed [a]j^ + 8. By treating this ketone with sodium amide and carbon dioxide l-methyl-cyclohexane-3-one-4-carboxylic acid was formed which was dehydrated yielding d l-methyl-A 8 -cyclohexene-4-carboxylic kicid of high optical activity [a]j)+ 150.1 l-methyl-h 3 -cyclohexene-4- -carboxylic acid, [a] - + 150.1 The following physical properties of A s - 8 < 9 >-p-inenthadiene were noted by Perkin and Wallach: boiling-point 184-185, d 0.858, 20 n 1.4924 from which the molecular refractivity is 46.02, calculated for C 10 H 16 /= 2 is 45.24 showing the exaltation due to the conjugated position of the double bonds. THE PARAMENTHANE SERIES 333 Terpinolene and the Terpinenes. The constitution of the terpinenes has been a matter of consider- able controversy but researches of recent years, particularly the work of Wallach, has solved the puzzle in a very satisfactory manner. Til- den, Armstrong and others had studied the action of mineral acids on turpentine or pinene, also limonene and the alcohols, terpineol and terpin, but the chief result of their investigations was to the effect that a new terpene, C 10 H 16 , was probably formed. It was not defi- nitely characterized either by physical constants or chemical deriva- tives, and it was given a variety of names. In 1885 Wallach 58 applied his tetrabromide method, which he had used in the identification of limonene and dipentene, to the high-boiling fraction boiling from 179-190, obtained by the action of alcoholic sulfuric acid on tur- pentine. From this fraction he prepared a new tetrabromide, C 10 H 16 Br 4 melting at 116-117, thus proving the existence of a new terpene, which he named "terpinolene." In the mixture of hydrocar- bons resulting from the action of alcoholic sulfuric acid on terpin hydrate, dipentene, phellandrene or cineol, he showed that the fraction boiling from 179-182 contained what he termed "terpinene." This fraction did not give a crystalline tetrabromide. A fairly good yield of terpinolene was also obtained by the action of hot concentrated oxalic or formic acid on a-terpineol, the terpineol being slowly dropped into the acid and the terpinolene removed by distilling with steam as fast as formed, as otherwise the new terpene underwent further change. Being influenced by the constitution for a-terpineol which he had pro- posed, Wallach 59 suggested the following structure for terpinolene, 'CM, Although von Baeyer had accepted Wallach's a-terpineol formula, he nevertheless advanced his now generally accepted A 1>4(8 > structure for "Ann. tin, 283 (1885) ; 230, 262 (1885). "Ann. 227, 145 (1893). This formula was later put forward by Harries, Ber. 95, 1169, as the constitution of terpinene (q.v.). 334 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS terpinolene. Baeyer 60 considered that this structure was indicated by the formation of blue nitroso derivatives. The position of the second double bond was indicated by its relation to a-terpineol and its optical inactivity. In view of the fact that other isomeric hydrocarbons are also simultaneously formed and that radical changes in constitution are known to be brought about by heating with acids, these considera- tions would have little weight were it not for other evidence. Baeyer made terpinolene by brominating limonene dihydrobromide and treat- ing the resulting tribromide with zinc dust; saponification of the resulting mono-acetate yielded a new terpineol, y-terpineol, melting- point 69-70. Baeyer had shown that other substances containing the group >C = C(CH 3 ) 2 , for example, tetramethylethylene, give blue nitroso compounds. He had also shown that generally dibromides in which the two bromine atoms are in the 1.2 position are reduced by zinc dust and acetic acid to the olefine group. Br "Ber. 7, 436 (1894), CH, XH, y-terpineol M.-P. 69-70 CH, X CH, terpinolene THE PARAMENTHANE SERIES 335 Also since the nitrosochloride of y-terpineol was also blue, Baeyer reasoned that this terpineol contained the >C = C(CH 3 ) 2 group as in terpinolene. Dehydrating agents were shown to convert y-terpineol to terpinolene. Semmler 61 has made a terpinolene of unusual purity by reducing terpinolene tetrabromide (which can be isolated from impure material) by treating with zinc dust in alcohol instead of acetic acid. Semmler's terpinolene had the following physical properties: Sp. Gr. 20, 0.854, 20 n ^-1.484, boiling point at 10mm. 67-68, boiling-point at 760mm. 183-185, optically inactive. Heat converts terpinolene to dipentene and acids partially convert it to a mixture containing a and yterpi- nenes, dipentene and terpinolene. Terpinolene is apparently not found in nature; Clover 62 reported it in Manila elemi, but Bacon, 63 working on material from the same source, was unable to confirm this. Terpi- nolene is obtained as a by-product in the manufacture of commercial terpineol and is occasionally found as an adulterant of lavender and other oils. Terpinolene has been synthesized from nopinone and methyl nopinol (q.v.). 64 The Terpinenes. The formation of "terpinene" by the action of alcoholic sulfuric acid on pinene, terpin hydrate, cineol, dipentene and phellandrene has been mentioned in connection with terpinolene, which is also formed in the reaction mixture. Its occurrence in nature was first noted in the case of oil of cardamoms by Weber. 65 It has been reported to occur in Manila elemi, but according to the researches of Clover and Bacon, 66 on over a hundred specimens of authentic material, different individual trees yield an oleoresin containing either limonene or phel- landrene of remarkable purity ; the commercial oil accordingly contains both of these terpenes, but since Clover and Bacon worked with fresh material it is probable that the terpinene reported by others was formed from phellandrene by the action of formic or other acids developed by air oxidation. "Terpinene" has usually been identified by means of the nitrosite melting at 155. According to Schimmel Ber. 42, 4644 (1909). 93 Philippine J. Sci. 1907, 1. Philippine J. Sci. A. 1909, 93. "Wallach. Ann. 856. 244 (1907). **Ann. 238, 107 (1887). Philippine J. Soi. 1909, 93. 336 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS & Co., 6T Wallach's a-terpinene occurs in coriander oil and y-terpinene in ajowan, lemon and coriander oil. Terpinene was frequently confounded with dipentene, in the earlier literature. Both yield crystalline addition products with halogen acids, but although the melting points of the corresponding dihydro- halides are very close together, a marked lowering of the melting- point results when the two are mixed. The terpinene dihydrohalides are best prepared from sabinene. "Terpinene" Dipentene C 1 H M .2HC1 51-52 50 CioH w .2HBr 58-59 64 CioH w .2HI 76 77 Wallach 68 has shown that with aqueous alkali the dihydrochloride is converted into a terpin melting at 137 and not identical with cis or trans- 1.8-terpin which corresponds to limonene dihydrochloride. Wallach reasoned that if the new terpin was a di-tertiary alcohol, as its behavior indicated, it must have the structure of 1 . 4-dihydroxy- p-menthane, if it was in fact a para-menthane derivative. It was further distinguished from ordinary or 1.8-terpin by the formation of an oxide differing from cineol (eucalyptol). It should be men- :a l.8-Terp in M.-P. cis- 102-105 H 2 O trans- 156-158 1.8 cineol L-t-terpm )H M.-P. cis- 116-117 trans- 137 H 2 1.4 cineol Gildemeister & Mtiller, Wallach-Festschrift 1909, 443. Ann. 350, 157 (1906) ; 356, 200 (1907). THE PARAMENTHANE SERIES 337 tioned that on dehydrating ordinary terpin, one of the products is the oxide, cineol, which has been shown to be an oxygen ether, or oxide, the oxygen atom of which is attached to the carbon atoms 1 and 8. The difference between the two terpins and their oxides are, as sug- gested by Wallach. Additional evidence that this is the constitution of the new terpin was furnished by its synthesis 69 from sabina ketone. (Sabinene and sabina ketone, q.v., had already been shown to contain an unstable tri-carbon ring, as shown.) CH 3 CH 3 -OH JL-OH Reduction of ascaridol (q.v.) also yields 1.4-terpin. It will be evident that 1.4-terpin, or the corresponding dihydro- chloride, can conceivably decompose with loss of two molecules of water or hydrochloric acid, respectively, to give four different para- menthadienes, i.e., H 3 XH 3 a-terpinene $-terpinene y-terpinene terpinolene Of the hydrocarbons represented above IV has the constitution which had already been shown to be that of terpinolene. A hydrocarbon having the structure represented by II, A 3 - 1 ( 7 >-p-menthadiene, has been synthesized by Wallach 70 from sabina ketone and found to yield a tetrabromide melting at 154-155, and boiling at 173-174. It is Wallach, Ann. 357. 64 (1907). "Ann. 357, 68 (1907) ; 362, 287 (1908). 338 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS therefore not present in ordinary "terpinene" boiling at 179-182 and which does not yield a crystalline tetrabromide. The constitution of "terpinene" therefore resolves itself into I or III, or a mixture of these two hydrocarbons. On oxidizing ordinary terpinene with permanganate, a a'-dioxy- a-methyl-a'-isopropyladipic acid, melting-point 189 is formed. This acid can only be derived from A 1 - 3 -p-menthadiene. OH a-terpinene erythrite adipic acid derivative The structure of this acid has been proved beyond question by its synthesis, in the following manner, CH 3 CH 3 CO C OH. H 2 C +2HCN > H 2 C CN H 2 C H 2 C CN CO C\ I | OH. C 3 H 7 C 3 H 7 followed by hydrolyzing the nitrile to the adipic acid derivative. The structure of this acid can also be shown by following the oxidation of terpinenol- (4) CH 3 THE PARAMENTHANE SERIES which also yields this adipic acid derivative. 71 Further oxidation yields dimethyl-acetonylacetone whose dioxime melts at 137. Additional evidence of the presence of A 1 3 -p-menthadiene in "ter- pinene" has been furnished by the conversion of terpinene nitrosite to carvenone, first by reducing the nitrosite in alcohol solution by zinc, 72 and later with particularly good yields, by reducing the nitrolamine by zinc and acetic acid. 73 CH 3 H .OH terpinene nitrolamine carvenone oxime carvenone In the investigation of terpinene from various different sources or made in different ways, it was observed that those specimens which give good yields of the a a'-dioxy-a a'-methylisopropyladipic acid, melting at 189, also give good yields of the above crystalline nitro- site. 7 * On the other hand it has been noted that specimens which yield little or no nitrosite also yield very little of the adipic acid derivative melting at 189. The presence of A 1 4 -p-menthadiene in terpinene was made prac- tically certain by the discovery of Gildemeister and Miiller 75 that one of the oxidation products was isopropyl tartronic acid. This specimen of terpinene was isolated from ajowan oil and Gildemeister and Miiller were unable to isolate a crystalline nitrosite, nor could they detect the adipic acid derivative among the oxidation products. All experience with terpinene, particularly as brought out by Wallach's extensive investigations, 76 indicate that "terpinene" is a mixture containing vary- ing proportions of the two isomers, which Wallach designates as a and y-terpinene (see above). Auwers has studied the terpinene question from the standpoint of their physical properties, particularly the re- 71 Wallach, Ann. 362, 266 (1908). "Amenomiya, Ber. 38, 2730 (1905). "Wallach, Ber. 40, 582 (1907). "Wallach, Ann. 374, 229, 250 (1910). 7 5 Schimmel & Co. Semi-Ann. Rep. 1909 (2), 16. 'Ann. 362, 261, 285 (1908) ; 368, 13 (1909) ; 574, 224 (1910). 340 CHEMISTRY OF THE NON-BENZENOID- HYDROCARBONS fractive index. In accord with Wallach's findings, Auwers showed that a terpinene having a particularly high refractive index, as would be expected in the case of a-terpinene, also gave a very large yield of crystalline nitrosite and the adipic acid derivative. Wallach believes that the preparation of strictly pure terpinenes, terpinolene and the phellandrenes is impossible. 77 Carvenene is probably an impure a-terpinene. It was so named by Semmler, 78 who prepared it from carvenone by the action of PC1 5 followed by reduction. According to Semmler, alcoholic sulfuric acid converts carvenene to an isomeric hydrocarbon isocarvenene which he considers is identical with (3-terpinene. Auwers does not agree with Semmler as to the supposed purity of carvenene and claims that it is not identical with a-terpinene which Auwers made from 0-cresol. Henderson and Sutherland 79 have made what appears to be mainly a-terpinene by reducing thymohydroquinone to 2 . 5-dioxy-p-menthane and decomposing this with removal .of two molecules of water. Their a-terpinene showed a boiling-point of 179, specific gravity about 0.840 and refractive index 1.4779. Pickles 80 isolated a terpene "origa- nene" from the volatile oil of Origanum hirtum, which he considers is probably a-terpinene. Crithmene. This terpene is mentioned in connection with the terpinenes since it yields terpinene dihydrochloride on treating with hydrogen chloride. It is contained in the volatile oil of Crithmum maritimum. 81 Its boil- 20 ing-point is 178-179, specific gravity, 0.8658, n~- 1.4806. It does not yield a crystalline tetrabromide, the nitrosite melts at 89-90 and the nitrosate at 104-105. It yields two nitrosochlorides which can be distinguished by their different crystal forms, although the melting-points are very close together, 101-102 and 103 and 104. The discoverers have suggested that crithmene is probably A 1 ** 7 *-** 8 )- p-menthadiene. "Fairly pure a-terpinene has been synthesized by Wallach, Ber. tf, 2404 (1909). 19 J. T Chem. Soc. 97, 1616 (1910). 80 J. Chem. Soc. 93, 862 (1908). 81 Fransesconi & Sernagiotto, AtU accad. Lincei. 1913. 231, 312 : Delepine & Belsunce Butt. soc. chim. (4) S3, 34 (1918). THE PARAMENTHANE SERIES 341 CH 2 H 2 C " CH 2 H 2 C CH 2 \ / C CH 3 CH 3 crithmene The Oxides. 1.8-Cineol, 1 . 4-Cineol, Pinol and Ascaridol. These oxides of the terpene series are usually described without reference to other intramolecular ethers, or oxides, of the non-ben- zenoid hydrocarbons. Unfortunately the number of such organic oxides, to use the customary term, which are known, is so small that it is not possible to show such close relationships between those which happen to have been first prepared from the terpenes, and those pre- pared from other non-benzenoid cyclic or open chain hydrocarbons, as might be desired. In the first place it may be noted that the ethylene oxide ring is considerably less stable than the tri-carbon ring in cyclopropane and its derivatives. Thus ethylene oxide reacts with water on heating to give glycol and this reaction is catalyzed by a trace of a mineral acid. 82 CH 2 CH 2 OH + H 2 - , CH 2 OH. Oxides of this type have been carefully studied in the case of the oxides of ethylene, propylene, the butylenes, amylenes and hexylenes. They react with hydrogen chloride with considerable energy, forming the chlorohy drins ; with ammonia to form the corresponding amino alcohols, with nascent hydrogen to give alcohols and with a variety of other substances, as, for example, sodium malonic ester, Henry, C&mpt. rend. W, 1404 (1907). 342 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CH 2 Na CH 2 ONa | >o+ >C.(C0 2 C 2 H 5 ) 2 >| CH 2 H CH 2 CH(C0 2 C 2 H 5 ), The greater instability of the ethylene oxide ring as compared with the tri-carbon ring is brought out in the case of isobutylenoxide (CH 3 ) 2 C CH 2 , which reacts with water to form the glycol merely on shaking together with water at ordinary temperatures. (The group (CH 3 ) 2 C< in cyclopropane usually results in greater stability.) 83 As with carbocyclic rings a great increase in stability is noted, com- pared wth ethylene oxide, when the oxide ring contains five or six atoms. Thus diethylene oxide CH 2 CH 2 is the principal CH 2 CH 2 product resulting when ethylene glycol is distilled with 4 per cent aqueous sulfuric acid; in the same manner, but with smaller yields, 1.8-terpin yields ordinary cineol and 1.4-terpin yields 1.4-cineol. Diethylene oxide, in contrast to ethylene oxide, yields a series of well- defined addition products 84 of the type which Baeyer suggested were derivatives of quadrivalent oxygen; the sulfate, C 4 H 8 2 .H 2 S0 4 melts at 101, the dibromide C 4 H 4 2 Br 2 melts at 60, etc. When the con- stitutions of these two substances are compared it is apparent that the reason for the greater stability of diethylene oxide is the hexatomic ring. CH 2 CH 2 CH 2 I >0 | | CH 2 CH 2 CH 2 diethylene oxide. In connection with the stability of diethylene oxide its comparatively high melting-point + 9.5, and boiling-point 100-101 are significant. If the valence directions of quadrivalent oxygen are in the directions of the four corners of a regular tetrahedron as we assume to be the case in the carbon atom, then we should expect a close parallel with non-benzenoid carbocylic substances, as regards stability and ease of formation and rupture. Derick and Bissell 85 have called attention "Ingold, J. Chem. Soc. 119, 305 (1921). "Paterno & Spallino, Gaez. chim. Ital. 37 (1), 106 (1907) ; Faworski, Chem. Zentr. 19ffl (I), 16. Am. Chem. Soc. 1916, 2478. THE PARAMENTHANE SERIES 343 to the fact that trimethylene oxide CH 2 CH 2 CH 2 O, which contains four atoms in the ring is markedly more stable than ethylene oxide. [The stereochemistry of oxygen has been very little studied. There would seem to be no reason why substances containing asymmetric oxygen cannot be resolved into optically active forms, possibly by Pasteur's method of mechanically picking out crystals of opposite hemihedral development.] It has been noted 86 that 1.4 and 1.5- glycols and their oxides behave in a manner markedly different from the 1.2-glycols. The former are readily converted into their oxides of five and six membered rings respectively by heating with 60 per cent sulfuric acid and these oxides are quite stable to water; in fact, they can be heated with water to 200 several hours without forming glycols. It should be pointed out that their behavior is strictly parallel to the behavior of the better known oxides, cineol and pinol. Up to the present time the only oxides whose synthesis has been attempted with the idea of industrial utilization are the simpler 1.2 oxides, i.e., ethylene oxide for the manufacture of phenylethyl alcohol by the Grignard reaction, 87 C 6 H 5 MgBr + (CH 2 ) 2 C 6 H 6 CH 2 CH 2 OH, and other organic preparations, 88 and the 1.2 oxides of butylenes and amylenes which have been proposed as solvents for cellulose esters. 89 However, the 1.4 and 1.5 oxides are quite stable and should prove industrially valuable if they could be made economically. Tetramethylene oxide. CH., CH 2 boils at 67 and is I >0 PITT OUT U1 2 v-;.tl 2 easily soluble in water. It is not reacted upon by water at 150 but is attacked by fuming hydrobromic acid. 1 A-Oxidopentane. CH 2 CH CH 3 is a liquid of agree- 1 C H 2 CH 2 able ethereal odor boiling at 77-78 and soluble in 10 parts of water at ordinary temperatures. It can be made by heating the 1.4-glycol with 60 per cent sulfuric acid or by the action of caustic alkali on the M Petrenko-Kritschenko & Konschin, Ann. Sl,2, 51 (1905). 87 Grignard, Compt. rend. 136, 1260 (1903) ; Altwegg, U. S. Pat. 1,315,619. Accord- ing to the writer's experience this reaction gives yields 75 to 80 per cent, of the theo- retical ; ethylene oxide is best made by solid caustic soda on nearly anhydrous ethylene chlorohydrin. 88 Cf. Soc. chim. du Rhone, Brit. Pat. 128,552; 128,553; 128,554 (1919), for aminobenzoic acid derivatives; Brit. Pat. 128,911 (1919) for chloroethyl esters; Brit. Pat. 128,908 (1919) for ethanolamines and aminophenol ethers. "Walker, U. S. Pat. 972,952. 344 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS corresponding chlorohydrine. The dimethyl derivative is a product of the action of sulfuric acid on diallyl, CH 2 CH = CH 2 CH 2 CH CH 3 > I >0 boiling-point 90-92 CH 2 CH = CH 2 CH 2 CH CH 3 1 .6-Oxidopentane (Pentamethylene oxide). 90 CH 2 CH 2 CH 2 CH 2 CH 2 can be obtained by heating 1 . 5-dibromopentane with water at 100 or from the glycol by the sulfuric acid method. The oxide boils at 81-82. The same methods of preparation have proven successful in the conversion of 1.4 and 1.5-dihydroxy-n.hexane 91 to the corre- sponding oxides. Most of the glycols containing six or more carbon atoms are quite difficult to prepare and since interest in them is so narrow, they are not described here. The above examples, however, are given in support of the general thesis of the present volume, that the chemistry of the so-called hydro-aromatic hydrocarbons is ration- ally a part of the larger division of non-benzenoid hydrocarbons. * Cineol is the name given by Wallach and Brass 92 to the substance C 10 H 18 0, boiling-point 172, which they isolated from the volatile oil of wormseed, "Oleum cince" from Artemisia maritime, L. It was also shown that "cajeputol" from cajeput oil and "eucalyptol" were identi- cal with cineol. Gladstone had shown that cineol could be distilled over metallic sodium without change and Hell and Ritter 93 obtained an addition product with hydrogen chloride and accordingly suggested that the oxygen was bound as in ethylene oxide, but recognized that cineol is much more stable than ethylene oxide. The following additive compounds have been prepared from cineol, (C 10 H 18 0) 2 .HC1; (C 10 H 18 0) 2 .Br 2 ; C 10 H 18 O.Br 2 ; (C 10 H 18 0) 2 .I 2 ; C 10 H 18 O.HBr; C 10 H 18 O.H 3 P0 4 ; C 10 H 18 O.H 3 As0 4 , 93 also well crys- tallized products with hydroferricyanic and hydroferrocyanic acids, a and (3-naphthol, 94 iodol and resorcin. This property is utilized for the detection and quantitative estimation of cineol, methods based 90 Hochstetter, Monatsh. 23, 1073 (1902). "Franke & Lieben, J. Chem. Soc. Abs. 1913. I, 491. "Ann. 225, 291 (1885). M Merck, German Pat. 132,606. "Henning, German Pat. 100,551; Chem. Zentr. 1899 (1), 764. THE PARAMENTHANE SERIES 345 upon the reaction with phosphoric acid and with resorcin bejng most favored for quantitative estimation.' 5 Composition of Product" Melting-Point OoHuO.HBr ......................................... 56. doH.80.CJ.NH. (iodol) ..... > ........................ 112. (C 10 H M O) 2 .C 9 H 4 (OH) 2 (resorcin) ...................... 80. (CioH 18 ),.CH4(OH),(hydroquinone) .................. 106.5 aHsOH ..................................... 8. e^OH.CH, (orthocresol) ................... 50. . thymol ...................................... 4.5 Baeyer 97 first pointed out that these addition products, which are generally decomposed easily into their original constituents, are prob- ably derivatives of quadrivalent basic oxygen. Baeyer regarded the comparatively stable compound of ethyl ether and magnesium-alkyl X halides as "oxonium" compounds of the constitution (C 2 H 5 ) 2 0< MgR but more recent investigations point to the structure C 2 H 5 MgX C 2 H 5 R which was proposed by Grignard. When cineol is used as a reaction medium instead of ether, the reaction of magnesium, ethyl iodide and cineol takes place with almost explosive violence unless the cineol is diluted with benzene. 98 When cineol is added to a solution of mag- nesiumethyl iodide in ethyl ether, the ethyl ether is displaced and the cineol compound, (C 10 H 18 0) 2 MgC 2 H 5 I, is precipitated. When this compound is decomposed by dilute acids cineol is almost quantitatively regenerated, but if the complex is heated to 170-190 and then care- fully decomposed by cold dilute acids, a-terpineol is formed. If the conditions are reversed and magnesium-ethyl-bromide is poured into cineol, an oil is produced the products of hydrolysis of which have not been investigated. l-Ethyl-p-menthanol(8) should be formed if the re- action product, like many other Grignard ether complexes which have been investigated, is capable of decomposing in several ways. In the case of ethylene oxide and phenyl magnesium bromide, phenyl ethyl alcohol is the principal product. Cineol reacts with acetic anhydride in the presence of metallic chloride ZnCl 2 or FeCl 3 to form terpinyl acetate and terpin diacetate." 8 Cf. Parry, "Chemistry of Essential Oils," Ed. 3. Vol. I, 321, Vol. II, 256; Gildemeister, "Aetherische Oele," Ed. 2, Vol. I. 547. "Belluci & Grass!, Chem. Zentr. 1914 (1), 884. " Ber. S4, 2679 (1901) ; 35, 1201 (1902). M Pickard and Kenyon, J. Chem. Soc. 91, 896 (1907). Knoevenagel, Ann. 402, 111 (1919). 346 CHEMISTRY OP THE NON-BENZENOID HYDROCARBONS The constitution of cineol is clearly indicated by its formation from 1.8-terpin by oxalic acid and other dehydrating agents, by the formation of dipentene dihydrochloride from cineol by the action of hydrogen chloride (in acetic acid), by the absence of a double bond and the absence of a carbonyl group. Cineol was formerly regarded as the 1.4 oxide but when a-terpineol was shown to be A^p-men- thenol (8) and ordinary terpin to be the 1.8 derivative, the constitu- tion attributed to cineol was revised to accord with these facts. Hot permanganate solution oxidizes cineol to cineolic acid, which sub- stance retains the 1.8 oxide grouping. cineolic acid, M.-P. 197 ( Cineol occurs in the volatile oil of many species of eucalyptus and the commercial valuation of eucalyptus oils is usually determined by their cineol content. Commercial oils are derived from a number of different species and earlier references to the essential oil of Eucalyp- tus globulus undoubtedly refer to the mixed oil from several species. The genuine oil of Eucalyptus globulus contains 50 to 70 per cent of cineol, the balance being d-a-pinene, and minor percentages of a ses- quiterpene alcohol which has been named globulol, an unidentified terpene and very small proportions of butyric, valeric and caproic aldehydes. R. T. Baker and H. G. Smith have made a systematic survey of the various species of eucalyptus and the volatile oils derived from them, the results of which have been published in a comprehensive monograph 10 and in a series of papers in the Journal of Proceedings, Royal Society of New South Wales. Baker and Smith find 58 species yielding oils whose principal constituents are cineol and pinene, 14 in which pinene and sesquiterpenes are the chief compo- nents, 9 which contain notable percentages of a new aldehyde "aro- "x>"Eucalypts'of Tasmania, " 1912; J. Soc. Chem. Ind. 32, 710 (1913). THE PARAMENTHANE SERIES 347 madendral," 101 33 species which yield oils characterized by phellan- drene and piperitone (q.v.), and several other species whose oils differ markedly from those above mentioned. 102 Cineol crystallizes on chilling the fraction boiling at 174-178 from good eucalyptus oil and its isolation in this way is comparatively easy, though naturally not quantitative. A ketone derivative of cineol has been made from a-terpineol by first preparing a-terpineol nitrosochloride, treating this with hydrox- ylamine, thus replacing Cl by NH.OH. and hydrolyzing the resulting product by water. 103 :H, CH, CH, CH } - tH 3 a-terpineol 1.4-Cineol: This isomer of ordinary cineol has not been found in nature, but was discovered by Wallach as one of the products of the dehydration of 1.4-terpin by oxalic acid. 104 It boils at 172 but does not crystallize on cooling to 15. It is quite stable to per- manganate solution. Ascaridol is one of the most remarkable organic compounds known. It was discovered by Schimmel & Co. in 1908 105 in the volatile oil of American wormseed or Chenopodium ambrosioides, L., var anthel- minticum. It was found to contain two atoms of oxygen and on heat- ing to 130-150 decomposes with explosive violence. On reducing by Paal's method, four atoms of hydrogen are taken up and one of the stereoisomeric forms of 1.4-terpin are formed. This 1.4-terpin, melting at 116-117, is regarded by Wallach 106 as the cis form. (The identity of this terpin was clearly shown by its conversion to terpinene dihydrochloride and by its decomposition to 1.4-cineol.) 101 J. Proc. Roy. Soc. N. S. W. 1900, 1. 102 An excellent review of the eucalyptus oils is given in Parry, Essential Oils," Ed. 3, Vol. I, pp. 319-358. 103 Cusmano & Linari, Gazz. & (1), 1 (1912). 104 Ann. 392, 62 (1912). 105 Reports, 1908 (1), 108. 10 *Ann. S92 t 59 (1912). 'Chemistry of 348 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The absence of double bonds and hydroxyl or carbonyl oxygen, its peroxide-like properties and its relation to 1.4-terpin, indicates, in Wallach's opinion, a peroxide structure. Nelson 107 first showed the peroxide character of ascaridol but suggested that the two peroxide oxygen atoms connected the 2.5 positions. By treating with ferrous sulfate solution in the cold two glycols C 10 H 18 3 , melting at 62.5-64 and 103-104 are obtained. Heating the higher melting glycol with dilute sulfuric acid yields p-cymene. According to Wallach's consti- tution for ascaridol, the erythritol C 10 H 20 4 , melting at 128-130 should yield a, a'-methyl-isopropyl-a, a'-dihydroxyadipic acid, and Nelson obtained an oxidation product of this acid, i.e., 2-methylhep- tane-3.6-dione. By acid permanganate Nelson has split the acid C 10 H 16 6 , from the lower melting glycol, to 2-methyl-heptane -3.6- dione. Nelson suggests the following relationships. CH ascaridol erythritol glycol 1 .4-cineolic acid Ascaridolic acid, which possesses the structure of a 1. 4-cineolic acid, was resolved by Nelson by means of its cinconidine salt to the d. and I. forms. The physical properties of ascaridol are, boiling-point 83 under 20 5 mm. pressure, sp. gr. (20) 1.008, [a] 4 14 and n_ 1.4731. Pinol: The resemblance of pinol to the two cineols is indicated by its chemical behavior and its methods of preparation. Thus terpi- neol dibromide, 108 on treating with aniline or alcoholic alkali loses one molecule of hydrogen bromide to form an unsaturated bond and loses a second molecule of HBr after the fashion of the bromo-hydrines and chlorohydrines to form the oxide pinol. 1OT J. Am. CTiem. Soc. 33, 1404 (1911) ; 35, 84 (1913). 108 Wallach, Ann. 253, 254, 261 (1889). THE PARAMENTHANE SERIES 349 a-terpineol dibromide pinol Like cineol the oxide ring is quite stable but the double bond reacts normally, being oxidized by permanganate to terebinic acid, adds bromine to form pinol dibromide, melting-point 94, gives a nitrosochloride, etc. The odor of pinol resembles cineol and camphor. With mineral acids it readily yields cymene. Its physical properties 20 20 are as follows, boiling-point 183-184, d 0.942, n -1.4714. When pinol is treated with hydrogen bromide in acetic acid solu- tion, the oxide ring is broken, as with other oxides, and so-called pinol hydrobromide is formed, which on treating with alkali yields "pinol hydrate," H OH .OH ^ pinol pinol hydrate The dibromide of pinol hydrate on treating with alkali yields a dioxide, H 350 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS On treating pinol tribromide with zinc in acetic acid pinolone is formed. This reaction has been a puzzling one and is worth noting as an instance, now established beyond question, of the conversion of the six carbon ring to the cyclopentane ring. Wallach 109 has proven that dihydropinolone is acetylisopropylcyclopentane. Hydrogenation of pinolone yields the saturated ketone dihydropinolone, the constitu- tion of which has been shown both by decomposition studies and by synthesis, to be 1 . 3-acetylisopropylpentanone. The synthesis is of interest as employing reactions of quite general application. H, AH, N.OH i AH, dihydropino lone The fact of the formation of the cyclopentane ring is thus clearly established. Wallach suggests an explanation of this change which is based upon the conversion of the glycid group to a ketone group, many examples of which are known, particularly as shown by recent researches of Darzens. 110 H H pinol tribromide 109 Ann. 384, 193 (1911). Compt. rend. 152, 443, 1105 (1911). THE PARAMENTHANE SERIES 351 CO -CH or pinolone As mentioned above, chlorohydrines yield oxides when treated with caustic alkali. Slawenski m has made pinol and pinol hydrate by the action of caustic potash upon the chlorohydrin of terpineol (the latter substance being made by the direct addition of hypochlor- ous acid to terpineol). The formation of pinol and pinol hydrate shows that the chlorine is in position 6. An oxide of the diterpene series has been discovered in Java, citro- nella oil. The oxide, C 20 H 34 0, boils at 182^-183 (at 12 mm.). It contains two double bonds and is reduced by hydrogen and platinum black to C 20 H 38 0: it yields a monohydrochloride melting at 107.5. When citronellal is heated with oxalic acid, one of the products is an isomeric oxide C 20 H 34 0. 112 Other Alcohols of the Paramenthane Series. It is evident that by the partial decomposition of ordinary terpin or terpin hydrate, four isomeric terpineols can theoretically be pro- duced, i.e., H. a- terpineol fi-terpineol I Inactive M.-P. 35 M.-P. 32 \Active M.-P. 37-38 y-terpineol b-terpineol M.-P. 69-70 (unknown) Chemik. Polski, 15, 97 (1917) ; Chem. Ate. IS, 887 (1919). 112 Semmler, Ber. 47, 2077 (1914) ; Spornitz, Ber. tf f 2478 (1914). 352 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS a-Terpineol has been described in the foregoing pages on account of the importance of its constitution to the structure of the related substances of this series. $-Terpineol is a constituent of commercial terpineol 118 made by the partial decomposition of terpin hydrate. It has not been found in nature. Its physical properties are as follows, melting-point 32-33, boiling-point 209-210, d 150 in supercooled state 0.923, 20 1.4747. Its phenylurethane melts at 85, the nitrosochloride at n _ 103 , 11 * the nitrolaniline derivative at 110 and the nitrolpiperidine derivative at 108. (3-Terpineol was made synthetically by Perkin, 115 in a manner which clearly confirms its structure. Incidental to the synthesis of a-terpineol, described above, a small amount of hydroxyisopropyl- cyclohexane-4-one was formed, which ketone was dehydrated in the usual manner. When the resulting unsaturated ketone was treated with magnesium-methyl iodide, |3-terpineol was formed. ^-terpineol The decomposition of (3-terpineol by oxidation has been studied by Stephan and Helle 116 and by Wallach. 117 One of the products of oxidation, l-methyM-acetyl-A^cyclohexene has been utilized by Wallach 118 for the preparation of a number of saturated and un- & C * Semi ' Ann> Re P- 1901 m Wallach, Ann. 345, 128 (1903). 118 /. Chem. Soc. 85, 659 (1904). 118 Z/oc. cit. 111 Ann. S24, 88 (1902). Ann. klk, 202 (1918). Stephan & Halle, Ber. 35, 2147 THE PARAMENTHANE SERIES 353 saturated alcohols. Thus magnesium-ethyl iodide yields homo-a- terpineol - CH 8 -/ >C(OH)< X / CH Nascent hydrogen reduces the carbonyl group and when the result- ing product is treated with dilute sulfuric acid 1.8-dihydroxy-l- methyl-4-ethylcyclohexane, melting at 94, is formed. Two other 1.8-terpins were described in the same paper, i.e., addition of water to homo-a-terpineol yields HO - CH 3 v C(OH)< CH 3 - C 2 H melting-point 65-67, and secondly hydration of // - v C 2 H 5 -C(OH)< N - / CH 3 by dilute acids gave the corresponding 1.8-terpin, crystallizing with one molecule of water, melting at 75-76. y-Terpineol has not been found in nature, but is one of the minor reaction products when ordinary 1.8-terpin is partially decomposed by oxalic or phosphoric acids. It was prepared by Baeyer incidental to his investigation of the constitution of terpinolene (q.v.). It is characterized by its relatively high melting-point, 69-70, and its blue nitrosochloride melting at 82. On heating with about one volume of concentrated formic acid terpinolene results. 119 & 3 -p-Menthenol(8), was made synthetically by Perkin and Wal- lach 12 incidental to their synthesis of A 3<8(9) -p-menthadiene. It melts at 41, boils at 205 with slight decomposition and yields a phenylurethane melting at about 128, the melting-point varying somewhat with the rate of heating owing to decomposition at this temperature. A 2 -p-Af enthenol ( 1 ) , is related to the phellandrenes (q.v.). It is "'Wallach, Ann. 368, 11 (1909). 120 Ann. 37+, 198. 354 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS easily decomposed by dehydrating agents to give a-phellandrene. It was made synthetically by Wallach 121 by the action of magnesium methyl iodide on 4-isopropyl-A 2 -cyclohexenone. It boils at 92 (10mm.). The Terpinenols: Several terpene alcohols are known which can be derived from the 1.4-terpin of the terpinene series, and are hence called terpinenols. As in the case of 1.8-terpin, noted above, loss of one molecule of water from 1.4-terpin can theoretically lead to the formation of four isomeric alcohols. CH 3 OH Terpineol-4 Terpinenol-1 unknown y-Terpineol Of the substances indicated above, it will be noted that IV is identical with y-terpineol. Terpinenol-4 is found in nature in a number of essential oils, juniper, Ceylon cardamon, nutmeg and zedoary. 122 It is formed by the hydration of sabinene by cold dilute sulfuric acid. 123 The physi- cal properties of optically active terpinenol-4 are as follows, boiling- 19 point 209-212, d 0.9265, [a] +25 4', n 1.4785. It has iy D D not been obtained in crystalline form. Both terpinenol-4 and terpi- neol-1 give terpinene dihydrochloride when treated with hydrogen chloride in glacial acetic acid, and also give 1.4-terpin on hydra ting with cold, dilute sulfuric acid, although this hydration takes place much more slowly than with a-terpineol. The two terpinenols are however distinguished by their oxidation products, 124 r . 111 Arm. 559, 283 (1908). 122 Terpinenol 4 is present in comparatively large proportions in one of the Formosan lauracese closely resembling the camphor tree : Schimmel & Co. Semi-Ann. Rep. 1915 (2), 42. "Wallach, Ann. S60, 94, 97 (1908) ; 362, 279 (1908). * Wallach, Ann. 356, 210 (1907). THE PARAMENTHANE SERIES 355 (a) terpinenol-4 1.2.4. -trioxy-p-menthane h?-carvenone (fr-p-menthenol 4) M.-P. 116-117 (H 2 0). terpinenol-1 1, 8, 4- trioxy-p-menthane ^-menthenone 125 M.-P. 120-121 Terpinenol-1 occurs in commercial terpineol and can be isolated from the forerunnings obtained when large quantities of crude terpin- eol are distilled with steam. It has also been synthesized by the action of magnesium methyl iodide on A 3 -4-isopropyl cyclohexenone. I Its physical properties 126 are as follows, boiling-point 208-210, d lo i 0.9265, and n _ 1.4781. Dihydrocarveol, A 8 < 9) -p-menthenol (2) , is found in nature in oil of \ caraway, associated with carvone. Its importance in the work of \ determining the constitutions of limonene and related substances has already been pointed out. It can be made by the reduction of carvone by sodium and alcohol, 127 or by the reduction of carvoxime to dihy- drocarvylamine and treating the latter with nitrous acid. Complete 125 A'-P- men thenone is characterized by its boiling-point 235-237, the oxime melting at 77-79 and its semicarbazone melting at 210. 1M Wallach, Ann. S56, 218 (1907). ^Wallach, Ann. 275, 111 (1893). 356 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS reduction yields carvomenthol. 128 The relationships between the more important substances of this carvone series may be indicated as fol- lows, "a carvacrol CH carveol dihydrocarveol carvomenthol CH, carvotanacetone dihydro carvone tetrahydro- carvone Dihydrocarveol is characterized by oxidation by chromic acid to dihy- drocarvone, whose oxime melts at 88-89 (inactive) or 115-116 (active form), and by its physical properties, 129 boiling-point 224, 20 d 0.9368, and n 1.4836. By shaking with 3 per cent sulfuric lo D acid it is hydrated to 2 . 8-dioxy-p-menthane (M.-P. 112-113). Carveol is one of the principal products of the oxidation of limo- nene by air in the presence of water. 130 It boils at 226-227, yields a phenylurethane melting at 94-95, and a phthalate melting at 136-137. Its methyl ether (by CH 3 ONa on 1,2,8-tribromomen- thane) boils at 208-212. Isopulegol, A 8(9) -p-menthenol(3). This alcohol is not found in nature but is readily formed from citronellal by the action of acids; 128 Cf. Henderson & Schotz, J. Chem. Soc. 101, 2565 (1912). ""Schimmel & Co. Semi-Ann. Rep. 1905 (1), 51. ""Blumann & Zeitschel, Ber. tf, 2623 (1907). THE PARAMENTHANE SERIES 357 when oils containing citronellal are heated with acetic anhydride this aldehyde is almost quantitatively converted into the acetate of iso- pulegol. Heating with sodium ethoxide 131 converts isopulegol to citronellol and also decomposes it to acetone and methylcyclohexanol (3). Isopulegol is characterized by oxidation to the corresponding ketone, isopulegone, 132 which ketone yields an oxime melting at 121 (active) and 140 (inactive form). The acetate boils at 104-105 under 10 mm. pressure. Isopulegol boils at 91 under 13 mm. pressure, 0.9154, n 1.4729. Menthol, para-menthanol(3). This saturated alcohol is a com- mon article of commerce. Up to the present it has not been manu- factured synthetically but is obtained from oil of peppermint, par- ticularly Japanese peppermint. Peppermint has been under cultivation in Japan since about 900 A.D., and in Europe for a period probably equally long, and, as is usually the case with cultivated plants, there are numerous varieties and the number of distinct species is as yet an open question. Yet, with the exception of oil of spearmint 133 which is characterized by considerable proportions of carvone, the various peppermint oils owe their characteristic flavor and aroma to menthol and the corresponding ketone menthone. The menthol occurs in these oils partly free and partly in the form of esters of acetic and other acids, and on chilling part of the free menthol crystallizes from the oil. The melting-point of menthol is 42.5. According to F. E. Wright, 134 ordinary /.menthol crystallizes in four different forms, a, p, y and 6, only one of which, the a-form, is stable between zero and its melting-point, 42.5. The other three forms are monotropic and have lower melting temperatures, i.e., 35.5 (3, 33.5 y> and 31.5 6; all the unstable forms invert finally into the stable a-form. Previous work of Schaum, 135 , Pope, 136 Hulett 137 and others had shown the existence of at least two unstable forms, which investigations were confirmed and extended by Wright. Menthol boils at 215-216, 45 d. 0.881 and when derived from peppermint is laevorotatory. Pick- 181 Schimmel & Co. Semi-Ann. Rep. 1913 (2), 91. 1S2 Wallach, Ann. S65, 251 (1909). 183 In the United States oil of spearmint is derived from Mentha viridis (Mentha spicata). The peppermint oil industry is very fully described by Parry, "Chemistry of Essential Oils," Ed. 3, Vol. I, 205-231. 134 J. Am. Chem. Soc. S9, 1515 (1917). 188 Ann. 308, 39 (1899). /. Chem. Soc. 75, 463 (1899). m J. Phj/s. Chem. 28, 667 (1899). 358 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS ard and Littlebury 1S8 made menthol by the catalytic hydrogenation of thymol, and by resolution of the brucine salt of the monomenthyl- phthalic ester ^.menthol was obtained [a] 48.76 and d.menthol [ C < J* ui the possible number of optically inactive isomerides is increased to four and the number of optically active isomerides to eight. By the hydrogenation of thymol in the presence of catalytic nickel, which had been carried out by Brunei, 141 a mixture containing 60 per cent of "menthols," 30 per cent of menthones and 10 per cent of unchanged thymol is obtained. After removing the thymol, the alcohols were separated by phthalic anhydride, in the usual manner. The semi- < barbazones of the mixture of menthones proved to have widely dif- ferent solubilities in alcohols, one nearly insoluble in cold alcohol J. Chem. Soc. 101, 109 (1912). , . 1905 " 8 nomenclature was use <* by Aschan, "Chemie d. alicyklischen Verbindungen, ' M Compt. rend. 140, 252 (1905), et seq. THE PARAMENTHANE SERIES 359 and melting at 217, previously described by Wallach, 142 and a more soluble one melting at 158. Fractional crystallization of the hydro- gen phthalate esters yielded two pure products, one melting at 177 and one melting at 130. The ester melting at 177 on hydrolysis yields an optically inactive menthol melting at 51, "neomenthol," previously isolated by Beckmann. 143 Hydrolysis of the menthyl hydrogen phthalate melting at 130 yields an inactive menthol melt- ing at 34, which can be resolved, by means of the cinchonine or brucine salt, to ordinary i.menthol, melting-point 43, and d.menthol, melting-point 43. These relations are evidently parallel to and of the same nature as those between the borneols, isoborneols and cam- phors (q.v.). The following diagram summarizes the findings of Pickard and Littlebury, Thymol, menthone semicarbazone M.-P. 158 isomenthone semicarbazone M.-P. 217 i-Menthol, M.-fr. 34 (hydrogen phthalate, M.-P. 130) i-neo-menthol, M.-P. 67 (hydrogen phthalate, M.-P. 177) 1 -menthol d. menthol d. leomenthol 1 . neomenthol M.-P. 43 M.-P. 43 oil oil [a] 48.76 [a] +48.15 [a] +19.6 [a] 19.6 D D D D \ \ / / \ ^ oxidation / \ / \ / \ K Z-menthone d-menthone By the hydrogenation of pulegone by the Sabatier and Senderens method, Haller and Martine 144 obtained two menthols which evi- Ann. 263, 272 (1908). 143 J. prakt. Chem. (2) 55, 30 (1897). "'Comfit, rend, itf, 1298 (1905). Haller's 0-pulegomenthol is probably d-neo- menthol and his a-pulegomenthol evidently belongs to the isomenthol series. 360 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS dently are identical with ^.menthol and d.neomenthol described above. By electrolytic reduction of menthone in solution in about equal parts of 94 per cent alcohol and 75 per cent sulfuric acid a yield of 25 per cent of the theory of menthol has been reported. 145 Beckmann 146 has described an isomenthol melting at 78-81, obtained by the reduction of a specimen of d.menthone made by inverting Z.menthone by 90 per cent sulfuric acid. By passing synthetic "menthol (from thymol) over copper at 300, it is converted to menthone. 147 Substituted menthols of the general constitution indicated below have been made from menthone by the Grignard reaction, and by zinc and alkyl halides, CH 3 CH, CH CH 2 OH H 2 CH C< I R CH 3 -CH-CH 3 Magnesium cyclohexyl bromide 147 gives the cyclohexyl derivative melting at 92 together with a cyclohexylmenthene boiling at 265. Methyl iodide and magnesium give chiefly the methyl tertiary alcohol; but ethyl iodide and magnesium or zinc yields chiefly a hydrocarbon C 12 H 22 . 148 Allyl iodide and zinc 149 give the expected allyl derivative, boiling-point 246-252. Hallers' reaction has been employed by Boedtker 15 for the prepa- ration of alkyl derivatives of menthone in which the alkyl groups are in position 2. Thus ethyl iodide and sodium amide, reacting with menthone, give 2-ethyl-p-menthanone(3) from which by reducing in moist ether by sodium the 2-ethyl menthol was made. The methyl, n-propyl, isoamyl and benzyl derivatives were prepared in the same manner. The stereochemistry of substances such as menthone and menthol is somewhat involved and has led to some confusion in the description " 8 Matsui, J. S&c. Chem, Ind. 19Z1, 162A. *Ber. 42, 846 (1909). " T Murat, J. Chem. Soc. Abs. 100 (1), 890 (1911). ' Ab8 ' m (1)> 474 ^Bul. Soc. chim. (4) 17, 360 (1915) ; Haller, Compt. rend. 156, 1199 (1913). Dimethylmenthone -- > dimethylmenthol, a liquid boiling at 245-247. THE PARAMENTHANE SERIES 361 of these substances and their derivatives. It should be pointd out that aside from cis-trans relationships discussed above, menthone pos- sesses two and menthol three asymmetric carbon atoms. CH 3 CH 3 I H | H C/ C/ H 2 C CH 2 H 2 C CH 2 H 2 C C = H 2 C *C< \*/ \*/ OH. C H C\ I I H CH CH Theoretically, therefore, menthol should be capable of existing in four spatial configurations of which each would have two optical antipodes and one racemic form. Kursanov 151 finds that when ordi- nary menthol is treated with phosphorus pentachloride, in benzene solution, a mixture of menthyl chlorides is obtained, which are of markedly different stability to caustic alkali. The stable chloride reacts with magnesium in ether to give a menthene, para-menthane, a crystalline dimenthyl (C 10 H 19 ) 2 melting at 105-106, [a] 51.42 (which is identical with the dimenthyl obtained by the action of metallic sodium on this chloride) and when the menthyl-magnesium chloride thus formed is treated with carbon dioxide, a crystalline menthanecarboxylic acid results. The unstable menthyl chloride yields a liquid menthanecarboxylic acid and the crude menthyl chloride consequently must contain two stereoisomers. Kursanov concludes that only the carbon atom (3) attached to the hydroxyl group in the original menthol is inverted by the reaction. When Lmenthone, corresponding in spatial configuration to ordinary Z.men- thol, is treated with 90 per cent sulfuric acid, it is partially inverted to d.isomenthone and according to Beckmann 152 the asymmetric car- bon atom involved in the change is the one to which the isopropyl group is attached. When the potassium derivative of menthol is heated with phenyl bromide or iodide the products are benzene and menthone, but in the 151 J. Chem. Soc. Abs. 108 (1), 420 (1915). 162 Ber. W, 846 (1909). 362 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS presence of finely divided copper the reaction gives a high yield of menthylphenyl ether. When this ether is heated with concentrated hydrochloric acid to 170 this ether is isomerized to menthyl phenol. Menthol is relatively very stable and its esters can accordingly be easily prepared. The benzoate, melting-point 54, can be prepared by heating with benzoic acid in autoclave to 170. Although the benzoate cannot be made by mixing the alcohol with benzoic and sulfuric acids, the phenyl acetate and phenyl propionate can be made in this way. 153 In studying the action of esters on magnesium alkyl halides Stadnikow 154 found that magnesium menthyl iodide reacts with ethyl acetate to give a practically quantitative yield of menthyl acetate; ethyl propionate gave an 80 per cent yield of menthyl pro- pionate and ethyl benzoate gave 64.6 per cent menthyl benzoate. Tschugaeff 155 prepared a series of menthyl esters by acting upon men- thol by various acid chlorides in slight excess. A great many esters of menthol have been employed in the study of optical activity. The following table gives the boiling-point or freezing-point of a number of menthyl esters. Density 20 0.9359 4- 0.9185 0.9184 0.9114 0.9074 0.9033 0.9006 0.8977 Ester Melting-Point Boiling Formate PS -Point (15mm.) (25mm.) (20mm.) (20mm.) (20mm.) (20mm.) .(20mm.) Acetate 108. 118. 129. 141. 153. 165. 175. Propionate Butyrate Valerate n . Hexoate n . Heptoate n . Octoate Dimenthyl oxalate 67. -68. 62. 110. 132. 168. 79. 83. 61. 83. 62.' V ' Dinienthyl succinate Menthvl-H-phthalate Dimenthy] phthalate Dimenthyl muconate .... Dimenthyl , y-hydromuconate Dimenthyl, a, |3-hydromuconate Dimenthyl adipate Menthyl piperate Monthyl ft, y-hydropiperate Menthyl a, |3-hydropiperate . . . Dimenthyl malonate 263. 270. Menthyl glutarate 240-3 248. -252 257-9 254.6 256. -258 Menthyl pimelate Menthyl suberate Menthyl azelate .... Menthyl sebacate Menthyl benzoate 54.5 . rend. 155, 1! 1113 (1915) 163 Senderens & Aboulenc, Compt 154 /. RUBS. Phys.-Chem. Soc. lit, Ber. 31 t 360 (1898). 254 (1912). ; J. Chem. Soc. A6. M 975 (1915). THE PARAMENTHANE SERIES 363 Ester Melting-Point Boiling-Point Menthyl phenylacetate Menthyl phenylproprionate . . . 28.5 Menthyl acetoacetate 36. Menthyl propyl acetoacetate Menthyl phenyl acetoacetate . . 69. 205.5 216. 154. 162. 131-3 (25mm.) (25mm.) (10mm.) (8mm.) (O.lmm.) Density 20 The freezing-point curves of menthyl mandelate indicate "the existence to a considerable extent, of undissociated racemate in the liquid state." 186 Ketones of the Para-Menthane Series. There are two saturated ketones and one known diketone derived from para-menthane. Menthone: The stereo isomers of menthone have been discussed in connection with menthol. Ordinary menthone isolated from oil of peppermint and regenerated from the semicarbazone (melting-point 184), boils at 208, has a density 0.894, and refractive index 1.4496. The oxime of menthone is of interest on account of the fact that it undergoes a Beckmann rearrangement, 157 with rupture of the cyclo- hexane ring, to give menthoneisoxime (by treating with concentrated sulfuric acid). By dehydrating agents the isoxime yields mentho- nitrile, which on reduction yields menthonylamine and from this amine, by heating the nitrile with water, menthocitronellol is formed, as indicated in the following outline, H 3 t+H =NOH I C=0 :H ' ^CH-^H CH, CjH 7 C^OH "CH wit menthocitronellol "Findley & Hickmans, J. Chem. Soc. 91, 905 (1907). ' Wallach, Ann. 296, 124. 364 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The reduction of Z.menthone oxime yields a single menthylamine, but by heating menthone with ammonium formate a mixture of crystalline d. and Z.formylmenthylamines C 10 H 19 NH.COH is formed, which can be separated, and which yields two isomeric menthylamines of the cis and trans types and from which a series of isomeric derivatives have been prepared. 158 When menthone oxime is heated to 220 with caustic potash 159 thymol is formed together with about 45 per cent of the open chain acid, (CH 3 ) 2 CH. (CH 2 ) 3 CH(CH 3 ) .CH 2 CO 2 H. Menthone has been synthesized by Kotz 16 from (3-methyl-a-iso- propylpimelic acid in two ways (1) distillation of its calcium salt with soda lime, and (2), intramolecular condensation of its ester by sodium ethoxide after the manner of the acetoacetic ester condensa- tion, CH, CH, CH 2 CH CH 2 C0 2 R CH 2 CH C0 2 R i, CH 2 CH CH 2 CH, CH CO C 3 H 7 Another synthesis of menthone by Wallach and Churchill 161 is of interest. Reformatsky's reaction was employed for the condensation of l-methylcyclohexane-4-one with a-bromoisobutyrate. The unsatu- rated acid, derived from the resulting ester, yields A 4 < 8 >-p-menthene. Substances containing a double bond in this position rearrange under the influence of dilute acids, the double bond shifting to the ring, as in the case of terpinolene. In the present instance i.A 3 -p-menthene is formed from which i.menthone was synthesized. CH, CH, 1M Wallach, Arm. Stf, 67 ; Stf, 259. 188 Wallach, Ann. 389, 185. 180 Ann. S57, 209. 191 Ann. 360, 26 (1908). THE PARAMENTHANE SERIES CH 3 CH 3 365 i dil.acid & 3 -p-menthene N.OH H H J ' * nitrosochloride k*-p-menthenone ,,H 7 i. -menthone In a study of the chlorination and bromination of cyclic ketones Kotz 162 finds that the halogen always substitutes in an ortho position to the carbonyl group. In the case of menthone, 4-chloro or 4-bromo- menthane-3-one are obtained, from which aniline or aqueous potas- sium carbonate followed by dehydration by oxalic acid, yields A 4 -p- menthenone. Similarly carvomenthone (p.menthane-2-one) yields l-chloromenthane-2-one. Wallach has studied the conversion of 2.4-dibromomenthone to the cyclopentane hydroxy-carboxylic acid. (Cf. "Rearrangements.") Oxidation of menthol by chromic acid 163 gives the ketonic acid (CH 3 ) 2 CH.CO.CH 2 CH 2 CH.CH 2 C0 2 H. The same acid is obtained CH< by treating menthone with amyl nitrite and hydrochloric acid and hydrolyzing the nitrosomenthone thus formed. 164 This ketonic acid is also formed by the air oxidation of menthone, in sunlight. Sun- light in the absence of oxygen, however, decomposes menthone giving a decoic acid and an aldehyde 165 which is different from Wallach's menthocitronellaldehyde, i. e., . 397; 1 (1911). Kotz effects this halogenatlon by diluting the chlorine or bromine with air and adding water and calcium carbonate to the ketone ie3 Beckrnann, J. Chem. Soc. Abs. 1896 (1), 312. 1M Baeyer & Oehler, Ber. 29. 27 (1896). 165 Ciamician & Silber, Ber. 42, 1510 (1909). 366 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS (CH 3 ) 2 CH . CH CH . CH 2 . CH (CH.) . CH 2 CH0. 165 Oxidation by potassium permanganate yields oxomenthylic acid (1) and p-methyladipic acid, and Caro's reagent gives a lactone of di- methyloctanolic acid (2) , (1) (2) CH 3 CH 3 CH, - CH - CH 2 CH 2 - CH - CH 2 I 1 I - > CO CH 2 - -CO C0 2 H. CH 2 - -CH - CH 3 CH CH 3 CH 3 CH CH 3 Menthone can be alkylated by Hallers' reaction 166 (sodium amide and alkyl halide), the alkyl group being substituted in position (2). One or both of the hydrogen atoms in the CH 2 group (2), can be replaced by alkyl groups. 167 With hydrazine hydrate, menthone reacts to form menthylidene hydrazine, which on heating with caustic potash loses N 2 and gives p-menthane. 168 Menthone condenses with formic acid esters (amyl formate and sodium) to form oxymethylene menthone. Benzaldehyde reacts slowly in the presence of alkalies, but rapidly with hydrochloric acid to form benzylidenementhone (hydrochloride melting at 140). Reduction of this compound yields benzylmenthol, M.-P. 111-112. Menthone, 169 like camphor, carvomenthone and cyclohexanone, acts as a catalyst in the combination of sulfur dioxide and chlorine to form S0 2 C1 2 . The optical inversion of Z.menthone by sodium ethylate has been suggested as a means of determining the per cent of Z.menthone in pine oils. 170 Menthone reacts normally in the Reformatsky reaction, 171 wi bromacetic ester and zinc, to give the ester of mentholacetic acid, th free acid readily losing water and carbon dioxide to give homomen- thene C n H 20 , boiling at 186-187. in . he lw Boedtker, Bull. soc. chim. Ft, 360 (1915). 167 Holler, Compt. rend. 156, 1199 (1913). 188 Kizhner. J. Russ. Phya. Chem. Soc., M, 1754 (1912). ""Cusmano, Gazz. Chem. Ital. 50 (2), 70 (1920). "Gruse & Acree, Science 44, 64 (1916). Tubandt [Ann. 377, 284 (1910] shows that the rate of inversion of menthone by acids is not proportional to the H ion con- centration. The speed of inversion is greatly retarded by water. m Wallach, Ann. 353, 313. THE PARAMENTHANE SERIES 367 CH S CH 3 CH CH H 2 C CH 2 H 2 C CH 2 > OH > homomenthene m C = H 2 C C< ' \ / \ / CH 2 C0 2 H +H 2 + C0 a CH CH C 3 H 7 C 8 H 7 Normal Menthone, l-methyl-4-propylcyclohexane-3-one. This ketone, synthesized by Wallach, 172 does not smell like ordinary men- thone, illustrating the marked influence of slight differences of con- stitution on odor, noted also in the case of unsaturated ketones similar CH a H 2 C CH H 2 C C = AH. to ionone and also in the case of artificial musk when the tertiary butyl group in Musk Bauer is replaced by similar alkyl groups. Nor- mal menthone boils at 215-217. fr-p-Menthenone, has been found in Japanese oil of peppermint 178 and in one of the Cymbopogon grass oils, C. senaarensis, Chiov. 174 19 It boils at 235-237, density _ 0.9375, [a] D 1.4875. It forms a very sparingly soluble semicarbazone melting at 212 and yields a dibromide which, by the action of aqueous caustic potash and heat, is converted almost quantitatively to thymol. Piperitone: This menthenone occurs in the essential oils from several species of eucalyptus, particularly in E. dives, the oil of which contains 40 to 50 per cent of this ketone. It occurs associated with the corresponding alcohol "piperitol." As this oil is available in large m Chem. Zentr. 1915 (2), 824 Schimmel & Co. Semi-Ann. Rep. 1910 (2), 79. * Roberts, J. Chem. Soc. 1915, 1465. 368 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS quantities it is probably only a question of a short time before men- thol, and possibly thymol, will be manufactured from this material; in fact, Smith and Penfold 175 have reported that with hydrogen and catalytic nickel menthone is formed almost quantitatively. The constitution of piperitone is not definitely known. Smith and Penfold suggest that it may prove to be identical with A A -p- menthenone. They describe it as boiling at 229-230 at 760 mm. or 106-107 at 10 mm., d 2Q0 0.9348, [a] D 40.05 and n 1.4837. It forms a semicarbazone melting at 219-220 and an oxime melt- ing at 110-111. The most characteristic derivative is the compound formed with benzaldehyde, benzylidene cU.piperitone, C 10 H 14 O.CH.C 6 H 5 , melting at 61. By oxidation, -by boiling with ferric chloride in dilute acetic acid, a yield of thymol corresponding to 25 per cent of the theory was obtained. Pulegone, A 4 * 8(9) -p-menthenone: This ketone occurs in the essen- tial oils of Mentha pulegium and Hedeoma pulegioides and .imparts its characteristic odor to oil of pennyroyal. Its physical properties, 176 are, boiling-point 224 (750mm.) or 93-94 at 869mm., d,^ 0.9405, 20 15 [ct] D 20 28' and n_ 1.48796. By reducing energetically with nascent hydrogen 177 menthol may be obtained. When reduced by sodium and alcohol, about 30 per cent of a yellow resin, C 20 H 34 2 , is formed, 178 a similar product being formed when employing the aluminum-mercury couple. 179 Paolini has separated the alcoholic reduction products (by sodium and alco- hol) and has identified ordinary ^.menthol of peppermint, a solid dmenthol melting at 88-89, boiling-point 214, [a] 11.7, and Z.pulegol. Pulegone is of special interest as furnishing an example of the conversion of the cyclohexane ring into the cyclopentane ring. When pulegone dibromide is heated with sodium methylate in alcoholic solu- tion pulegenic acid results, and when pulegenic acid is oxidized by permanganate a glycol is formed which then forms a lactone; by a pinacoline rearrangement the cyclopentane ring is enlarged to give C0 2 (loss of 1 carbon atom) and pulenone, C 9 H 15 0. 180 Pulegenic "J. Proc. Roy. Soc. N. 8. W. 5k, 40 (1920). " Gildemeister, "Die Aetherischen Oele," Ed. 2, Vol. I, 463. "'Beckmann & Pleissner, Ann. 262, 30 (1891). 178 Paolini. J. Chem. Soc. A6. 1920 (1), 171. "Harries & Roeder, Ber. 32, 3357 (1899). " Wallach, Ann. 329, 82 ; 376, 154. THE PARAMENTHANE SERIES 369 acid also decomposes with loss of one molecule of carbon dioxide to give pulegene, C 9 H 16 . H 3 CH 3 Xx CH, byKMnO< CH, f CH, > CH/ The ring in pulenone can be broken by converting the oxime into the isoxime, in the same manner as menthone, described above. Thus, heating with acetic anhydride gives the nitrile of a nonylenic acid, CH> CH 3 CH, ^C.H pidenoneoxime isoxime ll Wallach, Ann. S29, 100 (1903). nonylenic CH, CH, acid 181 nitrile 370 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Pulegene yields a nitrosochloride, melting-point 74 to 75, which on decomposing with alkali yields the oxime of a ketone, pulege- none, 182 CH 3 CH 3 CH CH H,C CH H 2 C H 2 C CH 3 CH H 2 C C = N.OH !H(CH,) : C Cl H(CH 3 ) C pulegene nitrosochloride H(CH 3 ) 2 oxime of pulegenone The corresponding saturated ketone, l-methyl-3-isopropylcyclopen- tane-2-one, is identical with camphorphorone. When a halogen, chlorine or bromine, is introduced into a ketone, in the ortho position to the carbonyl group, the resulting halogen deriva- tive is unstable. Kotz 183 noted that the bromo ketones are particu- larly unstable, fuming in the air and decomposing rapidly when warmed. Wallach 184 showed that the dihalogen ketones react rapidly with aqueous alkali at room temperature and that cyclohexanones could be converted into cyclopentanones by evaporating the alkaline solution, thus obtaining an oxy acid which on distilling with lead peroxide and sulfuric acid yields the pentanone, C=0 182 Wallach. Ann. 327. 133 (1903). Ann. 400, 47 (1913). Nachr. Goettingen 1915, 244; J. Chetn. Soc. Alia. 110 (1), 487 (1916). THE PARAMENTHANE SERIES 371 By a similar series of reactions methylcyclohexane-2-one gives 1 - methycy clopentane - 2 - one ; 1 - methylcyclohexane - 3 - one gives l-methycyclopentane-3-one; the ketone 1 . 3-dimethylcy clohexane-5- one yields 1 . 3-dimethylcy clopentane-2-one ; from 1 . 3 . 3-trimethyl- cyclohexane-5-one there was obtained 1.3.3-trimethylcyclopentane- 5-one. CH, CH, CH CH /\ /\ H 2 C C = H 2 C C = H 2 C CH 2 H 2 C CH 2 Y H 2 CH 8 CH 8 in CH H 2 C CH 2 H 2 C CH 2 H 2 C C = H 2 C- -C = C H 2 CH 3 CH 3 CH CH H 2 C CH 2 = C CH 2 I I CH 3 > I I CH 3 = C C< H 2 C C< \/ CH 3 CH 8 H 2 Menthone similarly gives l-methyl-3-isopropy Icy clopentane-2-one (dihydrocamphorphorone) , and carvomenthone gives the same oxidation products. Menthone can also be converted into pule- genone by slightly modifying the above procedure. Wallach finds that menthone dibromide first yields two isomeric substances C 10 H 16 2 , one of which proved to be buchu camphor. 372 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS menthone buchu camphor The dibromide of buchu camphor, when treated with aqueous alkali, gives an oxy-acid which on distillation yields pulegenone. 185 The hydrazine derivative of menthone yields para-menthane when heated with caustic potash. Pulegohydrazine, however, yields the bicyclic hydrocarbon carane, under these conditions. 186 CH 3 pulegone carane This substance, also called diosphenol, is found in nature in the essential oil of buchu leaves. It is of considerable interest in that its chemical behavior gives no indication of the existence of the tau- tomeric diketo form, analogous to camphor quinone, although its for- mation from the dibromine substitution products of both menthone and carvomenthone indicate that the diketone must be an intermediate product. 187 188 Wallach, J. Chem. Soc. Als. 114, 544 (1918). 1M Kizhner. J. Russ. Phys.-Cliem. Soc., J t S ) 1132 (1911). '"Cusmano, J. Chem. Soc. Ala. 191^ (1), 303; Atti accad. Lincei (5), H, 520 (1915). CH 3 CH CH H 2 C CHBr H 2 C CH.OH Hf~\ fl ___ f\ TT/^1 f*\ /"\ 2 W V^ \J 1\J V>/ \J Y-B, Y C 3 H 7 C 3 H 7 THE PARAMENTHANE SERIES CH 3 C] 373 H 2 C H 2 C CH 3 C Br /\ CH 3 C /\ \H 2 C CH in /\ CH 3 C C = OH,C C OH ^H,C C = H 2 C C = ^ C = HC C = CHBr H 2 C CH.OH \x CH CH C 3 H T HO TT , 7 ^a 11 ? Carvone, A 6 - 8 < 9 >-p-menthadiene-2-one. The constitution of car- vone and much of the chemical behavior has been shown above, in connection with the discussion of limonene and the terpineols. Car- vone is of further interest, however, on account of its conversion to the cylcoheptane derivative eucarvone, and derivatives of the bicyclic carane series. Baeyer 188 showed that the hydrobromide of carvone gave, by loss of HBr, an isomeric ketone which he called eucarvone. Baeyer originally suggested that the constitution of eucarvone was that represented as I, below, but Wallach was able to show that eucarvone is a cycloheptane derivative II, and that Baeyer's structure for eucarvone is in fact possessed by an intermediate product in the reaction. Wallach 189 represents these reactions as follows, CH 3 eucarvone 374 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Baeyer regarded dihydro and tetrahydroeucarvone as cycloheptane derivatives and showed that the completely reduced tetrahydroeucar- vone was broken up by oxidation to (3, p-dimethylpimelic acid which in turn yields 1 . l-dimethylhexane-3-one, in the usual manner: CH, cio H 2 C C = H 2 C C0 2 H C0 2 H H 2 C CH 2 * H 2 C CH 2 - H 2 C C0 2 H. H 2 C C(CH 3 ) 2 H 2 C--C(CH 3 ) 2 H 2 C CH 2 H 2 C C(CH 3 ) 2 tetrahydro-eucarvone, (1 . 4 . 4-tri-methylcycloheptane-2-one,) I H 2 C C = H 2 C CH 2 H 2 C--C(CH 3 ) 3 dimethylcyclohexanone Eucarvone boils at 85-87 (12mm.); its density at 21 is 0.949 20 and n 1.5048. It yields a semicarbazone melting at 183-185 and a benzylidene derivative melting at 112-113. The oxime melt- ing-point, 106, may be reduced by sodium and alcohol to dihydro- eucarvylamine and, by more energetic reduction, to the saturated amine C 10 H 19 .NH 2 . Reduction of the oxime by hydrogen and pal- ladium yields the dihydro and tetrahydrooximes, melting at 122-123 and 56-57 respectively. The alcohol, tetrahydroeucarveol [1.4.4- trimethylcycloheptanol(2)], a product of reduction by sodium, boils at 216. When carvone is reduced by nascent hydrogen the double bond next to the CO group is first reduced, dihydrocarvone being A 8 < 9 >-p- menthene-2-one. Wallach, Albright and Klein 190 have made the interesting observation that when the CO group is converted to the Ann. tfS, 74 (1914). THE PARAMENTHANE SERIES 375 oxime and the resulting carvoxime then reduced by one mole of hydrogen, in the presence of PaaFs colloidal palladium, the double bond in the side chain is first reduced. In this case the oxime of carvotanacetone (A 6 -p-menthene-2-one) is formed. Vavon 191 also showed that, in the presence of platinum black (Willstatter's method) , carvone itself was reduced first to carvotanacetone. This ketone is also formed by the rupture of the three carbon ring in thujone, effected by heating. 192 Carvone boils at 230 and occurs in d. and I. forms [a] 60 D Carvone, dihydrocarveol and limonene are the principal constituents of oil of caraway, used in making the liqueur "kiimmel"; carvone is also an important constituent of dill and spearmint oils. Like citral and other substances containing the group CH = CH CO carvone reacts. to form an unstable bisulfite compound from which carvone can be regenerated by alkali, and also forms stable salts of the dihydrosulfonic acid derivative, from which the ketone cannot be regenerated. Carvone forms a crystalline compound with hydrogen sulfide, (C 10 H 14 0) 2 .H 2 S, from which carvone can be regenerated. 193 The bisulfite method is preferable for the isolation of carvone. For its identification, the following derivatives are characteristic: the d. and Z.oxime, melting-point 72, i-carvoxime melting-point 93 (when too great an excess of hydroxyamine is employed a compound of car- voxime and hydroxylamine, C 10 H 14 NOH.NH 2 OH., melting at 174- 175, is formed. It will be of interest to note that in the preparation of carvoxime a Walden inversion occurs, d. carvone yielding Z.carvox- ime and i.carvone yielding dcarvoxime. With phenylhydrazine car- vone forms a phenylhydrazone melting at 109-110 and semicarbazid forms a semicarbazone melting at 162-163 in the case of d. or {.carvone but i-carvone yields the racemic semicarbazone melting at 154-156. Jfhe original ketones are readily regenerated by warm- ing the semicarbazones with oxalic acid. Carvoxime rearranges to amido thymol when treated with con- centrated sulfuric acid, l Compt. rend. 153, 68 (1911). 192 Semmler, Ber. 27, 895 (1894). 193 Wallach, Ann. SOS, 224 (1889) ; Claus & Fahrion, J. prakt. CJiem. (2), S9, 365 (1889). The product from d. or 1. carvone melts at 210-211 ; that from i. carvone melts at 189-190. It is dimolecular in benzene but monomolecular in glacial acetic acid. Deussen, Arch. Pharm. 221, 285 (1883). 376 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS HC H 2 CH 3 C CH 3 C HC C = N.OH CH, C CH 2 HC CH, YH i V CH 3 CH 2 CH CH, CH, HC C NH 2 HOC CH Y ' AH CH 3 CH 3 amidothymol By the Grignard reaction, using methyl-magnesium iodide, a hy- drocarbon, CnH 16 , results. 194 Klages regards this hydrocarbon as 2-methyl-A 2t6 - 8(9) -menthatriene on account of the ease with which it is isomerized to 2-methylcymene by boiling with a 2 per cent solu- tion of hydrogen chloride in acetic acid. 195 The reaction is worth noting as one of the many instances of the migration of a double bond from a side chain to the cyclohexadiene nucleus to give a benzene derivative. carvone 2-methylcarveol 2-methylcymene When dihydrocarvone is similarly treated, 2-methyldihydrocarveol results which can be decomposed directly, or better by converting to the corresponding chloride, to 2-methylhomolimonene, or hydrated by the action of alcoholic sulfuric acid to 2,8-dihydroxy-2-methyl- menthane. 196 The main product of the action of magnesium-benzyl- 184 Rupe & Liechtenhan, Ber. 39, 1119 (1906). 198 Ber. 39, 2306 (1906) ; Rupe & Emmerich, Ber. 41, 1393 (1908). 189 Rupe & Emmerich, loc. cit. THE PARAMENTHANE SERIES 377 chloride is a ketone, 6-benzyl-A 8 -p-menthene-2-one, [or 6-benzyldi- hydrocarvone] , melting at 73. The a-naphthyldihydrocarvone, melting-point 150, was prepared in the same manner. 197 Semmler 198 has employed the reaction of carvone with magnesium- isoamylbromide for the synthesis of a hydrocarbon of the sesquiter- pene class. (When ether is used as a solvent in the Grignard reac- tion, instead of benzene, a large proportion of isoamyldihydrocarvone is formed.) The synthetic sesquiterpene thus prepared is monocyclic, contains three double bonds and has been named isoamyl-a-dehydro- phellandrene by Semmler. Carvone is isomerized by sunlight forming a resin and a crystal- line camphor-like substance, melting at 100, which Ciamician and Silber named carvone-camphor. 199 This substance has been carefully investigated by Sernagiotto, 200 who showed that, in sealed tubes, the sunlight causes both double bonds in carvone to combine to form a four-carbon ring. Ciamician had suggested that the isomerization of carvone resembled that of the condensation of two molecules of cin- namic acid to form truxillic acid. The work of Sernagiotto shows that the chemical behavior of the substance may be indicated as follows, CH 2 CH CH 2 CH, H, CH 9 CH carvone -CH 2 CH 3 /L CH CH 2 C = O C CH 3 CH 2 A: CH- C0 2 H carvone-camphor CH 3 ketonic acid lw Rupe & Tomi, Ber. 47, 3064 (1914) : ZelinskI, German Pat. 202,720 (1909). Ber. 50, 1838 (1917). 399 Ber. 11, 1928 (1908). *>Atti accad. Lincei 23 (2), 70 (1914) ; 26; 238 (1917). 378 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Carvone camphor melts at 100, boils at 205.5, forms an oxime melt- ing at 126-128 and a semicarbazone melting at 239. Like menthone, pulegone and camphor, carvone can be condensed with aniline by heating with aniline in the presence of a little zinc chloride [or ZnCl 2 . (C 6 H 5 NH 2 ) 2 ]. The resulting carvoneanil is an oil. 201 It is pointed out by Lapworth 202 that when the double bond in the ring of carvone combines with hydrogen cyanide, the resulting CH a CH, H NC H \C /*\ HCN YH H 2 H oV^ V_/XJ-o ' \*/ CH CH /\ CH, CH 3 CH 2 cyanodihydrocarvone should theoretically be capable of existence in four stereoisomeric forms having three asymmetric carbons, as shown. By employing Aschan's scheme of representing the section of the ring plane by a line, these four isomerides can be represented as fol- CH 3 lows, Pr representing the group C \H 2 I Pr Pr Pr Pr CN CH 3 GEL CN CN CH 3 CN CH 3 (1) (2) (3) (4) The ordinary form, melting-point 93.5-94.5, is obtained in excellent yields when an alcoholic solution of carvone and potassium cyanide is treated with acetic acid in amounts which insure the presence of a 201 Reddelien & Meyn, Ber. 53, B 345 (1920) 202 J. Chem. Soc. 89, 946 (1906). THE PARAMENTHANE SERIES 379 little excess potassium cyanide. When the addition of hydrogen cyanide takes place in hot solutions a different crystalline isomeride is produced in considerable quantity which has a rotatory power in the opposite sense to that of the substance described above [a],-. 39, instead of []pv + 15.4. The Z.isomeride exhibits slight mutarota- tion and Lapworth considers that the four isomerides may be divided into two pairs, the two members of each pair being dynamic isomer- ides at ordinary temperatures. 203 Perillic aldehyde, occurring in the essential oil of PerUla nan- kinensis, has been found to contain two double bonds in the A 6 - 8 < 9 > positions, as in limonene. By reduction with zinc dust and acetic ! acid an alcohol is produced which is identical with the so-called dihy- : drocuminic alcohol previously found in gingergrass oil. Its constitu- ! tion was shown by converting the CH 2 OH group of perillic alcohol j to the chloride which on reducing by sodium and alcohol gave Uimo- ; nene. On dehydrating the oxime of perillic aldehyde the correspond- i ing nitrile is formed which on hydrolysis yields perillic acid. Reduc- ! tion of the ester of perillic acid with sodium in absolute alcohol reduced one of the double bonds and gave dihydroperillic alcohol. 204 CHO CH 2 OH C0 2 H CH 2 OH k A A H-A HC CH 2 HC CH 2 HC CH 2 H 2 C CH, [ 2 C CH 2 H 2 C CH 2 H 2 C CH 2 H 2 C CH 2 CH CH CH CH A A A A CH 3 CH 2 perillic aldehyde 265 CH 3 CH 2 perillic alcohol CH 3 CH 2 perillic acid CH 3 CH 2 dihydroperillic alcohol 203 Lapworth & Steel, J. Chem. Soc. 99, 1877 (1911). 2 <*Semmler & Zaar, Ber. 44, 52 (1911). zee The physical properties of these substances are as follows : Perillic Perillic Perillic Perillic aldehyde alcohol nitrile acid Boiling-point (10mm.) .. 104-105 119-121 (llmm.) 116 (llmm) 164 Density 0.9617(18) 0.9640(20) 0.9439 [a] D 146 68.5 115 20 (25% "D 1.5074 1.4996 1.4977 a ^ h . ol) Perillic acid melts at 130-131. CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The Phellandrenes. As early as 1842, Cahours noted a terpene boiling at 173-175 in the essential oil of bitter fennel, which gave a crystalline nitrosite. The correct empirical formula of this nitrosite, C 10 H 16 N 2 3 , was first established by Pesci, 206 who showed that "phellandrene" from the essential oil of water fennel, Phellandrium acquaticum, also yielded a nitrosite of the same empirical formula. Both phellandrenes were dextro-rotatory and yielded laevo nitrosites of nearly identical physi- cal properties; the phellandrenes from both sources were unstable and even by repeated distillation were converted to limonene. Wallach took up the study of the phellandrenes in 1902 and a little later 207 showed that d.phellandrene from elemi oil and d.phellandrene from bitter fennel oil were identical in every respect and that Z.phellandrene from Eucalyptus amygdalina oil was the corresponding laevo form; that the dphellandrene from water fennel oil is a different hydro- carbon, and therefore designated the two hydrocarbons as a-phellan- drene and (3-phellandrene, respectively. That a-phellandrene belonged to the paramenthane series of hydro- carbons was early recognized by reason of the easy conversion of the dibromide to cymene. 208 The constitution 209 of a-phellandrene is indicated by its conversion to carvotanacetone (A 6 -p-menthene-2-one) . When a-phellandrene nitrite is heated with alcoholic caustic potash, nitro-a-phellandrene is formed which, on careful reduction by zinc and acetic acid, yields carvotanacetone, the constitution of which had previously been estab- lished, a-phellandrene a-phellandrene nitrite **Gazz. chim. Ital. 16, 225 (1886). 207 Ann. 336, 9 (1904). 208 Ann. 287, 383. 209 Wallach, loc. cit. nitro-cn- phellandrene carvo- tanacetone THE PARAMENTHANE SERIES 381 The constitution of ct-phellandrene shown above is confirmed by its synthesis from 4-isopropyl-A 2 -cyclohexenone. 210 CH 3 CH 3 & A-OH i H 2 C CH H,C MgCH.l H 2 C CH H 2 C CH N/L \/ V r*Tj OTI r^tr v^xl \jJUL v^H C 3 H T C 3 H 7 C 3 H In carrying out the above synthesis the intermediate alcohol is not liberated as such but the magnesium-methyl halide addition product decomposes in the reaction mixture to give the hydrocarbon. Hydro- carbon formation is often observed under these conditions. a-Phellandrene nitrite is known in two forms. It is best pre- pared 211 in ligroin solution, shaking the ligroin-phellandrene mixture with concentrated aqueous sodium nitrite acidified by acetic acid. The two nitrites may be separated by crystallization from dilute acetone. The sparingly soluble a-nitrite, melts at 112-113, [a] D +136 to 143, [a] D 138. The p-nitrite is more soluble and melts at 105 [a] + 45.8 and [a] 40.8. On reduction these nitrites give a diamine, the hydrochloride of which decomposes on distillation yielding cymene. (3-Phellandrene also yields two known nitrites, the so-called a-nitrite melting at 102 and the p-nitrite melting at 97-98. When p-phellandrene nitrite is converted to nitro-p-phellandrene and when this is reduced by sodium and alcohol, dihydrocumin aldehyde is pro- duced, 212 indicating that its structure is either 310 Wallach, Ann. S59, 285. 211 Wallach, Ann. 313, 345 ; S36, 13. 212 Wallach, Ann. 340, 9. 382 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS H 2 C r /\ CH or H H H 2 C CH YH C 3 H 7 C 3 H 7 However, p-phellandrene is optically active which eliminates II, which has no asymmetric carbon atom. The decomposition of nitro deriva- tives with the formation of aldehydes has been observed by Kono- walow 213 and others. Careful oxidation of p-phellandrene by per- manganate yields first a glycol which on decomposing with acids, yields tetrahydrocuminaldehyde, H.OH - CHOH _ CHO p-phellandrene glycol tetrahydro- cuminaldehyde Isopropylcyclohexenone is formed by air oxidation of p-phellandrene in the presence of moisture, 214 which may be expressed, according to Engler's theory of air oxidation, as follows, " Ohem. Zentr. 1889, I, 1238. w Wallach, Ann. Stf, 29 (1905). THE PARAMENTHANE SERIES 383 The following physical properties of the phellandrenes have been noted: l.a-phellandrene, boiling-point 65 (12mm.), d^ 0.8465, n D 1.488. d.a-phellandrene, boiling-point 61 (llmm.), d 0.844, n.^ 1.4732. Synthetic a-phellandrene, boiling-point 175-176 d 0.841, D 1.4760, M D 45.61. 20 dfi-phellandrene, boiling-point 57 (llmm.), d 2QO 0.8520, n-^- 1.4788, [a] D + 18.54 (?). The essential oil of Bupleurum fruticosum yields d(3-phellandrene showing an optical rotation of [a] _^ 65.2, from which fact, together with evidence obtained by a study of the oxidation products and the nitrosochlorides, the discoverers 215 conclude that the d.(3-phellandrene of Pesci and Wallach, which showed a much lower rotatory value, is a mixture of the two optical antipodes. A terpene fraction boiling at 169-171, isolated 216 from the volatile oil of Rubieva multifida of California, and consisting "largely" of (3-phellandrene, showed [a] D +46.4. Hydrogen chloride passed into an alcoholic solution of (3-phellan- drene gives a-terpinene dihydrochloride. 217 218 Fransesconi & Sernagiotto, Gazz. chim. Ital. 46 (1), 119 (1916). "Nelson, J. Am. Chem. Soc. 42, 1286 (1920). 217 Fransesconi & Sernagiotto, Gazz. chim. Ital. 44 (2), 456 (1914). Chapter X. Cyclic Non-benzenoid Hydrocarbons. Ortho- and Meta-Menthanes and Their Derivatives. Sylvestrene: The most important derivative of this series is syl- vestrene, a terpene discovered by Atterberg in Swedish oil of turpen- tine, from Pinus sylvestris. Its physical properties are nearly identi- cal with those of limonene, boiling at 175-176. It is one of the most stable of the terpenes and is not isomerized to other terpene hydrocarbons either by the action of heat or dilute acids. Its rela- tion to meta-cymene was shown by Baeyer by first reacting upon it by hydrogen bromide forming the dihydrobromide (melting-point 72), introducing a third bromine atom and treating the tribromide with zinc dust and alcoholic hydrochloric acid when meta-cymene was produced. Under these same condition limonene gives para-cymene. The inactive form of this terpene has been called carvestrene and bears the same relation to sylvestrene that dipentene bears to limo- nene. Baeyer 1 made i-sylvestrene from carvone by reducing this ketone by sodium and alcohol to dihydrocarveol, oxidizing this alcohol to dihydrocarvone and adding hydrogen bromide to the latter ketone; when the hydrobromide of dihydrocarvone is treated with cold alco- holic caustic potash, carone is formed, which substance has been shown to have a cyclopropane ring. The oxime of carone is reduced in the usual manner to the corresponding amine, and warming witfc dilute acids ruptures the three-carbon ring. When the hydrochloride of this amine is heated, ammonium chloride is split off and i-sylves- trene (carvestrene) is produced. l Ber. 27, 3485. 384 ORTHO AND METAMENTHANES 385 carvone dihydrocarveol dihydrocarvone carone vestrylamine i-sylvestrene The nature of the reaction taking place when the hydrobromide of dihydrocarvone is treated with alkali to form carone, and the con- stitution of carone, was first suggested by Wagner, whose views were accepted by Baeyer. In conjunction with Ipatiev, Baeyer investi- gated the oxidation of carone by permanganate 2 and showed the formation of two isomeric dibasic acids, C 5 H 8 (C0 2 H) 2 , one of which readily forms an anhydride (when boiled with acetyl chloride) but the other does not form an anhydride under these conditions. Their research led Baeyer and Ipatiev to the conclusion that these two caronic acids were cis and trans modifications of the following struc- ture, C(CH 3 ) 2 C(CH 3 ) 2 C C C0 2 H H C C H H0 2 C trans-caronic acid *Ber. 29, 2796 (1896). C0 2 H C0 2 H cis-caronic acid 386 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The correctness of the constitutions shown above was proven by W. H. Perkin, Jr., and J. F. Thorpe, 3 who synthesized the caronic acids from bromodimethylglutaric ester, C(CH 3 ) 2 C 2 H 5 2 C - CHBr H.CH.C0 2 C 2 H { R0 2 C - CH C(CH 3 ) 2 - CH.C0 2 R Trans-caronic acid is converted to the anhydride of cis-caronic acid by heating with acetic anhydride at 220. Inactive sylvestrene has been synthesized by Perkin 4 by means of the Grignard reaction. Starting with meta-hydroxybenzoic acid, which was reduced to cyclohexanol-3-carboxylic acid and this oxi- dized to the corresponding ketone, the reactions may be represented as follows, being parallel to the reactions employed by Perkin for the synthesis of limonene. It is evident that Baeyer's i-sylvestrene can be only the A 1 hydro- carbon, but Perkin's synthetic hydroc'arbon may, from the method of its preparation, be either A 1 or A 6 hydrocarbon, although Perkin's results indicate that his synthetic sylvestrene consists at least mainly of the A 1 product. *J. Chem. Boc. 75, 49 (1899). OR THO AND METAMENTHANES 387 Sylvestrene cannot be isolated from Swedish oil of turpentine by fractional distillation on account of the presence of other terpenes of practically the same boiling point. It has usually been prepared by making the crystalline dihydrochloride from the fraction boiling at 173-178 and decomposing this with an alkali or an organic base. Wallach observed that the terpene so prepared was not pure but by fractional distillation of the product obtained by decomposition of the dihydrochloride obtained a sylvestrene of the following physical prop- erties. Blg.-pt. 175-176; d 0.848; n 1.4757; [a] +66.32. 20 D D It is well known that the decomposition of dipentene dihydro- chloride or ordinary terpin, and also terpinene dihydrochloride or terpinene-terpin (1.4 terpin) yields mixtures of terpenes and it would therefore appear probable that the decomposition of sylvestrene dihy- drochloride would also yield a mixture of hydrocarbons. The first definite demonstration that sylvestrene dihydrochloride is 1.8-di- chloro-m.-menthane was the conversion of dl.-m.-h l -menthenol(8) and d/.-ra.-A 6 -menthenol(8) into this dihydrochloride, 5 A-m-menThenol(8) CH 3 dih_ydrochloride The decomposition of this dihydrochloride could possibly yield the following six isomeric meta-menthadienes. * Perkin & Tattersall, J. Chem. Soc. 91, 481 (1907) ; Perkin & Fisher, ibid., 93, 1888 388 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CH a CH 3 CH 2 i HC CH. (1) V H 2 CH 8 C HC CH S CH 3 GIL \ CH a CH / \ CH 3 The hydrocarbons represented by (4), (5), and (6) have no asym- metric carbon atom and since sylvestrene is optically active its struc- ture cannot be (4), (5) or (6). Also, sylvestrene does not show the chemical behavior of a substance having a semicyclic >C = CH 2 group, which renders the structure (3) very improbable. Haworth, Perkin and Wallach 6 have shown that repeated fractionation of the crude sylvestrene, made by heating the dihydrochloride with diethylaniline, yields a sylvestrene boiling at 175 (751mm.) and [a]-~ + 83.18 at 18. A higher boiling fraction was also isolated, boiling at 182-184 and [a] + 45.42. This terpene resinifies rapidly on exposure to air, or in contact with sodium, and the authors conclude that it contains a considerable proportion of inactive syl- veterpinolene together with some isomeride of similar boiling-point but optically active. The purest sylvestrene thus obtained, boiling at 175, is regarded as a mixture of the A 1 and A 6 isomerides, (1) and (2) above. All efforts to obtain a pure sylvestrene of definite con- J. Chem. Soc. 103, 1230 (1913). ORTHO AND METAMENTHANES 389 stitution by the dehydration of sylveterpin, under different conditions, were without success owing to the marked tendency of the sylveterpin to form meta-cineol, The complexity of the problem is indicated in the foregoing discus- sion but, nevertheless, Haworth and Perkin 7 were able to synthesize both optically active forms of sylvestrene and their research, cul- minating with the synthesis of d. and . sylvestrene, is one of the most interesting examples of refined experimental method and application of the theories of organic chemistry. The removal of hydrogen bromide from 1-bromo-l-methylcyclo- hexane-3-carboxylic acid yields a mixture of the A 1 and A 6 unsaturated acids. By fractional crystallization of the brucine salt an optically active acid [a]^ + 108 was isolated, and from the mother liquors, by em- ploying Z.menthylamine, another acid [a]^ 49.7 was obtained. In order to show which of these acids was the A 1 and which the A 6 acid, the latter was synthesized from l-methylcyclohexane-6-one-3- carboxylic acid, /. Chem. Soc. MS, 2229 (1913). 390 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS -CO Z H and this, on resolution by brucine, also gave an acid [a] +108 and the laevo form [a]^ 98.6. This dextro-rotatory acid, from both sources, was converted into d-A 6 -m-menthenol(8), which in turn was changed to d.-sylvestrene dihydrochloride, []-pv + 22.5, which on decomposition by diethylaniline gave d-sylvestrene, [a] The Isevo A 6 acid, [a] D 98.6, and the laevo A 1 acid, [a] D 49.7 both gave l-sylvestrene, by similar reactions, the rotation being 66.5 in one case and 68.2 in the other. Sylveterpin and Sylveterpineols: When sylvestrene dihydrochlo- ride is shaken with dilute aqueous caustic potash the corresponding terpin is formed. Like ordinary terpin, sylveterpin exists in two modifications of the cis and trans type, the cis form melting at 137-138, being less soluble, was discovered first, 8 and the more solu- ble trans form, melting at 70-75, was recently discovered 9 in the mother liquors after separating the first or cis form. The cis and trans forms of sylveterpin are the d. constituents of the inactive or cis and trans carveterpins. TVans-carveterpin, melting at 126-127, was discovered by Baeyer during his researches on i-sylvestrene (or "carvestrene"). 10 Sylveterpineol, the chief product of the action of dilute alkali on sylvestrene dihydrochloride, has been shown, 11 by study of its oxida- tion products to be a mixture of A 6 -m-menthenol (8) and A^m-men- thenol(8). The mixture distills at 214. The menthenols obtained by synthesis, employing the Grignard reaction as described in the foregoing pages, are usually obtained quite pure. All of the six theoretically possible raeta-menthenols, having the hydroxyl group in position (8) are known. 12 When these meta-menthenols are decom- 8 Wallach, Ann. 357, 73 (1907). Haworth, Perkin & Wallach, J. Chem. Soc. 103, 1234 (1913). Ber. 27, 3490 (1894). 11 Haworth, Perkin & Wallach, loc. cit. "Perkin, J. Ohem. Soc. 07, 2129 (1910). ORTHO AND METAMENTHANES 391 posed a mixture of hydrocarbons results except in the case of A 2 or A 3 -m-menthenol(8) ; which can decompose with loss of water only in one way. CH 3 CH H 2 C CH 2 CH, H, C C OH CH 3 & 3 -m-menthenol(8) CH 3 C Y V ui 2 H t 9) -m-menthadiene This hydrocarbon is of interest as showing the effect of the conjugation of the two double bonds upon the physical properties, as compared with the isomeric m6a-menthadienes. 13 Its physical properties closely resemble the similarly constituted A 3 8 < 9 >-p-menthadiene. I. A 2 8 < 9 > m-menthadiene 14 A 3:8(9) m-menthadiene A 3 8 < 9 > p-menthadiene II. III. IV. A 1 8(9) p-menthadiene (limonene). /. //. Boiling-point ....... 182 181-182 20 d- ............... 0-8624 0.8609 184-185 0.8580 1.4924 46.02 IV. 175-176 0.8460 1.4746 45.23 nD 1.5030 1.4975 M 46.6 46.3 M. calc. for C 10 H 16 /= 2 45.24. Dihydrosylveterpineol [m-menthanol(8)] possesses two asym- metric carbon atoms and accordingly exists in two slightly different isomeric forms, the activity of one being due to the carbon atom (1), and in the isomer carbon atom (3) is active. The latter substance is obtained by the catalytic hydrogenation of sylveterpineol and the former is prepared by synthesis from l-methyl-3-acetylcyclohexane. 15 13 Luff & Perkin, J. Chem, Soc. 97, 2154 (1910). "Haworth, Perkin & Wallach, J. Chem. Soc. 99, 120 (1911). 13 Haworth, Perkin & Wallach, J. Chem. Soc. 103, 1228 (1913). Wallach, Ann. S81, 392 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS [a] form, 10.35 ph. ur ethane, M.-P. IT d(l) form, [a] D +1.96 ph. ur ethane, M.-P. 83 ( Ortho-Menthane Derivatives: Menthenols and menthadienes of the ortho series have not been found in nature but we have a fairly complete knowledge of them due largely to the systematic researches of Perkin, Jr., and his assistants. The methods of synthesis employed by Perkin to obtain the substances of this series are quite closely analogous to those already described in connection with the para and raeta-menthane derivatives. Of the six possible or/io-menthenols, in which the hydroxyl group occupies postion (8), five are known. Their boiling-points under 30mm. pressure are given for the known ortho- menthenols, Of A 5 110 ORTHO AND METAMENTHANES Like the m-menthenols, these of the ortho series have odors closely resembling a mixture of terpineol and menthol. No attempt has been made to resolve the synthetic inactive o-menthenols into their active d and I constituents. A 1 -o-menthenol (8) was synthesized from ortho toluic acid, which will serve to illustrate a typical synthesis of this series. Reduction by sodium and amyl alcohol gave 1-methyl-cyclo- hexane-2-carboxylic acid which was then brominated and then decom- posed to the unsaturated acid which was proven to be 1 -methyl- A 1 - cyclohexene-2-carboxylic acid by oxidation with permanganate to 3-acetobutyric acid. CH 1 8/9) ^^"^ A-0-menthenol(8) A -0-merithadiene /CO.H . 3- aceto butyric acid As in the case of A 2 -m-mentheriol (8) , this o-menthenol can decom- pose with the formation of a double bond in only one direction and accordingly the resulting A 18(9) -o-menthadiene is quite pure. It ex- hibits the usual characteristics of a hydrocarbon containing conjugated double bonds, combines with only one molecule of a halogen or halo- gen acid, exhibits exaltation of the molecular refraction, has a boiling- point higher than its isomers which do not have their double bonds in conjugated position, resinifies rapidly in contact with air or on warm- ing with metallic sodium, etc. The same o-menthenol and o-men- thadiene was synthesized in quite a different manner by condensing diacetylpentane (by means of sulfuric acid) and treating the resulting unsaturated ketone with magnesium-methyl-iodide in the usual manner. 394 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS H 9 C C COCH, H 2 C C C(CH 3 ) For the preparation of the A 5 and A 6 -o-menthenols and the o-men- thadienes resulting from their decomposition Perkin was compelled to make use of an ingenious method of separating the 1-methyl-A 5 and l-methyl-A 6 -cyclohexenecarboxylic acids. Haworth and Per- kin 16 had observed that of the following two acids the A 4 acid esteri- fies much more rapidly and the ester is hydrolyzed or saponified much more rapidly than the A 5 acid. CH 3 CH HC CH C0 2 H HC CH 2 A* acid CH 8 C HC CH C0 2 H C H CH, A 5 acid It was found that the methylcyclohexenecarboxylic acids showed a parallel behavior, the 6-acid esterifying much less rapidly than the 5-acid. 18 J. Chem. 8oc. 93, 577 (1908). ORTHO AND METAMENTHANES CH 8 CH 3 CH C HC CHC0 2 H. HC CHC0 2 H. A 6 H 2 C CH 2 C H 2 A 5 -esterifies much more rapidly than A 6 . Perkin was able to effect a fractional separation of these two acids by making use of this fact, and then synthesized the corresponding o-menthenols in the usual manner. O-Menthane-5-One: The first o-menthanone to be described was prepared by reduction of l-methyl-2-isopropyl-A 6 -cyclohexene-5- one. 17 This o-menthone boils at 204 and yields an oxime melting at 75. "Kotz and Anger, Ber. 44, 466 (1911). Chapter XL Cyclic Non-benzenoid Hydrocarbons. Bicyclic and Tricyclic Non-benzenoid Hydrocarbons. Camphene, bornylene and the pinenes are bicyclic hydrocarbons which might be considered as derivatives of cyclohexane but on account of their importance and the volume of their literature these hydrocarbons are considered in separate chapters. The three simplest bridged cyclohexane hydrocarbons are not known. \ C H norcamphane H norpinane These hypothetical hydrocarbons have the cyclic structures of cam- phene, pinene and carene respectively. A ketone having the structure of norpinane has recently been made by heating the calcium salt of cyclohexane-1.3-dicarboxylic acid and it would probably not prove difficult to prepare the hydrocarbon from the ketone. Norpinane and particularly norcarane would probably prove to be unstable, lacking the gem. dimethyl group. 1 When indene is reduced by sodium and alcohol, two atoms of hydrogen are added, forming hydrindene, a large number of deriva- tives of which are known. Willstatter and King noted that the double *Cf. Ingold, J. Ohem. 8oc. 119, 952 (1921). 396 CYCLIC NON-BENZENOID HYDROCARBONS 397 bond in styrene was reduced by hydrogen and platinum very much more rapidly than the benzene ring and by interrupting the hydro- genation good yields of ethylbenzene could be obtained. Similarly, indene may be hydrogenated in contact with nickel at 200 to hydrin- dene, 2 boiling-point 177. At 300, in contact with nickel and hydro- gen, hydrindene is not further hydrogenated but is partly decomposed and partly converted to indene and hydrogen. At 250-260, in the presence of nickel oxide and hydrogen under 110 atmospheres pres- sure, indene and hydrindene are completely reduced to octohydroin- dene, 3 a stable oil, boiling-point 165-167 (757mm.), d ono 0.8334, 20 UD 1.46287. Santene, C 9 H 14 . This hydrocarbon, discovered in oil of sandal- wood by Miiller 4 and in Siberian pine-needle oil by Aschan 5 is note- worthy as being one of the few hydrocarbons, occurring in essential oils, having other than ten or fifteen carbon atoms. Santene is char- acterized by the formation of a nitrosochloride melting at 109-110, a nitrosite melting at 125 and a hydrochloride melting at 80-81. The alcohol, santenol, formed by treating with glacial acetic and su!- furic acids (Bertram and Walbaum's method) melts at 97-98 and distills at 195-196 (phenylurethane melting at 61-62). The acetate has an odor resembling bornyl acetate and distills at 215- 219. The physical properties of santene noted by different observers are as follows. Boiling-point 31 -33.(9mm.) 140. d 0.863 0.8698(15) 20 n 1.46658 1.4696 D The constitution of santene has been shown by Semmler and Bar- telt 8 by means of oxidation by ozone and by permanganate in dilute acetone to be as represented in the following, 2 Padoa & Fabris, J. Chem. Soc. A6. 1908, I, 255. *Ipatiev, J. Ru88. Phys.-Chem. Soc. 45, 994 (1913). *Arch. Pharm. 238, 366 (1900). 6 Ber. 40, 4918 (1907). 6 Santene from sandal-wood oil, Semmler, Ber. 40, 4591 (1907). 7 Santene from Siberian pine-needle oil, Ashcan, Ber. 40, 4918 (1907). Ber. 41, 385, 866 (1908). 398 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS glycol, M.-P. 193' the constitution being clearly indicated by the formation of cyclo- pentane rarw-dicarboxylic acid melting-point 86, which acid was previously known. The formation of santene from camphenilone has recently been accomplished by Komppa and Hintikka 9 by decompos- ing the corresponding alcohol, camphenilol, by sodium acid sulfate and also by heating camphenilyl chloride with aniline. A mixture of hydrocarbons is obtained but santene is the principal product. A little confusion is cleared up by Komppa and Hintikka by showing that santenol is identical with isocamphenilol and Semmler's n-nor- borneol, and that santenone is identical with isocamphenilone and Semmler's jt-norcamphor. In the conversion of camphenilone or camphenilol a rearrangement occurs, such as is frequently observed among the terpenes and cyclohexane derivatives. camphenilone camphenilol santene Sabinene and Thujene may be considered as derivatives of para- menthane but they are both bicyclic and the bridged ring, common to both hydrocarbons, is a three carbon ring. Thujene (Semmler's tan- 9 Butt. 8oc. cMm. 21, 13 (1917). CYCLIC NON-BENZENOW HYDROCARBONS 399 acetene) has not been found in any essential oil but sabinene occurs in the essential oil of savin and as a subordinate constituent in a number of other essential oils. Sabinene purified by fractional dis- tillation, carried out by Schimmel and Company, showed a boiling- point of 163-164 and an optical rotation of [a] + 63. Although the active hydrocarbon does not appear to have been obtained in a high degree of purity, it can be differentiated from other hydrocarbons of approximately this boiling-point, by its low specific gravity 0.8480 (15). The molecular refraction owes its exaltation over the calcu- lated value C^H^/^ 1 to the presence of the three carbon ring. M (observed) 44.88, calculated, 43.53. It is readily converted to 1.4- terpin, terpinenol (4) and terpinene by the action of dilute sulfuric acid. On oxidation by alkaline permanganate sabinene behaves very much like p-pinene and other substances having a semicyclic methene group; it yields first sabineneglycol (melting-point 54), then sabi- nenic acid, the sparingly soluble nature of the sodium salt making its isolation easy. 10 Sabinenic acid melts at 57 and on further oxida- tion by lead peroxide and sulfuric acid yields sabina ketone. 11 The three carbon ring in sabina ketone is readily broken by hydrogen chloride in methyl alcohol and when the product is heated with aniline two isopropylcyclohexenones are produced. These ketones have been useful as serving for the synthesis of ct-terpinene and a and (3-phellandrene. CH sabinene glycol M.-P. 54 sabinenic acid M.-P. 57 sabina ketone "Wallach, Ann. 359, 266 (1908). 11 Sabina ketone boils at 218 -219 and yields a semicarbazone melting at 141-142V 400 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Thujene has been made indirectly from thujone, a ketone occuring in the oils of thuja, wormwood, tansy and sage. (The ketone has also been called tanacetone.) The ketone can be isolated by its bisul- fite compound, using ammonium bisulfite and adding a little alcohol to increase the solubility of the ketone, and allowing to stand. The ketone may be liberated from the crystalline bisulfite compound by adding alkali and distilling with steam. There are two physically isomeric thujones, designated as a and (3, and when they occur to- gether they can be separated by fractional crystallization of the semicarbazones and regeneration of the ketones from the purified semicarbazones. The a-thujone, which is the chief ketone in thuja oil, boils at 200-201, specific gravity 0.9125, [a] 10 23' and 1.4510. It appears to yield two dextro-rotatory semicarbazones melting at 110 and 186-188. Heating with alcoholic caustic alkali or alcoholic sulfuric acid converts a-thujone partially into p-thujone. When p-thujone is liberated from its semicarbazone (melting-point 170-172 or 174-176) it is dextro-rotatory [a] + 76.16. Its oxime melts at 54-55. The conversion of a to P-thujone by alco- holic alkali is reversible. Both ketones yield the same bisulfite compound. When the three-carbon ring of thujone is broken by heating with 40 per cent sulfuric acid an isomeric ketone, isothujone (boiling-point 231-232, d 0.9285) is formed, which change is represented by Wal- lach 12 and by Semmler 13 as follows, "Ann. 323, 371 (1902). "Ber. S3, 275, 2454 (1900). CYCLIC NON-BENZENOID HYDROCARBONS CH S CM, =0 401 isothujone Hydrogen chloride breaks the three-carbon ring in a different manner, a-thujene giving terpinene dihydrochloride. 14 Isothujone yields two physically isomeric thuj amenthols according to whether the reduction is carried out by sodium and alcohol (a-thujamenthol, boiling-point 212-214, d 0.8990) or by hydrogen and palladium which yields p-thujamenthone and then by farther reduction by alcohol and sodium p-thujamenthone 15 yields the p-alcohol, which boils about 2 higher than the a-form. Thuj one may be reduced to the corresponding alco- hol, thujyl alcohol (boiling-point 210-212, d ono 0.9265), which U alcohol is also formed by the action of nitrous acid on thujyl amine (the yields of alcohol by this reaction in most cases are very poor). Thujylamine is obtained in the usual manner, by reduction of thu- jone oxime. Oxidation of thuj one yields first a keto acid, melting- point 75-76, and then by further oxidation a dicarboxylic acid melt- ing at 141-142, both still retaining the three carbon ring but the .cyclopropane ring is much more stable in the dicarboxylic acid. COW C0 2 H a-thujaketonic acid M.-P. 75-76 labil a-dicarboxylic acid M.-P. 141-142 stable " Wallach, Ann. 360, 97. 402 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The constitution of thujene has not yet been clearly shown but it is believed to be as follows, The physical properties of the hydrocarbon, prepared by different methods, indicate that "thujene" is probably a mixture of hydrocar- bons, one of which probably has the constitution shown above. The name was formerly applied to the hydrocarbon made by the dry dis- tillation of the hydrochloride of thujylamine or isothujylamine. Tschugaeff prepared thujene by heating thujyl xanthogenate, and also by heating and decomposing trimethylthujyl-ammonium hydroxide, the latter method giving a hydrocarbon of considerably higher optical rotation than the former. The highest rotation observed is that noted by Kondakow and Skworzow, 16 i.e., + 109. The following physical properties have been noted, Observer Boiling-Point Density 16 n D Semmler " 60- 63 (14mm.) 0.8508 1.4760 Wallach 170-172 (760mm.) 0.8360 1.47145 Tschugaeff " 151 -152 (670mm.) 0.8275 1 .45042 Thujane was made by Tschugaeff and Formin 19 by catalytic hydrogenation, in the presence of platinum, of the thujene made by decomposing thujylmethyl xanthogenate. Thujane is readily oxidized by permanganate. Sabinene also gives the same hydrocarbon by hydrogenation under the same conditions. The following physical properties were noted, boiling-point 157 (758 mm.), d lfio 0.8190, Mol. 19 Chem. Zentr. 1910, II, 467. Ber. 25, 3345 (1892). "Ber. 33, 3118 (1900). 19 Compt. rend. 151, 1058 (1910). CYCLIC NON-BENZENOID HYDROCARBONS 403 Refraction^ 44.54 to 44.80, calculated 43.92, the difference being attributed to the presence of the cyclopropane ring. Thujane pre- pared by Kishner 20 from thujone by his hydrazine method showed the following physical properties, boiling-point 157.5 (741 mm.), d 0.8164, [a] D +53.41, n D 1.4398. Carene: This terpene, recently found in Indian turpentine (from Pinus longifolia, Roxb.) is one of the few hydrocarbons occurring in nature which contains a three-carbon ring. It has frequently been noted that this turpentine contained a terpene which yields sylvestrene hydrochloride and it is usually stated that sylvestrene is present in this oil, although Robinson 21 stated that the terpene was probably an isomer of sylvestrene. Simonsen 22 isolated the hydrocarbon, boiling- point 168-169 (750mm.) by fractional distillation, and had no diffi- culty in preparing d-sylvestrene hydrochloride from this fraction. The liquid hydrochloride mixture gave sylvestrene and dipentene on heat- ing with sodium acetate in acetic acid. Oxidation by permanganate gave a glycol melting at 69-70 which apparently contains no pri- mary alcohol group indicating the absence of the methene group. Oxidation by permanganate under the conditions recommended by Baeyer and Ipatiev gave trans-caronic acid, from which facts Simon- sen concludes that" carene has one of the two following structures, or is perhaps a mixture of the two. 30 d. Carene is slightly dextro-rotatory, [a] + 7.69, D - 0.8586, 30 n_ 1.469 and from the refractive index M 44.23; M calculated for *J. Ruas. Phys.-Chem. Soc. 42, 1198 (1910). J1 Proc. Chem. Soc. 27, 247 (1911). 21 J. Chem. Soc. in, 570 (1920). 404 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS doH^/^ 1 = 43.5 and adding the increment usually observed in cases where a cyclopropane ring is present M calc. becomes 43.5 + 0.69 = 44.19. Naphthalene is readily hydrogenated in the presence of finely divided platinum. When dihydronaphthalene is employed as the raw material, two atoms of hydrogen are very rapidly taken up and if the hydrogenation is then interrupted a good yield of tetrahydro- naphthalene can be obtained but when starting with naphthalene and stopping the operation after four atoms of hydrogen had been taken up, the product was found to be a mixture of unchanged naphthalene and decahydronaphthalene. 23 Tetrahydronaphthalene and the completely hydrogenated decahy- dronaphthalene were widely used in Europe, during the war period, as solvents, particularly as paint and varnish thinners. 24 A mixture of the hydrocarbons is manufactured under various trade names. Their solvent values are not accurately known but they are miscible with petroleum oils and are good solvents for coumarone resin, many natural resins, waxes, fats and oils. Their manufacture appears to be carried out in accordance with the well-known conditions of hydro- genation, employing temperatures within the range 120-150, and pressures within the range 3 to 100 atmospheres. 25 A preliminary purification from sulfur compounds by heating with metallic sodium, or with sodium amide is advised. 26 Tetrahydronaphthalene distills at 205-207, d 1 K0 0.975, flash-point 78. Decahydronaphthalene dis- lo tills at 189-191, d 2QO 0.8827, flash-point 57.3 . 27 Auwers 28 notes the molecular refraction (D line) of tetrahydronaphthalene as 42.91 and that of decahydronaphthalene as 43.85. The action of bromine on tetrahydronaphthalene is of interest as indicating the relative ease of bromine substitution in the two types of rings. No reaction takes place in the dark except in the presence of a catalyst such as iron or iodine when substitution in the benzene ring takes place. At higher temperatures, or in the light, the reduced ring is rapidly attacked but the only product isolated was a p-dibromo- tetrahydronaphthalene 29 (melting-point 70). 28 Willstatter & King, Ber. 46, 527 (1913). ** Frydlender, Rev. prod. chim. 23, 437 (1920). "Brit. Pat. 147,474 (1920). 2 Brit. Pat. 147,488 (1920); 147,580 (1920). "Vollman, Farter Ztg. 24, 1689 (1919). "Ber. 46, 2988 (1913). 29 v. Braun & Kirschbaum, Ber. 5$, 597 (1921). CYCLIC NON-BENZENOID HYDROCARBONS 405 The alcohols a and p-naphthanol were prepared by Ipatiev by his high pressure method. 30 0-Naphthanol C 10 H 17 .OH distills at 242- 244 and melts at 99-100; a-naphthanol, C 10 H 17 OH, distills at 245-250 and melts at 57-59, but Mascarelli 81 states that this alcohol can be separated into two stereo-isomers melting at 75 and 103. Both alcohols resemble cyclohexanol and the aliphatic sec- ondary alcohols in their chemical behavior. Naphthane-2 . 2-diol has been obtained in cis and trans forms; by the action of dilute caustic potash on 2 . 2-dibromonaphthane the cis form melting at 160 is obtained, while silver acetate on the dibromide yields the trans diol, melting at 141 . 32 The ketone |3-naphthanone has been very little studied but evi- dently undergoes the reactions of ether alicyclic ketones. Darzens and Leroux 33 condensed (3-naphthanone with chloroacetic ester in the presence of sodium ethylate to the glycidic ester, the free acid from which loses carbon dioxide on distillation giving p-naphthanoic alde- hyde (boiling-point 95-96 at 3 mm.). H 2 H 2 H 2 H 2 C C CO / \H/ \ / \H/ \ H H 2 C C C = H 2 C C C< -*RC CH.C0 2 R- I CHO H 2 C C CH 2 H 2 C C CH 2 ' \ /H\ / ' \ /H\ C C C C H 2 H 2 H 2 H 2 a, a-Dicyclohexylethane, CH 2 CH 2 This hydro- carbon was made by Sabatier and Murat 34 by the hydrogenation of diphenylethane in the presence of catalytic nickel and hydrogen at 220. Its physical properties are as follows: Boiling-point 256-257, 40 d 2flo 0.8271,n s - 1.511. The Nomenclature of spiro and other bridged ring hydrocarbons is in a very unsatisfactory state and none of the systems thus far proposed are very satisfactory except for certain types or classes of hydrocarbons. Probably the most flexible and least confusing is that 80 Per. 40, 1288 (1907). 11 Chem. Alts. 1912, 83. "Leroux, Compt. rend. 148, 1614 (1909). n Compt. rend. 154, 1812 (1912). "Compt. rend. 154, 1771 (1912), 406 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS recently proposed by Beesley and Thorpe. The scheme advocated by Baeyer 85 rests upon the fact that all dicyclic systems contain three bridged rings which makes it possible to distinguish them by prefixed numerals such as (0.1.2), (1.2.3), (0.1.4) and so on, depending upon whether the "bridge" is formed by the linking of two tertiary carbon atoms (0), or whether it is itself formed by 1,2 or more carbon atoms. When Baeyer's system is extended to tricyclic substances it becomes exceedingly cumbersome and complex. The plans suggested by Borsche 36 and by Bredt and Savelsberg 37 are open to the objection that terms such as methylene are used to denote ring formation and not unsaturation, and that the names of the compounds do not neces- sarily indicate to which of the cyclic systems they belong. Thus pinene by these systems would be named as follows, Borsche. l-Methyl-l-r- 2 ^ 4 >-dimethylmethylene-A 1(6 )-cyclohexene. Bredt and Savelsberg: m-meso-methylene-4 . 4 . 2i3-trimethy Icyclo- A^-hexene. Beesley and Thorpe (see below) : dimethylmethane-II 1 3 -4-methyl- A 4 -cyclohexene. The hydrocarbon of the following structure, CH, CH CH, would be named, according to Bredt and Savelsberg, p-wesometh- ylene-1 . l-dimethylcyclohexane-amphi-2 a . 3 a -methylene. By Beesley and Thorpe's system, the name would be methane-II 1 - 2 -cyclohexane- 1 -*II-dimethylmethane. Beesley and Thorpe's system appears to the writer to be much more easily grasped and easier to apply than the others, and much more definite. It may be briefly outlined as follows: A compound containing associated rings may be of two kinds. A. It may be formed from a simple ring compound having a side chain of carbon atoms from which another ring is produced by a link- ing between another carbon atom of the ring and another of the side chain, thus: *'Ber. S3, 3771 (1900). "Aim. 877, 70 (1910). "J. praM. Chem. (2) 97, 1 (1918). CYCLIC NON-BENZENOID HYDROCARBONS 407 (1) (2) (3) In these cases the side chain and the ring would be given their usual names, the number of linkings joining the two would be indicated by a Roman numeral, and the carbon atoms of the two series participat- ing in the ring complex would be indicated by means of index figures on which the particular residue is placed. Thus, the above hydrocar- bons (1), (2) and (3) would be named as follows, (1) is Ethylmethane-IP 2 cyclopentane. (2) is 2-Methylethane 12 IP- 2 cyclopentane. (3) is Propane * 3 IP 2 cyclopentane. The following is an example of the nomenclature of derivatives accord- ing to this system. GIL CH CH 9 CH 2 C Br >CH CH.Br CH 3 l-bromo-2-methylethane ^_ 1.2 n 1.2 _2-bromo-4-methyl- cyclopentane. CH CH CH I C H.CH 3 2-methylethane 1 - 1 - 2 III 1 - 2 - 4 cyclobutane. B. The associated ring may be considered as formed by linking pairs of carbon atoms in a ring to which another ring is already attached, as for example, the following, 408 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS H 9 C CH 2 H 2 C CH CH \ CH, CH CH<| CH. HC I CH CH \ cyclopropane-*- 2 II 1 - 2 - -cyclopentane The only rules which seem to be necessary are: (1) That one of the linked carbon atoms in the ring should be called 1, and that the corresponding carbon atom in the chain should also be called 1. The numbering would then proceed in the ring clockwise, and in the side chain in the usual manner. (2) That the name of the simplest por- tion of the chain entering into ring formation should be used first, and any attached groups should be named as derivatives of the simplest chain, for example, CH 2 CH CH 2 CH CH 2 CH 2-methyl 1 - 2 III 1 - 2 - 6 cyclohexane For further details and possible extensions of the system to hetero- cyclic compounds, the original paper of Beesley and Thorpe should be consulted. 88 8.3-Dimethyl-[0.1 .3]-Dicyclohexane: This hydrocarbon was synthesized by Zelinski 39 by reducing 1 . 1-dimethylcyclohexane 3.5- dione to the corresponding diol, converting the diol to the correspond- ing dibromide by phosphorus tribromide and finally treating the dibro- mide with zinc dust in aqueous alcoholic solution. The chemical and physical properties of the resulting hydrocarbon, boiling-point 115 20 20 (corr), d 0.7962, n 1.4331, particularly when compared with the isomeric 1 . l-dimethyl-A 8 -cyclohexene led Zelinsky to propose the bicyclic structure shown below. The three carbon ring is broken in two ways under different conditions. (1) By heating with hydri- odic acid to give a hydrocarbon boiling at 115-116 and indifferent to bromine and permanganate, probably 1.1.3-trimethylcyclopentane, and (2) by catalytic hydrogenation in the presence of platinum black, 88 J. Chem. Soc. 117, 591 (1920), 88 Ber. $6, 1466 (1913). CYCLIC NON-BENZEN01D HYDROCARBONS 409 yielding a hydrocarbon distilling at 109.5-110.5 ; which Zelinski claims is l-methyl-2-isobutylcyclopropane. CH CH 2 CH 2 CH 2 /I \ CH 3 \ CH 3 H 2 C C< > CH 3 C< \ / CH 3 \| / CH 3 CH CH, CH CH 2 CH CH 3 \ H 2 C<| CH 3 CH CH 2 CH< CH 3 The stability of this hydrocarbon to heat was not investigated but it is acted upon rather energetically by concentrated sulfuric acid. A similar dicyclohexane derivative was discovered by Kishner, 40 in quite a different manner. When camphophorone is treated with hydrazine a pyrazolone base is first formed, which on heating with caustic potash yields the hydrocarbon 2.6.6.-trimethyl-[0.1 .3.]- dicyclohexane. It has a petroleum-like odor, boils at 140 (752 mm.), 20 d - 0.8223 ; does not decolorize permanganate, dissolves in fuming nitric acid and reacts with hydrogen bromide to give a bromomethyl- isopropylcyclopentane. CH 2 CH 2 CH 3 >CH C< + KOH CH 2 CH 2 CH CH 3 -CH-CV I CH,- -* /\ CH 3 X N NH CH 3 CH CH C < CH 3 HBr CH 2 CH 2 CH 2 CH 3 CH C H\ CBr.(CH 3 ) 2 Caryophellene: The hydrocarbon oil described in the older litera- ture under this name has been shown by Deussen and his students to be a mixture of at least two and probably three hydrocarbons. The 40 J. Rues. Phya.-CJiem. Soc. kk> 849 (1912). 410 CHEMISTRY Of THE NON-BENZENOID HYDROCARBONS hydrocarbon mixture, isolated from copaiba balsam, clove oil and other essential oils, which distills at about 258-261, d 0.905 to 0.910 and which yields a crystalline dihydrochloride (by passing dry HC1 into a dry ethereal solution) melting at 69-70, has been called caryophellene. The easiest crystalline derivative to prepare is caryo- phellene alcohol, C 15 H 26 0, readily prepared from the hydrocarbon by Bertram and Walbaum's method. The alcohol melts at 94-96 and yields a phenylurethane, melting at 136-137. The work of Deussen and others on caryophellene clearly shows the difficulties of working with mixtures of hydrocarbons and the almost impossible task of determining the constitution of such sub- stances when present together and when they cannot easily be sepa- rated. It is worth while to examine Deussen's work as indicating to a limited extent the difficulties with which one would be confronted in attempting to ascertain the structure of the hydrocarbons occurring in the higher boiling distillates of petroleums. 41 Wallach obtained a crystalline nitrosochloride melting at 161- 163 from the hydrocarbon fraction boiling at 250-270, derived from oil of cloves. Kremers and Schreiner prepared the nitrosochlo- ride and after reacting with benzylamine, were able to separate the nitrolbenzylamine by fractional crystallization into fractions melting at 167 (named a-caryophellenenitrolbenzylamine) and at 128 (named (3-caryophellenenitrolbenzylamine). Deussen 42 found that by heating the crude nitrosochloride in alcohol, cooling and separating the crystals, the melting-point was raised to 177-179. The behavior of the nitrosochlorides led Deussen to suspect the presence of one or more other hydrocarbons. Repeated fractional distillation 43 resolved the crude caryophellene into three fractions, fractions I and III hav- ing different optical rotation and slightly different boiling-points, but otherwise very much alike, / /// Boiling-point .. 132.-134.(16mm.) 123.-124. (14.5mm.) Ca] D 4.67 25.03 d 2Q 0.90346 0.8990 n D 0.49973 1.49617 MR . 66.45 66.31 MR (for /= 2 ) 66.15 41 It is the writer's belief that the only practical way of throwing any light on the character of such petroleum hydrocarbons is to synthesize hydrocarbons of different types and compare the properties of such synthetic hydrocarbons with close cut petro- leum fractions. 42 Ann. 356, 5 (1907). **Ann. 859, 246 (1908). CYCLIC NON-BENZENOID HYDROCARBONS 411 Fraction I. was believed to be inactive, so-called a-caryophellene con- taminated with a small proportion of the laevo-p-caryophellene. The latter hydrocarbon yields a blue nitrosite from which Deussen con- cludes 4 * (from Baeyer's work on terpinolene) that p-caryophellene contains a double bond of the type shown in the following structure, which he proposed. CH 3 I H 2 CH C HC H 2 C CH CH ft HH CH When an excess of N 2 3 (from the reaction of arsenious acid and nitric acid) is passed into an ethereal solution of caryophellene a blue color first appears, followed by the formation of a voluminous yellow- ish white precipitate and the discharge of the blue color, 45 this be- havior resembling the formation of caoutchouc nitrosite. 46 The volu- minous precipitate from caryophellene crystallizes from ethyl acetate in silky needles melting with decomposition at 159-160, the separa- tion of this substance being regarded by Deussen as a delicate test for P-caryophellene. The formation of this substance is attended by the removal by oxidation of the isopropyl group. c_ NO I C ONO /\ CH 3 CH 3 CH TT ..I.... ONO C/ \ CH 2 NO HC ONO + C 3 residue Deussen advanced this explanation of the change by reason of the fact that the product is soluble in alkali, a property only of primary "Ann. 369, 55 (1909). 48 Deussen, Ann. 388, 138 (1912). "Harries, Ann. 383, 198 (1911). 412 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS and secondary nitro derivatives. Deussen 47 represents the deriva- tives of a and |3-caryophellene diagrammatically as follows, a-caryophellene (Humulene) \ \ / \ \ nitrosochloride nitrosate nitrosite M.-P. 177 M.-P. 161 M.-P. 116 + Na ethylate \ j initrolbenzylamine M.-P. 126-128 nitrosocaryophellene M.-P. 128 p-caryophellene glycol / / \ \ M.-P. 120.5 / nitrosite dihydrochloride y M.-P. 115 M.-P.69-70 nitrosochloride M.-P. 159 / \ \ \ \ \ \ N 3 6 isocaryophellene nitrolbenzylamine\ M.-P. 159.5 \ M.-P. 172-173 \ \ a-form < nitrosochloride M.-P. 122 M.-P. 120 \ p-form M.-P. 146 Although Semmler and Mayer 48 have proposed structural formulae for what he terms (using a curious nomenclature of his own) Terp-caryo- phellene and Lim-caryophellene, these structures can hardly be con- sidered as proven and will not be given space in this brief review. The above outline will indicate the variety of the isomeric derivatives and the difficulty of clearing up the constitution of such mixtures of oils. Humulene is the name given by Chapman to a sesquiterpene fraction isolated from oil of hops, but Deussen considers it to be identical with a-caryophellene. Cadinene is the name given to a hydrocarbon or rather a mixture of "Ann. 369, 41 (1910). "Ber. J,3, 3451 (1910) ; 44, 3651 (1911). CYCLIC NON-BENZENOID HYDROCARBONS 413 hydrocarbons occurring in camphor oil, cedar wood and other essen- tial oils; it is characterized by the formation of a dihydrochloride melting at 117-118 and this dihydrochloride may be prepared from the crude hydrocarbon mixture distilling at 260-280. Pure cadi- nene has never been obtained from natural oils but the sesquiterpene regenerated from the dihydrochloride (which is perhaps not identical with the natural hydrocarbon) is usually regarded as nearly pure "cadinene." The hydrocarbon may be prepared by decomposing the dihydrochloride by the usual methods, heating with alcoholic caustic alkali, with aniline, or with sodium acetate in acetic acid. The physical properties of regenerated cadinene are as follows, I* n" III 51 Boiling-point 274.-275. 271.-273. 271.-272. d^o 0.918 0.9215(15) 0.9183 [al D 98.56 -105. 30' 111. n D 1.50647 1.5073 Dextro-rotatory cadinene has been observed in the essential oil of the Atlas cedar. Cadinene resinifies very rapidly and is very easily polymerized, an indication that the two double bonds are in conjugated positions. The dihydrobromide, melting-point 124-125, and the dihydroiodide, melting at 105-106, are best made in glacial acetic acid solution. By catalytic hydrogenation, in the presence of platinum, tetrahydro- cadinene is produced, boiling at 125-128 (10mm.), d 0.8838, n D 1.48045. The constitution of cadinene is not known. By the distillation of galbanum resin Semmler and Jonas 52 ob- tained a sesquiterpene alcohol, cadinol, which on decomposition by potassium acid sulfate, formic acid or phthalic anhydride yields cadinene. Selinene: A sesquiterpene distilling at 262-269 was discovered in oil of celery seed by Ciamician and Silber 53 and the hydrocarbon was later recognized as a new hydrocarbon by Schimmel & Co., 54 who characterized it by the formation of a dihydrochloride melting "Wallach, Ann. 252, 150 (1889) ; 271, 297 (1892). 60 Schimmel & Co. ; Gildemeister, "Die Aetherischen Oele," Ed. II, Vol. I, 347. 61 J. Russ. Phys.'Chem. Soc. 40, 698 (1908). 82 Ber. 47. 2068 (1914). Cadinol distills at 155-165 (15mm.), d 09720 [a] D + 22. 63 Ber. SO, 496 (1897). "Schimmel & Co. Semi-Ann. Rep. 1910 (1), 32. 414 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS at 72-74. Semmler and Risse 55 prepared the dihydrochloride (by HC1 into the ethereal solution) and regenerated what they regard as selinene, identical with the original hydrocarbon but purer, by decom- posing the dihydrochloride by alcoholic caustic potash. The hydro- carbon thus obtained distills at 128-132 (11 mm.), d 2QO 0.919, H D 1.5092, [a] +61 36'. Reduction by sodium and alcohol yields tetrahydroselinene having the following physical properties, boiling- point 125-126 (10 mm.), d 2QO 0.888, M D + 1 12', n D 1.48375. By shaking the dihydrochloride with milk of lime for 36 hours at 95 an alcohol, selinol, C 15 H 26 0, is formed, which may be reduced by hydrogen (Willstatter's method) to dihydroselinol, C 15 H 28 0, melt- ing-point 86-87. On account of differences observed in the products obtained by treating natural and regenerated selinene with ozone and hydrolyzing the ozonides, Semmler regards natural selinene as a mixture of two hydrocarbons, both of which are believed to yield the same dihydro- chloride. Semmler regards these two hydrocarbons as related to each other in the same way as a and p-pinene, the hydrocarbon predomi- nating in natural selinene having a >C = CH 2 group, the double bond in the regenerated selinene being in the ring. Many will regard the constitutions proposed by Semmler as guesses, perhaps to be proven correct by further work but not clearly shown up to the present time. The two selinenes are bicyclic, contain two double bonds, and are believed by Semmler to be represented by the two following structures, CH 2 CH 3 H. II H 2 | C C CO CH 3 C C CH 2 CH 3 C C CH H 2 C C CH 2 H 9 C C! A : *' A CH 3 CH 2 CH 3 // \H 2 natural selinene regenerated selinene "Ber. J5, 3301 (1912) ; #>, 599 (1913). CYCLIC NON-BENZENOID HYDROCARBONS 415 Eudesmene : A sesquiterpene alcohol discovered by Smith 56 in numerous eucalyptus oils, and named eudesmol, yields the sesquiter- pene eudesmene, C 15 H 24 , when decomposed by heating with 90 per cent formic acid. The alcohol is a bicyclic unsaturated alcohol, melting- point 57 84 and distilling at 156 (10 mm.). It adds two atoms of hydrogen when reduced by Willstatter's method (hydrogen and plati- num black in acetic acid solution) and the resulting dihydro-eudesmol melts at 82 and distills at 155-160 (12.5mm.). When eudesmene or the alcohol is treated with hydrogen chloride in acetic acid a dihy- drochloride, melting at 79-80, is formed. The dihydrobromide melts at 104-105. Eudesmene also combines with four atoms of hydrogen when reduced by the Willstatter method. 58 The physical properties of the two hydrocarbons are as follows, Eudesmene Tetrahydro-eudesmene Boiling-point 122.-124.(7mm.) 122.M22.5 (7.5mm.) d 20 o 0.91964 0.8893 [a] D + 54.6 + 10.2 20 n^- 1 .50874 0.48278 Santalenes: The sesquiterpene fraction of East Indian sandal- wood oil apparently contains two hydrocarbons, which Guerbet 59 has called a and (3-santalene. Their physical properties do not differ widely, a-santalene distilling about 10 lower than (3-santalene. Both hydrocarbons give liquid hydrochlorides but a-santalene forms a nitrosochloride melting at 122 (nitrolpiperidide melting at 108- 109) and P-santalene forms a mixture of two nitrosochlorides which can be separated by fractional crystallization to one melting at 106 and another melting at 152. Probably neither hydrocarbon has ever been isolated in a very pure state. Semmler 60 gives the following physical properties of the two hydrocarbons. Boiling-Point d^ [ a ] D H D ,-santalene jglSftffi ^ " 0-santa.ene fig^f^ 0.892 -35' 1.4932 M J. d Proc. Roy. Soc. N. 8. W. S3, 86 (1899). "Semmler & Tobias, Ber. 46. 2026 (1913). The melting-point previously recorded by Semmler, Ber. 1,5, 1390 (1912J, was 78. 68 Semmler & Risse, Ber. 46, 2303 (1913). "Bull. Soc. chim. (3) 3, 217 (1900). 80 Ber. 40, 3321 (1907). 416 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Semmler 61 regards a-santalene as bicyclic and p-santalene as tricyclic. Associated with the santalenes in sandal-wood oil are two alcohols, a and p-santalol, but their relation to the hydrocarbons has not been shown and Guerbet prefers to distinguish the hydrocarbons formed by the decomposition of the alcohols by the names a-iso and (3-isosanta- lene. The santalols are both primary alcohols, yield an aldehyde, by oxidizing with chromic acid, whose semicarbazone melts at 230. Oxidation with permanganate yields chiefly tricycloeksantalic acid, CnHieOg, melting at 71-72. According to Guerbet a-santalol dis- tills at 300-301 and p-santalol at 309-310. The former is nearly inactive, [a]j) 1.2 and p-santalol has the rotation [a]-> 56. When a-santalol is reduced by hydrogen in the presence of platinum the hydroxyl group is replaced by hydrogen; the product is tetra- hydrosantalene, 62 a bicyclic hydrocarbon, C 15 H 28 , distilling at 115- 116 (9 mm.). p-Santalol behaves similarly, giving mainly tetrahy- drosantalene. By heating I. a-phellandrene and isoprene together in a sealed tube, Semmler obtained a hydrocarbon C 15 H 24 boiling-point 129-132 (at 15 mm.), d20o 0.8976, n-p 1.4949, which he regarded as p-santalene. Limonene and isoprene under the same conditions do not react. In a series of such experiments Semmler showed that generally conden- sation of isoprene with the terpenes can be effected at about 275 but at 330 and higher, the sesquiterpenes are decomposed. 63 Cedrene: This sesquiterpene, occurring in cedar-wood oil associ- ated with the closely related alcohol, cedrenol, is of unknown consti- tution although considerable effort has been spent in research on this hydrocarbon. It forms a dihydrocedrene when catalytically reduced in the presence of platinum. Cedrene distills at 262-263, or 124 to 126 at 12 mm., d 0.9354, [a] D 55, n 1.50233. Oxidation by permanganate (in acetone solution) yields a glycol melting at 160, also a diketone or ketoaldehyde of the empirical formula C 15 H 24 2 and a keto acid of unknown constitution, C 15 H 24 3 (oxime melting at 60 ) . By oxidation of cedrene by chromic acid in acetic acid solution a ketone, cedrone, is produced, this ketone having a strong odor of cedar wood, distills at 147-150.5, d 12 5 o 1.0110. 64 91 Ber. 1120 (1907). 8 Semmler & Risse, Ber. tf, 2303 (1913). ' . "Ber. 47 f 2252 (1914). * Semmler & Hoffman, Ber. 46, 768 (1913). CYCLIC NON-BENZENOID HYDROCARBONS 417 Cedar-wood oil appears to contain two sesquiterpene alcohols related to cedrene. 65 Dihydrocedrene, obtained from natural cedrene by catalytic hydro- genation, distills at 122-123 (10 mm.), d 2QO 0.9204, H D 1.4929. No crystalline hydrochlorides or hydrobromides of cedrene are known. Tricyclic non-benzenoid hydrocarbons have been made by the catalytic hydrogenation of tricyclic benzenoid hydrocarbons such as anthracene and phenanthrene. By the hydrogenation of phenan- threne at the remarkably high temperatures of 360, under high pres- sure, Ipatiev 66 obtained the completely reduced hydrocarbon C 14 H 24 , which he calls perhydroanthracene. It is an oil distilling at 270- 276 and does not crystallize at 15. It is inert to permanganate solu- tion and bromine in the cold, and also practically unacted upon by sulfuric-nitric acid nitrating mixture. Anthracene was reduced by Godchot 67 over nickel at 260 to tetrahydroanthracene, the constitution of which is unknown. It crys- tallizes from alcohol in plates melting at 89 and distilling at 309. At a little higher temperature, 200-205 octohydroanthracene, melt- ing-point 71 and distilling at 292-295, is formed, and at 260-270 and under about 125 atmospheres pressure Ipatiev 68 succeeded in reducing it to decahydroanthracene, melting-point 73-74, and finally to the completely reduced hydrocarbon, perhydroanthracene, an oily liquid. Copcene: This name has been given by Semmler and Stenzel 69 to a sesquiterpene occurring, together with caryophellene, in African copaiba balsam. The hydrocarbon was separated by fractional dis- tillation, its constants as thus isolated, being as follows, boiling-point 119-120 (10mm.), d 1KO 0.9077, [a]^ 13.35, n^ 1.48943. It lo J D gives a hydrochloride identical with that formed by cadinene. The new hydrocarbon is apparently tricyclic, combining with two atoms of hydrogen, by catalytic hydrogenation to' give dihydrocopaene, C 15 H 26 (boiling-point 118-121 at 12 mm., n D 1.47987, d lgo 0.8926). Semmler has proposed a constitution for copaene. Abietic Acid: Ordinary commercial rosin consists chiefly of abietic acid. Its constitution is not definitely known but it has been shown M Semmler & Mayer, Ber. J5, 1384 (1912). "Ber. 41, 999 (1908). "Ann. chim. phys. (8) 12. 468 (1907). 68 Ber. 41, 996 (1908). "Ber. V, 255 (1914). 418 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS to have the three ring carbon structure of phenanthrene. Heating with sulfur forms H 2 S, carbon dioxide and retene, which hydrocarbon is believed to be a methyl isopropyl derivative of phenanthrene. The fossil substance fichtelite, C 18 H 32 , is regarded as completely reduced retene. Schulze 70 showed that rosin oil, obtained by the destructive distillation of rosin, contains hydrogenated retene derivatives and by oxidation 1 . 2 . 4-benzene tricarboxylic acid was obtained. The for- mation of this acid from abietic acid would show that the methyl and isopropyl groups are not attached to the same ring. Although com- bustion analyses of abietic acid, reported by different observers, agree almost equally well with the empirical formula C 19 H 28 2 and C 20 H 30 2 , it may be pointed out that the formula C 19 H 28 2 agrees best with the known evidence that abietic acid has the carbon skeleton of phenanthrene together with a carboxyl group, a methyl and an isopropyl group, or 19 carbon atoms in all. Easterfield and Bagley 71 found that abietic acid was esterified with difficulty and therefore sug- gested that the methyl and isopropyl groups were in ortho positions with respect to the carboxyl group, thus assigning these two groups positions in one of the rings. Bucher 72 has reviewed the literature and, in view of the character of the oxidation products, states that one of the alkyl groups must be in position (8) and the other in posi- tion (2) or (3). Bucher also notes that an alkyl group in position (2) and the carboxyl group in position (1) would satisfy the condition which Easterfield and Bagley believed to be required by the slow rate of esterification. It need hardly be pointed out that much of this rests upon very slender evidence. As regards the difficulty of esteri- fying acids by saturating an alcoholic solution with hydrochloric acid gas, it may be pointed out that instances are known in which esteri- fication with the aid of hydrogen chloride proceeds with difficulty, but with relative ease when the alcohol and acid are heated together. The formula C 19 H 28 2 and the tricyclic structure of reduced retene leaves two double bonds to be accounted for. Grim, 73 who adheres to the C 20 H 30 2 formula, has recently pro- posed a constitution for abietic acid which has only one double bond 7 * Ann. S59, 132 (1908). 71 J. Chem. Soc. 85. 1238 (1904). 72 J. Am. Chem. Soc. 82, 374 (1910). /. Ctwm. Soc. Abs. 1921 (1), 344. 7 * Unpublished work of the writer has shown that when abietic acid, recrystallized from alcohol containing a little hydrochloric acid, is hydrogenated in dilute alcohol by Skita's method, the quantity of hydrogen absorbed is that required by two double bonds (within a very small experimental error). This, however, may nevertheless be in accord with Griin's formula and it may also be pointed out that Griin's formula may also account for the peculiar behavior noted in recrystallizing abietic acid. The CYCLIC NON-BENZENOID HYDROCARBONS 419 and has a bridged ring as in pinene, with which hydrocarbon abietic acid is associated in the natural oleo-resin. The formula which have been suggested are as follows, CH, COM Easterfield & Bagley Bucher (Double bonds not placed) Grun Rosin oil has been manufactured on a large scale and the heavier, neutral fractions used as a lubricant. As noted above such oils con- tain hydrogenated phenanthrene or retene derivatives. The crude oil contains about 30 per cent by volume of organic acids, has a marked greenish-blue fluorescence, and distills over a wide range of tem- perature. Its density varies from about 0.945 to 1.010. The lighter fractions, consisting of hydrocarbons of unknown character, are some- times distilled and collected separately, being known industrially as rosin spirit. The fraction distilling at 343-346 is believed to be a diterpene, C 20 H 30 . Rosin oil resinifies on air oxidation; its solu- bility in 96 per cent alcohol varies rather widely, i.e., 50 to 70 per cent at ordinary temperatures, depending upon the conditions under which the oil has been made. y the Chapter XII. Bicyclic Non-benzenoid Hydrocarbons. Turpentine and the Pinenes. Probably the most outstanding fact with regard to turpentine is its rapidly decreasing production. This is having the result that turpentine is being replaced in many of its applications by light petro- leum fractions, particularly in the case of paints and varnishes where it functions merely as a solvent. There are many industrial uses of turpentine, however, in which it appears to be indispensable, as in the manufacture of artificial camphor, terpineol and dammar varnish. The extent of the forests of the world, capable of producing turpen- tine, is well known and although the production of turpentine has been rapidly diminishing, reasonable sylviculture, as in France, 1 will insure a supply of turpentine easily adequate for chemical and other special purposes. The United States, the principal turpentine producing country, produced 27,073,000 gallons of oil of turpentine in 1914, but only 17,737,000 gallons in 1919 in spite of a considerable increase in the number of producing plants, much higher prices per gallon in 1919 and an increase in the output of "wood turpentine" and similar products of about one million gallons. 2 At the present time the United States produces 75 per cent of the world's turpentine supply. Not many years ago the greater part of the world's turpentine supply was derived from North Carolina alone, but the turpentine forests of that State have practically disappeared, North and South Carolina together now producing less than one per cent of the American output. It is worth while to call attention to these facts, and a knowledge of the physical properties and chemical behavior of the pinenes should be brought to bear upon every important industrial use of turpentine with the object of conserving the supply for uses for which it is indis- 1 The pine tree plantations in Southwestern France cover an area of about 2.5 million acres, of which about 2 million acres are privately owned. 2 Special Report on Turpentine, U. S. Bureau of the Census, Washington. May. 1921; Veitch, U. S. Bur. Chem. Butt. 898 (1920). 420 BICYCLIC NON-BENZENOID HYDROCARBONS 421 pensable and also affording relief by the substitution of cheaper material, so far as possible, in the case of consumers now handicapped by the high price of this solvent. In the United States the only important sources are the long leaf yellow pine, Pinus palustris, and Pinus heterophylla, both of which yield turpentine oils consisting of more than 90 per cent of the two pinenes. The terms gum turpentine, gum spirits or spirits of turpen- tine refer to the volatile oil, distilled unchanged, from the natural oleoresin collected from the trees. Wood turpentine 3 made by distill- ing the wood in closed retorts with steam, or recovered by extracting the wood with a low boiling solvent, can be refined so as to replace turpentine for practically all solvent purposes, but when old stump wood is distilled an entirely different product is obtained, known com- mercially as "long leaf pine oil" or "pine oil," the chief constituent of which is terpineol 4 but other minor constituents which have been identified in it include the pinenes, Uimonene and dipentene, a and y-terpinene, borneol, fenchyl alcohol and traces of cineol and camphor. The greater part of such pine oil distills from 190-220 and is useful for the manufacture of terpin hydrate and terpineol, for the flotation of copper sulfide ores and in certain solvent mixtures and cleansing compositions. Rosin spirit is a term employed for the mixture of hydrocarbons obtained by the destructive distillation of rosin. It contains very little of the pinenes, boils over a wide range of tem- perature and usually contains organic acids of unknown character; it usually gives the Liebermann-Storch color reaction with acetic anhy- dride and sulfuric acid. It will be obvious from their composition that neither pine oil nor rosin spirit can be substituted for turpentine in the manufacture of artificial camphor. Other products resembling turpentine find their way into com- mercial channels. "Recovered turpentine," a name sometimes applied to the mixture of terpenes, chiefly i-limonene, is produced by decom- posing the liquid hydrochlorides obtained as a. by-product in the manufacture of bornyl chloride and artificial camphor. Approxi- i mately 90 per cent of this product boils within the range 170-180, i depending upon the rectification and purification of the product. The [presence of chlorides, as indicated by the Beilstein or other halogen I tests, is indicative of such an origin. The oils given off during the Cf. Frankforter, J. Am. Chem. Soc. 28, 1467 (1906) ; Hawley & Palmer, U. S. .Forest Service Bull. 109 (1912) ; French & Withrow, J. Ind. d Eng. Chem. 6, 148 (1914). *Teeple, /. Am. Chem. Soc. SO, 412 (1908). 422 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS melting of varnish gums are sometimes recovered but, even after good purification, have never found favor as thinners with paint and varnish manufacturers. The softer grades of Manila copal 5 yield 10 to 12 per cent of its weight of oil, largely limonene and i-limonene, during the first part of the fusion, up to about 330. Fresh Queensland Kauri gum, from Agathis robusta, yields about 11.6 per cent of nearly pure a-pinene. 6 Parry 7 gives the following physical properties of turpentine as the result of the examination of a large number of commercial samples. Specific gravity at 15 0.862- 0.870 n 1.468- 1.473 D Initial boiling-point 154. -155.5 Distillate below 160 72.% - 74.5% Distillate below 170 95.% - 97.5% Iodine value, Hiibl 360. -375. Iodine value, Wijs 335. -350. The optical rotatory power is subject to considerable variation. Herty 8 found the oil from P. palustris to vary from 7 26' to + 18 18' and that from P. heterophylla, 29 26' to + 15'. The volatile oils of several species of pine found in the western States have been examined by A. W. Schorger, 9 who finds that the turpentine from P. Ponderosa (Laws) and P. Scopulorum (Eng.) con- sists largely of (3-pinene (q.v.) : that from P. Sabiniana is practically pure n. heptane and that from P. contorta consists largely of (3-phel- landrene. These oils are not likely to become of commercial im- portance. Pinus sylvestris is the chief source of Swedish and Russian tur- pentine and contains sylvestrene in addition to dipentene and (3-pinene 10 and possibly a-pinene and Z.eamphene. Russian turpen- tine is a very indefinite product containing considerable proportions of phenolic or acid substances and oil boiling above 180. French oil of turpentine, which constitutes nearly 20 per cent of the world's supply, is derived from Pinus pinaster (Pinus maritima) and is a true pinene turpentine consisting chiefly xl of Z.-a-pinene, [a] 20 to 38. It is suitable for the manufacture of artificial 6 Brooks, Philippine J. Sci. 1910, 203. Baker & Smith, "A Research on the Pines of Australia," Sydney, 1910, p. 376. T Chemistry of Essential Oils, Ed. 3, Vol. I, 17. 8 J. Am. Chem. Soc. 30, 863 (1908). 9 Bull. 119, U. S. Dept. Agriculture. 10 Chem. Ztff. 32, 8 (1908). "Darmois (Chem. Zentr. 1910 [1], 30) concludes, from studies on its optical rota- tion, that this turpentine consists of approximately 62% a-pinene and 38% /3-pinene. BICYCLIC NON-BENZENOID HYDROCARBONS 423 camphor or other uses to which a true pinene oil can be put. With this brief review of the character of commercial turpentines the chem- istry of the pinenes and their more important derivatives will be noted. With the elucidation of the constitution of (3-pinene and its synthesis by Wallach in 1908, the chemistry of the pinenes is prac- tically complete. a-Pinene is one of the most widely distributed of the terpenes, having been found in the essential oils of a large number of Coniferse, the grass oils, Lauraceae, Labiatae, etc. When it occurs together with other terpenes in oils used for the manufacture of flavoring extracts or perfumes, it is common practice to separate the terpenes by making use of their lesser solubility in dilute alcohol, 12 as compared with the esters, alcohols, aldehydes and the like which give such oils their aromatic value. The resulting ter- pene-free oils can be dissolved in much more dilute alcohol, thereby effecting considerable saving in the preparation of these solutions. Inactive a-pinene is one of the few terpenes which have been iso- lated in quite a pure condition. Fairly pure a-pinene can be obtained by fractional distillation of turpentine 13 but a purer product can be obtained by preparing the nitrosochloride, purifying this by fractional crystallization and regenerating the a-pinene by decomposing the nitrosochloride by aniline 14 in alcoholic solution. Such a sample described by Schimmel & Co. 15 had the following physical properties: 20 boiling-point 154.5-155, d 1 _ 0.8634, n_-1.4664, optically inactive. lo D The highest observed optical rotations of a-pinene are [] n + 51.52 in the case of pinene isolated by A. W. Schorger 16 from the oil of the Port Orford cedar (Chamcecyparis lawsoniana). This is probably the purest natural a-pinene thus far discovered. A very pure d. a-pinene [aU + 48.4 has been noted in the case of a specimen isolated from Greek turpentine (from the Aleppo pine, P. halepensis) , 17 and a laevo- pinene [a] 48.63 from one of the eucalyptus oils, E. lavopinea. 1 * 12 Bocker, J. prakt. CTiem. (2) 89, 199 (1914) ; Vezes & Mouline, Bull. soc. cMm. (3) 31, 1043 (1904). See section on physical properties (solubility). 13 Henderson & Sutherland recommend fractional distillation with steam, followed by ordinary distillation, taking the fraction boiling at 155-156 as a-oinene. (J. Chem. Soc. 101, 2289 [1912]). "Wallach, Ann. 258, 343 (1890). 35 Gildemeister, "Die Aetherischen Oele," Ed. 2, Vol. 1, 308. 16 J. Ind. & Eng. Chem. 6, 631 (1914). "Vezes, Bull. soc. chim. (4 5, 932 (1909). "Smith, J. d Proc. Soc. X. 8. W. 32, 195 (1898). 424 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Oils of high optical rotation give very poor yields of crystalline nitro- sochloride. a-Pinene is usually identified by preparing the nitrosochloride of the fraction boiling below 160, which preparation is carried out by slowly adding concentrated hydrochloric acid to a strongly cooled solution of the hydrocarbon in glacial acetic acid and ethyl or amyl nitrite. On standing the crystalline nitrosochloride separates. After separating the crystals and recrystallization by dissolving in chloro- form and precipitating with methyl alcohol, the nitrosochloride melts at 103 but the nitrolamines give melting-points which are more use- ful and reliable for identification purposes. The pinene nitrolpiperi- dine melts at 118-119 and the nitrobenzylamine melts at 122-123. In preparing the nitrolpiperidine a small proportion of nitrosopinene is simultaneously formed. 19 In the case of pinene of high optical activity recourse may be had to oxidation by permanganate to the pinonic acids. 20 The hydrochloride (bornyl chloride) made by pass- ing dry hydrogen chloride into cooled pinene, carefully dried by dis- tillation over sodium, has also been employed for the detection of pinene although both a and p-pinenes give the same hydrochloride, melting-point 127. The constitution of a-pinene has been determined largely by a study of its oxidation products. One of the most important advances made in clearing up the chemistry of the terpenes was the recognition, first clearly set forth by Wagner, that the hydroxyl group in a-terpi- neol is in position (8) and not position (4) . In this same remarkable communication of Wagner, 21 which was published in full in the Rus- a-pinene (Wagner) H + H.O H r H, ^ H Ss-r Ha r Jc. OH er) ^X^ X. CH^ X CH, "Wallach, Ann. 2tf t 252 (1888). Confirmed by Bushujew, J. Rusa. Phya.-Chcm. 800. 41, 1481 (1910). a Schimmel & Co. Semi-Ann. Rep. 1909 (1), 120. ai Ber. 27, 2270 (1894). BIG YC LIC NON-BENZENOID HYDROCARBONS 425 sian language, Wagner published what have proven to be the correct constitutions of limonene, carvone, dihydrocarvone, carone and a-pinene. According to Wagner's structure for ct-pinene, the forma- tion of cc-terpineol and terpin is formulated as shown on the preced- ing page. Wagner seemed to have an almost uncanny ability to visualize the constitution of such substances. Baeyer showed that a series of oxidation products obtained by him also are in accord with Wagner's a-pinene constitution, which oxida- tions he expressed as follows, 22 a-pinene a-pinonic acid pinoylformic acid CO,H Hqc pinic add norpinic acid 23 Just as the four carbon ring in pinene is broken by dilute acids to form terpineol, so also is the four carbon ring in a-pinonic acid broken to give the methyl ketone of homoterpenylic acid, identical with the product of the oxidation of terpineol itself. (See page 426.) a-Pinene is usually associated with the isomeric hydrocarbon, |3-pinene, and oxidation by permanganate gives the products of oxida- 2 *Ber. W, 2775 (1896). 23 The cyclobutane ring has about equal stability in pinene, pinonic acid and pinoylformic acid, being split with about equal ease by dilute acids. In pinic and norpinic acid it is very much more stable, this stablity being due apparently to the influence of the carboxyl group, parallel to the observations of Buchner on the effect of the carboxyl group on the stability of the cyclopropane ring. 426 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS c/' HO.C a-pinonic acid ,8, CM,' methyl ketone of ^ homoterpenylic acid HQ; ci a-terpineol tion characteristic of these two hydrocarbons. The oxidation is car- ried out 25 as follows: 5 cc. of the hydrocarbon are shaken for about three hours with an ice -cold solution of 12 g. potassium permanganate, 2.5 g. caustic soda, 200 cc. of water and 500 g. ice. After 3 hours saturate with carbon dioxide and remove the volatile unoxidized oil by distillation with steam, filter and evaporate in a current of carbon dioxide to about 200 cc. and extract several times with chloroform. On further evaporation the first salt to separate out is sodium nopinate, which on acidifying gives crystalline nopinic acid, melting at 125. This acid is characteristic for |3-pinene. The sodium salt of pinonic acid is more soluble than the nopinate. Barbier and Grignard 26 have investigated the optically active forms of pinonic acid obtained by the oxidation of d. and Z.a-pinene of high optical rotation. From Z.pinene, [a] 37.2 Lpinonic acid was obtained, by permanganate oxidation and after distillation in vacuo, 189-195 at 18 mm., sepa- rated in long crystals melting at 67-69, and [a] 90.5. From 24 The constitution of homoterpenylic and terpenylic acids is discussed in connec- tion with terpineol and limonene. "Schimmel & Co. Semi- Ann. Rep. 1910 (1), 165. "Compt. rend, lift, 597 (1908). BICYCLIC NON-BENZENOID HYDROCARBONS 427 d.a-pinene, [] n + 39.4, they obtained a mixture of racemic and dpinonic acids, the latter melting when recrystallized at 67-68, [a] n + 89.0 and when mixed with the i.pinonic acid the racemic acid melting at 104 was obtained. Harries 27 investigated the action of ozone on a-pinene and by heating the resulting ozonide with acetic acid to 90 obtained an oil boiling over the wide range of 100-142 under 12 mm., from which he prepared a semicarbazone melting at 214-215 which was "prob- ably" pinonic aldehyde. On standing in contact with moist air, as in loosely stoppered containers, particularly in sunlight, turpentine or a-pinene is oxidized to pinol hydrate (sobrerol) which crystallizes from the oil. 28 From d. or ^.turpentine the correspondingly active pinol hydrates, melting-point 150, are obtained. The d-l hydrate is formed on treating pinol with hydrogen bromide followed by hydrol- ysis by alkali. The relations of pinol hydrate and pinol are indicated by the results on oxidizing with permanganate. Each adds two hydroxyl groups, pinol to form pinol glycol, C 10 H 16 0. (OH) 2 and the hydrate to form sobrerythrite 29 C 10 H 16 (OH) 4 . Pinol glycol is also formed by the action of dilute acids on the dioxide, pinol oxide, C 10 H 16 2 . Sobrerythrite is also formed by the action of hypochlorous acid on pinene and hydrolysis of the dichlorohydrin. In accord with the general behavior of the higher alkylene oxides (q.v.) concentrated alkalies convert the dichlorohydrin to the dioxide, pinol oxide, and on treating pinol oxide with dilute acids the 1.2 oxide is hydrolyzed to pinol glycol, leaving the oxide ring of four carbon atoms unchanged. Parallel with the behavior described in connection with cineol and other oxides, heating pinol hydrate with dilute acids causes the /or- mation of the oxide pinol. These facts show clearly the relation be- tween the number of carbon atoms in the oxide ring and their relative stability. (See figure on page 428.) The behavior of turpentine or pinene on air oxidation is, in gen- eral, typical of the behavior of the olefines, including unsaturated petroleum oils. With all such substances air oxidation is accom- " Ber. tf, 879 (1909). 28 Formic acid is one of the products of the oxidation of turpentine by air and metal containers are accordingly sometimes corroded by old turpentine. Formic acid produced in this way is probably a product of the oxidation of /3-pinene, not o-pinene. 29 The sobrerythrite made from pinol hydrate melts at 156 ; a stereoisomeric sobrerythrite made by the action of hypochlorous acid on pinene melts at 194. 428 CHEMISTRY OF THE NON-BENZENO1D HYDROCARBONS CH, CH, pmol oxide binol olycol P 1 not panied by the formation of organic peroxides, water, carbon dioxide, simple organic acids, resinous substances and other oxidation products among which alcohols, aldehydes and ketones have frequently been noted. The oxidizing power of old turpentine was observed by Schon- bein and frequently investigated subsequently by others. The organic peroxides formed in this way are rapidly destroyed by heating to 140 and are hydrolyzed by water. As pointed out by Engler and Weiss- berg 81 the peroxides decompose, causing further oxidation of other 80 The positions of the chlorine atoms and hydroxyl groups may be reversed ; in the above constitution their positions are arbitrarily assigned. The dichlorohydrine of pinene may be a mixture of isomers, just as the addition of HOC! fo propylene gives a mixture of CH 8 CHC1.CH 2 OH and CH 3 CHOH.CH,C1. M Vorgange 0.9742. 436 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS hydrocarbon probably has two double bonds in a conjugated position as in a-terpinene. Dehydration of verbenol with phosphoric oxide or zinc chloride gave cymene. When /.verbenene, prepared from ver- benol, is brominated in chloroform the crystalline dibromide, melt- ing at 70-72, is formed. The d.dibromide from d verbenene nat- urally melts at the same temperature but the racemic dibromide melts at 50-52. Oxidation of verbenene by permanganate yields norpinic acid melting at 1 75.5 -l 76.5, and treatment with zinc chloride yields p-cymene. Blumann and Zeitschel 64 regard verbenene as having the constitution shown below, I; on reduction by sodium and alcohol two atoms of hydrogen are added (a reduction usually possible when the double bonds are conjugated) and the resulting hydrocarbon, dihydro- verbenene or "S-pinene" they regard as having the constitution indi- cated by II. /. verbenene 0.8867 II. dihydroverbenenef B.-P 158-159 (762 mm.) 20 n-p- 1.4980 B.-P 159M60' "Ber. 5k, 887 (1921). 20 0.8625 1.4662 BIG YC LIC NON-BENZENOID HYDROCARBONS CH, 437 verbenol verbenene f An alcohol isomeric with verbenol and also possessing the bridged ring structure of pinene, is myrtenol, an alcohol occurring in myrtle oil as the acetate. 65 Its constitution is shown by the fact that reduc- tion of the corresponding chloride by sodium and alcohol yields a-pmene 66 Oxidation by chromic acid yields the corresponding aldehyde myr- tenal, but permanganate oxidizes it to d.pinic acid. The conversion of pinene to derivatives of borneol is well known, the best example being the formation of bornyl chloride (so-called pinene hydrochloride) by the action of dry hydrogen chloride on "Soden & Elze, Chem. Ztg. 29, 1031 (1905). Semmler & Bartelt, Ber. W, 1363 (1907) ; myrtenol boils at 222-224' (760mm.) or 102-105 (9mm.). 438 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS pinene at about 15. When bornyl chloride is prepared from pinene in the usual manner, the product is usually optically inactive but Barbier and Grignard 67 noted a rotation of . 25.20 for the hydrochloride, melting-point 127, made from L-a-pinene from French turpentine, and a d.pinene hydrochloride (bornyl chloride), [a] + 33.19, melting-point 127.1, has been prepared 68 from d.a- pinene from Greek turpentine. The hydrochloride made by Lynn from highly active da-pinene from the Port Orford cedar was inac- tive, from which observations, together with many other observations of similar kind, it is evident that racemization occurs very readily, but under certain conditions, efficiency of cooling or rate of reaction, the activity may be partially preserved. In preparing limonene mono- hydrochloride partial racemization occurs, the degree of racemization apparently being influenced by the rate of introducing the hydrogen chloride, as shown by Vavon. 69 Pinene hydroiodide (bornyl iodide) was made by Aschan 70 by digesting bornyl chloride in ether with magnesium iodide; the iodide is easily reduced by zinc in acetic acid to camphane. True pinene hydrochloride has not been detected among the reac- tion products of a-pinene and hydrogen chloride, but was synthesized by Wallach from nopinone by the Grignard reaction, thus making niethymopinol, and replacing the hydroxyl group in this alcohol by chlorine by means of phosphorus pentachloride, 71 nopinone methyl nopinol pinene hydrochloride "Bull, soc. chim. (4) 15/26 (1914). "Tsakalotos, J. pharm. chim. 1J,, 97 (1916). 99 Bull. soc. chim. (4) 15, 282 (1914). 70 Ber. 1,5, 2395 (1912) ; d. or Z.pinene gives a hydroiodide melting at 3 to 5" ; d.l.pinene gives the racemic hydroiodide melting at 12. Silver oxide in dilute alcohol conyerts o the iodide into an evidently new unaaturated alcohol, CioH 17 OH, boiling-point l Wallach, Ann. 356, 246 (1907). BICYCLIC NON-BENZENOID HYDROCARBONS 439 True pinene hydrochloride, as contrasted with bornyl chloride, is very unstable and decomposes at its boiling-point, 200-205, and is very readily converted to dipentene dihydrochloride by the action of hydrogen chloride ; bornyl chloride is not affected by hydrogen chloride. By treating pinene dibromide with zinc in alcoholic solu- tion a tricyclic hydrocarbon, melting at 65-66, is obtained. A di- iodide, prepared by Frankforter and Poppe, 72 is very unstable, entirely losing its iodine merely on standing or by distilling a few times. Anhydrous oxalic acid gives a relatively small yield of bornyl esters, dipentene and terpinenes being the chief products (see Artifi- cial Camphor). Acetic acid at 200 also gives a certain amount of bornyl acetate. 73 The oxalic acid reaction was the basis of the first industrial process for the manufacture of artificial camphor. When pinene is treated with HC1 in the presence of moisture, or at too high temperatures, oily mixtures are obtained, the chief product being dipentene dihydrochloride. Under the best conditions the yield of crystalline bornyl chloride does not exceed 75 to 78 per cent of the theory. The liquid, oily chloride mixture contains bornyl chloride in solution, also dipentene dihydrochloride and lesser amounts of other substances. Barbier and Grignard 74 have investigated these hydro- chloride oils, converting these hydrochloride oils into the magnesium compounds and treating the latter with oxygen and also with carbon dioxide. In addition to bornyl chloride, they found indications of the presence of fenchyl chloride. Aschan 75 has carefully investigated these oily hydrochlorides, having at his disposal comparatively large quantities of material made incidental to the manufacture of artificial camphor. By the action of alkali on the chlorides he obtained a com- plex mixture of hydrocarbons and showed that the low-boiling fraction contained (1) d.Z.bornylene (which yields d.Z.camphoric acid on oxida- tion), (2) a bicyclic hydrocarbon boiling at 144-145 which he called a-pinolene, and (3), a tricyclic hydrocarbon, boiling-point 143, which was quite stable to permanganate and which he named fi-pinolene or tricyclene. This hydrocarbon, which has been obtained as one of the products of the decomposition of fenchyl chloride by Aschan 76 and by Sandelin, 77 is probably identical with the cyclofenchene of Quist. 78 *J. Am. Chem. Soc. 28, 1461 (1906). 'Austerweil, Compt. rend. 11,8, 1197 (1909). *Bull. soc. chim. (4) 7, 342 (1916). 5 Ber. 1,0, 2750 (1907) ; Ann. 337, 27 (1912). Ann. S87, 27 (1912). T Ann. 396, 297 (1913). Ann. i*7, 278 (1918). 440 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS By decomposing fenchyl alcohol by heating with potassium acid sul- fate, Quist obtained two hydrocarbons, one being the low-boiling "cyclofenchene" or (3-pinolene. Fenchyl alcohol cannot decompose to water and an unsaturated hydrocarbon, forming a double bond with an adjacent carbon atom, as will be evident from its constitution. Other hydrocarbons may be formed from (3-pinolene by rearrangement. Quist confirms Aschan as to its stability to permanganate but discovered that the three-carbon ring is evidently broken by the addition of bromine, forming a well crystalline dibromide of unknown constitu- tion. The chemistry of the fenchenes (q.v.) into which these deriva- tives of pinene lead, is still in a very unsettled condition. As regards their formation from a-pinene Aschan recalls that when hydrogen chloride reacts with tetramethyl ethylene, a rearrangement occurs. CH \ CH 3 HC1 CH C CHC1.CH 3 which is analogous to the addition of HC1 to pinene, and if we recall CH 3 that the CH 2 and >C< groups in the four-carbon ring are CH 8 equivalent as regards their spatial relations to the rest of the mole- cule, we may write the rearrangement of the initial hydrochloride as follows, H,C bornyl chloride H,C H 2 C BICYCLIC NON-BENZENOW HYDROCARBONS CH 3 CH 3 H 2 C H 9 C CH, CH 441 H C< Cl CH, C< CH 3 chloride of fenchyl alcohol According to Aschan and Quist the formation of p-pinolene (cyclo- fenchene), from fenchyl alcohol or fenchyl chloride is to be expressed as follows, fenchyl alcohol Isopinene is the name given by Aschan 7g to a hydrocarbon, boiling- point 154.5-155.5, d 0.8658, n 1.47025, obtained by reacting upon p-pinolene with hydrogen chloride and then decomposing the hydrochloride with aniline. Aschan identified os-apocamphoric acid among the oxidation products of iso-pinene. Aschan reasons that isopinene, barring rearrangements, can have only structure I or II T Chem. Zentr. 1909 (2), 26. 442 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS H 9 C H C = CH 9 H 9 C H -C- _C CH, H 2 C CH 3 C CH S CH CH ? CH C CH H,C in v^Xl - Wallach considers that d.Z.fenchene has the constitution represented by I, and also II best accounts for the formation of apocamphoric acid. H H H 2 C C C CH 3 H 2 C C CO-CH 3 CH, C CH. C H isopinene H -C CH 3 C CH 3 C H fenchenonic acid CCLH ccga H 9 C CH 3 C CH, C H apocamphoric acid C(XH The formation of isopinene by the rupture of the three-carbon ring in p-pinolene and the subsequent removal of HC1 may be understood by the following reactions, CH, ) CH, CH S H H fi-pinolene B.-P. 142-144 a>H CH 2 H fi-pinolene hydrochloride M.-P. 26 a CHpC-CH, XH, @- CH 3 -C-CH 3 H 2 C - CH - CH 2 C0 2 H H 2 C -- CH I* 7 Fenchone melts at 5 to 6. The physical properties of a specimen regenerated from the semicarbazone [Wallach, Ann. 562, 195 (1908)] were as follows, boiling-point I 192-193, d 18 o 0.948, [a] D + 62.76 (higher in alcohol solution) n^-1.46355. Fen- chone does not form a phenylhydrazone but readily gives an oxime, d. and I. melting at 164-165, inactive form melting at 158-160 [Wallach, Ann. 272, 104 (1893)]. The semicarbazone forms very slowly ; Wallach recommends allowing an alcoholic solution of the ketone, semicarbazid hydrochloride and sodium acetate to stand for two weeks [Ann. 353 } 211 (1907)]. The semicarbazone crystallizes from dilute alcohol in long prisms melting at 182-183. 88 Ann. SOO, 320. "Ann. 315, 293 (1901). 1(W Komppa, Ber. kk> 395 (1911). The racemic semicarbazone of fenchocamphorone f melts at 220. 448 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS By reacting upon a-fenchocamphorone with methyl magnesium iodide Komppa and Roschier 101 have synthesized-a- fenchene. The tertiary alcohol formed by this reaction is decomposed by distilling at ordinary atmospheric pressure to give the hydrocarbon. OH CH, CH C = CH 2 CH JH 2 CH 2 CH C CH 3 CH 2 CH C = CH 2 CH 3 -C-CH 3 [ or !H- >^ A, The physical properties and chemical behavior of the synthetic a-fenchene 102 are practically identical with Aschan's isopinene (q.v.), but Wallach considers that a-fenchene contains the >C = CH 2 group, on account of its formation together with p-pinene when nopinolacetic acid is dehydrated. 103 In a similar manner Komppa and Roschier have treated d.Z.p-fen- chocamphorone with magnesium-methyl iodide, thus forming methyl- p-fenchocamphorol. 104 When this alcohol is heated with potassium acid sulfate a mixture of two unsaturated hydrocarbons is obtained, consisting mainly of d.Lp-fenchene and an endocyclic hydrocarbon y-fenchene. The (3 hydrocarbon yields d.l. hydroxy-p-fenchenic acid melting at 124-125 on oxidizing with permanganate, and on further oxidation by the lead peroxide and sulfuric acid method d.Z.p-fencho- camphorone is obtained. The latter ketone by further oxidation yields apofenchocamphoric acid. A fourth fenchene, termed isoallo- fenchene by Semmler, and isofenchylene by Quist, is called 5 or iso- fenchene by Komppa. The p-fenchene of Komppa is Wallach's D, d. or L. I. fenchene, or Semmler's isofenchene. 101 J. Chem. Soc. in (1), 466 (1917). 102 The synthetic hydrocarbon of Komppa boils at 154-156, has a density ^ 20 0.8660 and refractive index ^- 1.47045. The hydrochloride, melting-point 35-37 is identical with that made from isopinene. Ozone gives racemic fenchocamphorone and r a-fenchenylanic acid, melting-point 105. 1C BICYCLIC NON-BENZENOID HYDROCARBONS H CH 3 H C 449 CH H,C CH 2 C H OH. methyl-p-fenchocamphorol CH 3 CH 3 H -C CH CH 3 H y-fenchene CH 3 Fenchone itself has been synthesized by Ruzicka 105 in the manner indicated by the following reactions. (1) Levulinic ester and ethyl bromoacetate are condensed by means of zinc and the resulting lactonic ester is converted into the I nitrile by treating with potassium cyanide, H 9 C CO CH, CH. H 2 C C CHXXXR HC C0CH BrCH 2 .C0 2 R \ OH >H,C C0 2 C 2 H 5 CH 3 CH 3 CH 3 |H 2 C C CH 2 C0 2 R H 2 C C CN H 2 C C C0 2 H. \ ( CO C, C, H,C0 2 R - 1 CH 2 -C0 2 H. H 2 C C0 2 R H 2 C C0 2 H . (2) The above tricarboxylic acid is condensed to a cyclopen- jtanone derivative by means of sodium in benzene, and the ester of the suiting product is condensed with bromoacetic ester. Ber. 50, 1362 (1917), 450 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CH 3 __i_C0 2 R 4-Br.CH 2 C0 2 R Hc HC C = H 2 C C C0 2 R I AH, H 2 C C CH 2 C0 2 R OH (3) The above hydroxy acid is converted to the unsaturated acid and the latter reduced to the saturated acid. , CH 3 H 2 C C C0 2 R CH 2 H 2 C C = CH.C0 2 R CH 3 H 2 C C C0 2 R CIL H. -As CHoCOoR (4) On heating the lead salt of the above acid methylnorcamphor is formed which on methylating by Haller's method, using sodium amide and methyl iodide, fenchone and fenchosantenone are formed. CH, H,C C C0 2 H CH 2 H. CH CH 2 C0 2 H CH S H 2 C C CO I AH CH 2 , H 2 C CH CH 2 methylnorcamphor H 9 C H. CH 3 Aro CH 3 HP p r CH 1 P I . rn rn nn -> | {H, ; prn CH r fenchosantenone fenchone CH 3 CH, The above structure of fenchone explains the formation of fencholic acid from fenchone by heating with caustic potash. 106 "Wallach, Ann. 369, 71 (1909). BICYCLIC NON-BENZENOID HYDROCARBONS CH 3 CH 3 H 2 C- -C- -CO H 2 C C CO,H. CH 2 | by KOH [ C _i H _i< CH3 CH 3 451 H H CH, CH CH< CH 3 fencholic acid By the action of sodium or potassium acid sulfates on fenchyl alcohol at 170-180, in a current of carbon dioxide, a fenchene is obtained boiling at 151-153, D 0.8660. On oxidation it yields hydroxyfenchenic acid melting at 138-139. 107 Fenchyl chloride, like the higher alkyl halides in general, reacts very slowly with magnesium in ether. After one week and treating with carbon dioxide the reaction mixture gives chiefly hydrofenchene carboxylic acid and hydrodifenchene. 108 C0 2 H CH 3 CH r r k | CH 2 C C< CH 3 X* S* H CH 3 hydrodifenchene H CH 3 hydrofenchenecarboxylic acid M.-P. 45^6 (racemic acid M.-P. 52-53) By the action of ozone on a-fenchene Komppa and Hintikka 109 obtained fenchocamphorone, which behavior is also readily explained by Wallach's formula for this hydrocarbon. H C = 107 ' o 1 000 V ? 1 . ine s & 5 ft II s s S S a s ^2 ^^ Cj rt ^J 3 .2 . Ill > >, ;-s rfl .JM ,3 ^_C! OQ fl o3 o flo a ^ s s _3 *ac I ^ 9-af^ t-id CO-^iO 13 BICYCLIC NON-BENZENOID HYDROCARBONS 455 Also when borneol xanthogenate is heated it decomposes to give chiefly bornylene, a method discovered by Tschugaeff. 5 Henderson and Caw 6 showed that when bornylene is prepared by Tschugaeff's method, the impurities can be removed by oxidizing with hydrogen peroxide, the bornylene so purified melting at 113, and boiling at 146-147 (750mm.). Bornylene has recently been made from cam- phor by Ruzicka 7 by the conversion of camphor to bornylamine by heating with ammonium formate in an autoclave at 60 atmospheres pressure; the resulting amine was subjected to the method of exhaus- tive methylation with methyl iodide (bornyltrimethylammonium iodide melts at 245). The free trimethyl base is gently decomposed to give bornylene melting-point 111-112. Bredt has also made a very pure bornylene from camphocarboxylic acid. By electrolytic reduction of this acid to the corresponding hydroxy acids 8 and distilling the acetylborneolcarboxylic acid the unsaturated acid, bornylenecarboxylic acid, 9 is obtained and on react- ing upon this acid with hydrogen bromide a mixture of a and (3-bro- mocamphanecarboxylic acids are obtained. The p-bromo acid is decomposed on heating with aqueous alkali to bornylene, bornylene- carboxylic acid and a lactone. CH.C0 2 H CH.C0 2 H C CO 2 H. -> C 8 H 14 < || .OH. CH . C 8 H 14 < | -> C 8 H 14 < | -> C 8 H 14 < || + HBr C = CH. CH.CCXH CBr.C0 2 H. - > C 8 H 14 < | and C 8 H 14 < | CHBr \ CH 2 fi-bromocamphene- \ carboxylic acid. \ CH C 8 H 14 <|| bornylene CH. The bornylene thus obtained had the following physical properties, melting-point 113, boiling-point 146 (740mm.), [a] 21.69 (10.4% in toluene), [a] 26.96 (4.42f in methyl alcohol). Bornylene gives camphoric acid on oxidation, which indicates that 5 Tschugaeff & Budrick, Ann. 388, 280 (1912). a 7. Chem. Soc. 103, 1543 (1912). 7 Helv. chim. Acta. 3, 48 (1920). 8 Bredt, Ann. 366, 1 (1909). These are ois and cis-trans isomeric forms of borneol- carboxylic acid, the cis acid melting at 102-103 and the cis-trana melting at 171. Melting-point 112-113, boiling-point 158 at 13mm. 456 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS the double bond is in the position shown, which structure is that for merly supposed to represent the constitution of camphene, CH 3 CH 3 -C CH CH 2 C CH CH, -C-CH, CH H CO,H. CH CH, -C-C -C H, H -C0 2 H. camphoric acid bornylene Camphene forms an ozonide which on decomposition gives formal- dehyde-camphenilone and dimethylnorcampholide. 10 Komppa and Hintikka " have synthesized the latter substance and shown its con-| stitution to be as represented in the following. CH 3 CH 3 CH 2 CH C< CH 2 in CH 3 camphene (Wagner) and O'-L- I>2 CH 2 ^^ o o r\ ^J-J^ V_^J.J. - \^ - V/ dimethylnorcampholide = camphenilone This constitution of camphenilone is supported by the fact that iti does not form a hydroxymethylene compound 12 and therefore does not possess a CH 2 group adjacent to the carbonyl group. Further evidence of the structure of camphenilone is given by the conversion of camphenilone by the action of sodamide, to the amide of the acid, CH 3 CH, CH- H 2 v^ i CH< CH, CH which substance has also been synthesized. 13 10 Harries & Palmen, Ber. A3. 1432 (1910). ll Ber. 42, 898 (1909). "Moycho & Zienkowski, Ann. 31,0, 54 (1905). "Bouveault & Blanc, Compt. rend. Uft, 1314 (1908), BICYCLIC NON-BENZENOID HYDROCARBONS 457 When camphene is oxidized by alkaline permanganate the chief product is camphenic acid, C 10 H 16 4 , an acid isomeric with camphoric acid. A great deal of work has been done upon the structure of this acid, based upon which other constitutions for camphene have been proposed. 14 However no reasonable doubt should exist as to its con- stitution since its synthesis by Lipp, 15 in the following manner: The ethyl ester of 1 . 3-cyclopentanonecarboxylic acid was condensed, by Reformatsky's reaction, with a-bromoisobutyric ester in the presence of zinc. By decomposition with loss of water an unsaturated acid was formed, whose constitution may be either III or IV but on hydro- genating the saturated acid 1.3-carboxylcyclopentylisobutyric acid was formed, which proved to be identical with dJ.cis-camphenic acid. 16 OH CH 3 CH 2 CO CH 2 C C< I CH 3 CH 2 > CH 2 C0 2 R. > CH 2 CH . C0 2 R CH 2 CH C0 2 R . I. II. CH 3 CH 3 ^ ^<^ 1 CH 3 CH C0 2 R v^n I - \u CH 2 ^<^ 1 CH C0 2 R ATI rjo T> L UUg is C0 2 R III. * IV. CH 3 CH CH -c< 1 CH 2 1 CH 3 C0 2 H r,TT A* ro TT camphenic acid When camphenic acid is distilled it loses C0 2 to form a ketonic acid, camphenonic acid, whose constitution may be inferred from the structure of camphenic acid. The d. and L forms of camphenonic JJCf. review by Haworth & King, J. Chem. Soc. 101, 1975 (1912). 18 Camphenic acid is practically insoluble in cold water, ligroin and carbon bisul- fide, but crystallizes from hot water, melting-point 135-137. 458 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS acid together with the d.l. acid are formed from camphenic acid made by the oxidation of strongly d. or I. camphene. 17 CH CH v_> I CH- H 2 CH, CH C< I CH 3 C0 2 H ^ C(XH CH 2 CH C< CH 3 CH 2 C C0 2 H. camphenonic acid. camphenic acid Fusion of camphenonic acid with caustic alkali or treatment with sodium and alcohol regenerates camphenic acid. Wagner's constitution for camphene is also supported by the work of Buchner and Weigand, 18 who condensed camphene with diazo- acetic ester and oxidized the acid so obtained to 1 . 1 . 2-cyclopropane- tricarboxylic acid. The camphene employed by Buchner melted at 44-45 and distilled at 156-157. Treatment with the ester at 160-165 in the presence of copper powder gave vigorous evolution of nitrogen and a good yield of the condensation product, 2 . 2-dimethyl- norcamphane-3-spirocyclopropanemethylcarboxylate. The relation of camphene to the condensation product and the oxidation product are as follows: CH ; CH, H camphene H rj CH 3 n PH CH 2 H \/ CH.C0 2 R CCLH. CH, C0 2 H. CH.C0 2 H. The spiro ester is stable to permanganate in suspension in sodium carbonate solution. [Buchner and Weigard also succeeded in mak- "Aschan, Ann. 1,10, 240 (1915). "Ber. 46, 759 (1913). BICYCLIC NON-BENZENOID HYDROCARBONS 459 ing the acid chloride, from which the amide was prepared, leaflets melting at 124.] The purified amide readily yields the pure acid, melting at 108. Reduction of the ester, by sodium in absolute alcohol, converts the C0 2 R group to CH 2 OH with rupture of the cyclopropane ring. Applying the same method to bornylene Buchner and Weigand 19 obtained 1 . 2 . 3-cyclopropane tricarboxylic acid. CH.C0 2 H CH.CXXH. When camphene hydrochloride is carefully treated with dilute alkali, camphene hydrate is formed, which can be decomposed to camphene having the same rotatory power as the original hydrocar- bon. The hydrate is therefore formed without causing any change in the carbon structure of camphene. Only two formula for this hydrate are possible, one being a tertiary and one a primary alcohol, but the properties of camphene hydrate are clearly those of a tertiary alcohol, i.e., camphene hydrate 19 Ber. 46, 2108 (1913). The trimethyl ester of 1.2.3-cyclopropanetricarboxylic acid melts at 56-57, which serves to distinguish it from its isomers. 460 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS A stereoisomeric form of camphene hydrate is methylcamphenilol, obtained by the action of magnesium-methyl iodide on camphenilone. Both forms yield camphene on dehydration and probably bear a stereochemical relation with each other comparable to borneol and isoborneol. 20 To explain the rearrangement which occurs when isoborneol is decomposed to form camphene, the intermediate formation of tri- cyclene has usually been assumed, =CH 2 /CH, CH 3 isoborneol ene tric^clene Tiffeneau 21 has proposed the theory that when alcohols are decom- posed with the formation of a hydrocarbon of a different carbon structure, as in the decomposition of pinacoline alcohol to tetramethyl- ethylene, that the intermediate product is a hydrocarbon having a bivalent carbon atom. In the case of isoborneol and its decomposi- tion to camphene this would be represented as follows, "Aschan, Ann. W, 222 (1915). gen. d. 8ci. 18, 583 (1907). BICYCLIC NON-BENZEN01D HYDROCARBONS 461 Meerwein 22 has tested both of these hypotheses. Camphor hydra- zone is decomposed by mercuric oxide, the intermediate compound CH 2 C 8 H 14 < I C = N.NH.HgOH being decomposed with evolution of nitro- gen, and it is a reasonable supposition that the bivalent carbon com- CH 2 C 8 H 14 < | pound C< whose transitory existence is assumed by the Tif- feneau theory, would be formed and immediately rearrange to cam- phene, if this theory is correct. It is found, however, that tricyclene is formed almost quantitatively. The properties of tricyclene clearly show that it cannot be an in- termediate product in the conversion of isoborneol to camphene. Thus Meerwein shows that under the conditions by which isoborneol is almost quantitatively changed to camphene (heating with 33% sulfuric acid at 100), tricyclene is practically unchanged. Also, as shown by Lipp, 23 heating tricyclene with fused zinc chloride is with- out effect although isoborneol is decomposed to camphene under these conditions. As regards the opposite reaction, the conversion of camphene to isoborneol (or acetate) , Meerwein shows that chloroacetic acid reacts more rapidly with camphene than with tricyclene, and consequently tricyclene cannot be an intermediate product in the conversion of camphene to esters of isoborneol. In these experiments evidence was found that tricyclene first forms an ester of camphene hydrate. When tricyclene or camphene is treated with hydrogen chloride in cold ethereal solution, a very unstable hydrochloride is formed which Meer- wein regards as the true chloride of camphene hydrate. It is so unstable that on merely shaking the chloride with water at ordinary temperatures, camphene hydrate is formed; in alcoholic caustic potash the neutralization of the alkali takes place so rapidly that the per cent of the hydrochloride present can be titrated in the cold with N/2 caustic solution during one half hour. The most striking prop- erty of this hydrochloride is its rearrangement to the chloride of isoborneol, melting-point 158, which takes place on warming with alcoholic hydrochloric acid, which probably accounts for the fact that this camphene hydrochloride was discovered only very recently. 2 * K Ber. 53, 1815 (1920). 2S Ber. 53, 769 (1920). 2 Meerwein, loc. cit. 462 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Thus, when hydrogen chloride is passed into a solution of camphene in alcohol the product is mainly isobornyl chloride, but also contains more or less true camphene hydrochloride, thus accounting for the various melting-points recorded in the literature for camphene hydro- chloride, i.e., 118 to 158. This chloride melting at 158 is the iso- bornyl chloride, evidence for which is its reduction by sodium and alcohol to camphane (dihydrobornylene) and its conversion to iso- bornyl acetate by treating with silver acetate in glacial acetic acid. Isobornyl chloride is much more stable than camphene hydro- chloride and is practically not affected by alcoholic caustic alkali at ordinary temperatures. Nevertheless isobornyl chloride is consider- ably less stable than bornyl chloride (made from pinene and HC1) since by heating for one hour with alcoholic caustic alkali bornyl chloride is scarcely attacked 25 but isobornyl chloride is completely decomposed. In such a chloride mixture it is therefore possible to estimate fairly accurately the per cent of camphene hydrochloride, isobornyl and bornyl chlorides, by making use of their relative stabil- ities to caustic alkali. The facts point to reversible reactions between camphene and esters of camphene hydrate (chloride or acetate), and between the latter and esters of isoborneol. 0/\c camphene acetate of camphene hydrate isobornyl acetate Thus, camphene hydrate can be prepared from isobornyl chloride. Methyl borneol and methyl fenchyl alcohol also appear to be in equilibrium in the presence of acids since Ruzicka 26 finds that by the action of sodium acid sulfate on either of these alcohols, the same mixture of methylcamphene and methyl-a-fenchene is obtained. "Hesse, Ber. 39, 1127 (1906). **Helv. chim. Acta. 1, 110 (1918). BICYCL1C NON-BENZENOW HYDROCARBONS CH 3 CM, 463 CH;C-CH, or methylborneol methylfenchyl alcohol Although Meerwein has produced good evidence to show that in the isoborneol 5 camphene rearrangement the intermediate formation of tricyclene or a hydrocarbon containing bivalent carbon is extremely improbable, the mechanism of the rearrangement is as obscure as ever. This rearrangement is to be classed with others such as that discovered by Kishner. 27 OH CH, / CH 2 CH C< CH, CH, CH CH 3 \ / C CH 2 CH 3 and the well-known retropinacoline rearrangements; for example, the chloride CH 3 \ CH 3 CH 3 CH, C.CH 2 C1 >C CH 2 CH 3 CH - A, and the like. By the hydrogenation of camphene and bornylene the correspond- ing saturated hydrocarbons are obtained, the nomenclature of which is unfortunate. By reducing bornylene by the method of Sabatier and Senderens a "camphane" melting at 150 and boiling at 161-162 was obtained by Henderson and Pollock 28 and the same hydrocarbon in a somewhat purer form was obtained from camphor by the decom- position of the hydrazone, 29 according to the method of Kishner, the hydrocarbon made in this way having a melting-point of 156-157 and boiling at 161 (757 mm.). From its relation to camphor it is 27 Chem. Zentr. 1908 (2), 1342; 1911 (1), 543. 28 J. Chem. Soc. 97, 1620 (1910). "Wolff, Ann. 394, 86 (1912). 464 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS perhaps proper to call it camphane, which however confuses it with the isomeric hydrocarbon derived from camphene. The terms dihy- drocamphene and dihydrobornylene are much to be preferred. By heating isoborneol with zinc dust at 220 Semmler 30 obtained a hydrocarbon, C 10 H 18 , boiling at 162 and melting at 85, and Vavon 31 reduced camphene in ether solution by platinum black and hydrogen and obtained a hydrocarbon probably identical with Semm- ler 's melting at 87. Sabatier and Senderens obtained a liquid mix- ture from camphene which Henderson and Pollock 32 showed was a mixture of the camphane of Vavon and unsaturated hydrocarbons. The crude product obtained by Henderson and Pollock melted at 64 and the product obtained by Ipatiev, 33 "isocamphane," obtained by passing isoborneol over a mixture of nickel oxide and alumina at 215-220 with hydrogen at 110 atmospheres, and described as melt- ing at 63-64.5, is probably the crude camphane. A product called cy do camphane has been made from cyclocam- phanone by Kishner's hydrazine method. Cyclocamphane melts at 117-118. 34 Angeli, 35 who first prepared the ketone by the action of nitrous acid on aminocamphor, regarded it as an unsaturated ketone and accordingly called it "camphenone." The presence of the three- carbon ring in camphanone was shown by converting it (through the oxime and nitrile and oxidation) to cyclocampholenic acid and by further oxidation to cycloisocamphoronic acid. CO CH 3 CCH 3 CQH CQH CQH I CO,H cyclocamphane cyclocamphanone cyclocampholenic cycloisocamphoronic acid acid, M.-P. 228 Reduction of the ketone yields a new borneol, cyclocamphanol, melt- ing-point 174-176. Camphene, like |3-pinene, does not form a nitrosite but the nitrosite "Ber. S3, 77 Q (1900). "Compt. rend. 149, 997 (1909). * 2 J. Ohem. Soc. 107, 1620 (1910). M Ber. ft, 3205 (1912). "Holz, Z. cmgew, Chem. 27 (1), 347 (1914). **6azz. chim. Ital. 24 (2), 44, 317 (1894). BICYCLIC NON-BENZEN01D HYDROCARBONS 465 conceivably formed as a labil intermediate product, decomposes to give nitrocamphene 36 (melting-point of d and I forms 84-85, d.l.- nitrocamphene melting at 64-66). Bromine also shows a similar behavior, the group >C = CH 2 adding Br 2 to form the labil >CBr CH 2 Br which immediately decomposes to give the mono- bromo derivative >C = CHBr, camphenylidene-6-bromomethane. 37 This bromide is capable of combining with hydrogen bromide (prob- ably reversible) to form 2-bromo-Q-bromo-camphene, melting at 90-91. The corresponding camphenylidene chloride is inert to hydrogen chloride. Camphene condenses with formaldehyde (trioxymethylene) in acetic acid, to form a primary alcohol, from which a large number of derivatives have been prepared. Thus, oxidation of the new alcohol yields the corresponding aldehyde, which can then react with mag- nesium alkyl halides to give a series of diethylenic hydrocarbons of the camphenic type. Camphene combines with hypochlorous acid in cold dilute solu- tions to give a nearly quantitative yield of camphenechlorohydrin, melting at 93. Reduction of this chlorohydrin with zinc and alcohol gives isoborneol; camphenechlorohydrin is therefore probably a-chlo- roisoborneol. 38 Camphenechlorohydrin reacts with caustic alkalies or moist silver oxide to form camphenilane aldehyde, which is also obtained by treating campheneglycol (obtainable by permanganate oxidation of camphene) with dilute acids. The conversion of 1.2- glycols to aldehydes or ketones by dilute acids is quite a general reaction. The principal oxidation products of camphene are shown in the following diagram, OH -C-CO.H CH, camphene H, CM J camphene glycol M.-P. 200 CH, a-oxycamphenilanic acid, M.-P. 171 *Lipp, Ann. S99, 241 (1913). 'Langlois, Ann. chim. 12, 265 (1920). " Henderson, Heilbron & Howie, J. Ctem. Soo. 105, 1367 (1914). 466 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS n C-CHO H CH \^ r l^ CH, | camphenilane aldehyde camphenilanic acid camphenilone M.-P. 70 M.-P. 65 M.-P. 43 Reduction of camphenilone by sodium and alcohol gives camphenilol, the corresponding alcohol, C 9 H 17 OH, melting at 84. Camphor. The essential oil derived from the leaves or wood of Cinnamomum camphora 39 is a complex mixture from which camphor is more or less perfectly separated before the oil comes into commercial markets. According to Bertram and Wahlbaum 40 and Schimmel & Co. this essential oil contains, in addition to camphor, pinene, phellandrene, camphene, dipentene, dfenchene, cUimonene, bisabolene, cineol, safrol, eugenol, terpineol, citronellol, borneol and cadinene. Ordinarily a light terpene fraction is separated from commercial camphor oil as this is of little value, and the heavier oil, containing large proportions of safrol and eugenol, constitutes the chief commercial source of safrol, employed for the manufacture of piperonal. In addition to the above named constituents Semmler and Rosenberg 41 isolated a bicyclic sesquiterpene boiling at 129-133 at 8 mm.; also a monocyclic diter- pene, C 20 H 32 , which they have named ce-camphorene, and a second diterpene named (3-camphorene which is distinguished from the a-hy- drocarbon by forming a liquid hydrochloride. So-called camphoro- genol reported and very imperfectly characterized by Yoshida, 42 evi- dently does not exist. The physical properties of camphor, as recorded in the literature, are as follows, melting-point 175 , 43 176.3 to 176.5 , 44 178.4 ; 45 "Parry ("Chemistry of Essential Oils," Ed. 3, p. 160 [1918]), has called atten- tion to a bulletin issued by the Monopoly Bureau, Formosa, according to which several varieties or species (?) of camphor trees are recognized but not yet distinguished botanically, whose essential oils do not yield camphor. 40 J. prakt. Chem. (2) 49, 19 (1894). 41 Ber. 46, 768 (1913). 42 J. Chem. Soc. Jft t 782 (1885) ; Cf. Bertram & Wahlbaum, loc cit. Landolt, Ann. 189, 333 (1877) ; Beckmann, Ann. 250, 353 (1889). "Foerster, Ber. 23, 2983 (1890). "Haller, Compt. rend. 105, 229 (1887). BICYCLIC NON-BENZENOID HYDROCARBONS 467 density at 18 0.9853 ; 46 boiling-point 204 , 43 209.1 ; 44 [a] D 44.22 in 20 per cent solution in alcohol. 47 The latent heat of fusion 48 is 8.23 calories and the latent heat of vaporization is 93.4 calories. Identification of camphor is best accomplished by preparing the oxime 49 melting-point 118 to 119. As pointed out by Beckmann (loc. cit.) the oxime of d. camphor is Ia3vo-rotatory and the oxime of ^.camphor is dextro-rotatory, amounting to 41.3, in alcoholic solu- tion. The semicarbazone, melting-point 236-238, the p-bromo- phenylhydrazone 50 melting at 101, the oxymethylene derivative melting at 80-81 and the benzylidene compounds, have also been employed for the purpose of detecting or identifying camphor. The benzylidene compound of inactive, or synthetic, camphor melts at 78 but that of d. or ^.camphor melts at 95-96. Camphor forms a series of compounds with mercuric iodide, 51 C 10 H 14 0. Hg 2 I 2 ; (C 10 H 14 0) 2 Hg 4 I 2 ; (C 10 H 14 0) 4 Hg 5 I 2 and (C 10 H 14 0) 3 Hg 6 I 2 . Nitric acid forms an addition product C 10 H 16 O.HN0 3 , melting at 24, and the existence of a second compound (C 10 H 16 O) 2 HN0 3 , melting at 2.2, is indicated by the freezing-point curves. 52 Hydriodic acid forms (C 10 H 16 0) .HI, melting at 29-30, and phosphoric acid forms an addition product C 10 H 16 O.H 3 P0 4 , melting at 29. Neither ordinary camphor nor its isomer epicamphor forms a cyanohydrin. Camphor, menthone, thujone and fenchone do not react with phenyl hydrazine. The Constitution of Camphor and Its Oxidation Products. The study of the constitution of camphor and its oxidation prod- ucts and derived substances constitutes a brilliant chapter of organic chemistry, and although the structure of camphor has been known with reasonable certainty for some time, the structures of some of the derived oxidation products are still subjects of research. Since this collection of researches is such a classic and has engaged the attention of many of the ablest organic chemists, it is worth while 48 Cnautard, Jahresber. 1863, 555 (for Z.camphor) ; Malosse, Compt. rend. 154, 1697 (1912), gives d^p- 0.963. 47 Beckmann, loc. cit. "Jouniaux, Bull. soc. chim. (4) 11, 993 (1912). "Auwers, Ber. 22, 605 (1889) ; Bredt, Ann. 289, 6 (1896). M Tiemann, Ber. 28, 2191 (1895). 61 Marsh and Struthers, Proc. Chem. Soc. 24, 267 (1909). 62 Shukoff & Kasatkin, J. Russ. Phys.-Chem. Soo. 41, 157 (1909). 468 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS to review the matter somewhat more fully than some other related subjects. For the sake of clearness the historical method of review will not be followed. In the earlier work undue emphasis was put upon the fact that under certain conditions camphor could be converted (with very small yields) into para-cymene, also to carvacrol. In 1893 Bredt 53 unraveled the structure of one of the important oxidation products of camphor, i.e., camphoronic acid. On heating, camphoronic acid breaks up into carbon dioxide, isobutyric acid and trimethyl succinic acid, a change which Bredt represented as follows: CH 3 H0 2 C C CH 2 :C0 2 :H > isobutyric acid (a) CH 3 C CH 3 C(XH. (b) CH 3 CH 3 | : CH 3 -CH-C0 2 H H0 2 C C : CH 2 C0 2 H H0 2 C CH CH 3 - C - CH 8 ' ' CH 3 C CH 3 or CH, | | >C-C0 2 H C0 2 H. C0 2 H CH 3 camphoronic acid trimethylsuccinic acid A little later Bredt's constitution for camphoronic acid was conclu- sively confirmed by Perkin and Thorpe, 54 who made the acid by well- known reactions of synthesis. Bredt represented the oxidation of camphor, through camphoric acid, to camphoronic acid, as follows, C0 2 H. camphoric acid "Ann. 292, 55 (1896) ; Ber. 26, 3047 (1893). "J. Chem. 800. 11, 1175 (1897). BICYCLIC NON-BENZENOID HYDROCARBONS CH, CH 3 CH, C CO.H CH 2 C C0 2 H 469 OH camphononic acid CH 3 CH 2 C -C0 2 H CH 3 C-CH 3 -I C0 2 H camphoronic acid The hydroxy acid shown as an intermediate product in the above series of oxidations is usually obtained in the form of the lactone, camphanic acid, CH 3 f~*\ /"*1 /"\ CH C CH, , GEL The correctness of Bredt's constitution for camphoric acid is very directly shown by the synthesis of this acid, first by Komppa 55 and later by Perkin and Thorpe. 56 By condensing ethyl oxalate with pp-dimethylglutaric ester, by means of sodium ethoxide, Komppa ob- tained the diethyl ester of diketoapocamphoric acid. C0 9 R C0 2 R CO CH CO 2 R CO "Ber. 34, 2472 (1901) ; S6, 4332 (1903). M J. Chem. Soc. 89, 795 (1906). CH 3 C CH 3 CH C0 9 R C0 2 R 470 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS By the action of metallic sodium and methyl iodide a methyl group was introduced and the two ketone groups were then reduced, first to the dihydroxy acid, then by hydriodic acid and red phosphorus to the unsaturated acid dehydrocamphoric acid. The double bond in the latter acid was then reduced by adding HBr and reducing the resulting product by the well-known method of reduction by zinc dust and acetic acid, the product proving to be racemic camphoric acid. CO- -CH C0 2 R CO C C0 2 R CH 3 C CH 3 CO CH C0 9 R dihydroxy ] (dehydrocamphoric acid acid CH 3 CH 2 - -C- -C0 2 H. CH 3 C CH 3 CH 2 CH C0 2 H r camphoric acid. Perkin and Thorpe's synthesis is even more conclusive. 57 Dimethyl- cyclopentanonecarboxylic ester was treated with magnesium-methyl iodide and the resulting alcohol converted first into the corresponding bromide and the latter into the nitrile, which, on hydrolysis, yields cU.camphoric acid. "The experimental details of Komppa's synthesis were published several years later (Ann. 368, 126; 370, 209 [1909]). Blanc and J. C. Thorpe published a paper calling in question the structure of the acid obtained by methylating diketoapocam- phoric ester, claiming this to be an o-methyl ether, not the c-methyl derivative (J. Chem. Soc. 97, 836 [1910]). After an explanatory reply by Komppa (IMd., 99, 29 [1911]), Blanc and Thorpe admitted their error (iUd., 99, 2010 [1911]). Komppa's synthesis has not since been questioned. BICYCLIC NON-BENZENOID HYDROCARBONS 471 -H 2 - - CO CH 3 C - ^TT PTT C -CH 3 > .C0 2 H C 3 T5~ f CH 3 HP OH CH 3 -C- HPJT CH 3 -5 CO H CH ^TT P CH 3 ^H C V^\_/ 2 Xi PN CH 3 C CH 3 ^TT OTI pn TT r CH 3 -C- CH 3 PO P OXH. The conversion of camphoric acid to camphor had already been effected by Haller. 58 Camphoric acid forms an anhydride, which on reduction by sodium amalgam, yields campholide, CO CO C 8 H 14 < >0 > C 8 H 14 < >0 CO CH 2 When campholide is heated with potassium cyanide it yields a nitrile which on hydrolysis is converted into homocamphoric acid. CO C0 2 K C0 2 H C 8 H 14 < > + KCN - C 8 H 14 < -> C 8 H 14 < CH 2 CH 2 CN CH 2 .C0 2 H. On heating the calcium salt of homocamphoric acid, Haller obtained camphor. C0 2 C0 2 C 8 H 14 < > Ca > C 8 H 14 < | CH 2 C0 2 CH 2 Dicarboxylic acids whose carboxyl groups are separated by two or three carbon atoms readily yield anhydrides, as in the case of succinic Compt. rend. 122, 446 (1896). 472 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS and glutaric acids, and the significance of the formation of camphoric anhydride was first pointed out by Baeyer. 59 Homocamphoric acid, however, does not form an anhydride, indicating that the two carboxyl groups are separated by at least four, carbon atoms, facts which agree with Bredt's constitution and the syntheses of camphoric acid, noted above. Camphoric acid exists in a form which does not yield an anhy- dride and to distinguish this form it has been called isocamphoric acid. Like ordinary camphoric acid the iso acid exists in d. and I. and a racemic form. These facts also harmonize with Bredt's con- stitution for camphor and the chemical evidence as to the structure of camphoric acid. The four active camphoric acids may be repre- sented as follows, CH 2 C< CH 3 CO,H CH 2 C< CH 2 C< C0 2 H d. and I. camphoric acid CH, CH 2 C< CO,H CH 2 C< CH 2 C< CH 2 C< H CH 2 C< C0 2 H CH 3 *-6 C0 2 H H C0 2 H CH 3 *6 H COoH d. and I. isocamphoric acid In camphor, however, the two carbon atoms represented by the car- boxyl groups in the camphoric acids, are bound to each other and therefore there are only two active forms of camphor, i.e., d. and ^.camphor, corresponding to d. and ^.camphoric acid. In camphor the asymmetry is due to the CO group and optical activity disappears if 6 *Ann. 276, 265. Camphoric anhydride may readily be prepared by heating the acid above its melting point, or by dehydrating by means of acetyl chloride. B1CYCLIC NON-BENZENOID HYDROCARBONS 473 this ketone group is reduced to CH 2 , as was shown experimentally by Aschan. 60 Epicamphor, or (3- camphor. It will be evident from the structure of ordinary camphor that another isomeric ketone should be capable of existence, and, in accordance with the nomenclature suggested for the hydrocarbon camphane, ordinary camphor would be cc-camphor and its isomer |3-camphor. The two ketones are related structurally to each other as follows, CH 2 ordinary, or a-camphor Epicamphor, or ^-camphor In the conversion of camphor to epicamphor a reversal of the sign of optical rotation is observed, which may be summarized thus, d-camphor, [a] + 39.1 5 ^-epicamphor, [a]-Q 58.2. CO Hydroxymethylene epicamphor, C 8 H 14 < | , like ordinary C = CHOH hydroxymethylene camphor, is formed when Z.epicamphor is treated with isoamyl formate and sodium in the presence of ether. 61 It ex- hibits muta -rotation, increasing on standing, particularly in the presence of sodium ethylate; freshly prepared material showed [a] - 125.5 and after adding a trace of sodium ethylate the rota- D tion increased to [a] 146.7. The decomposition of Z.epicam- D phoroxime by dilute sulfuric acid proceeds in a similar manner to that of camphoroxime, forming epicampholenonitrile, with rupture of the ring as in ordinary camphoroxime. The nitrile may be hydrolyzed to l.a-epicampholenic acid, but the behavior of this acid differs from ordinary a-campholenic acid in not rearranging to an isomeric acid corresponding to (3-campholenic acid. An interesting attempt to pre- 80 Ann. 316, 229 (1901). 61 Perkin & Titley, J. Ctiem. Soc. 119, 1090 (1921). 474 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS pare p-camphor was made by Haller and Blanc. 62 Campholide, made by reduction of camphoric anhydride, yields homocamphoric acid and camphor as follows, CO C0 2 K C0 2 H CO C 8 H 14 < > -> C 8 H 14 < -> C 8 H 14 < -> C 8 H 14 < | CH 2 CH 2 CN CH 2 C0 2 H CH 2 Haller and Blanc prepared (3-campholide but this substance did not react with potassium cyanide. Wagner 63 believed that he had suc- ceeded in preparing p-borneol by applying the Bertram-Wahlbaum reaction to bornylene, but it was later shown that his bornylene must have been very impure and his results are considerably variant from those of Perkin and Bredt. Wagner's method was carefully tested by Bredt and Hilbing, 64 who employed a very pure bornylene and a mix- ture of borneols was obtained which they were unable to separate and on oxidation, ordinary camphor only could be identified. Epi- camphor was first described in a preliminary paper by Perkin and Lankshear 65 and almost simultaneously by Bredt. 66 However much the best method of preparation was worked out by Perkin and Bredt jointly, their method consisting in treating methyl d-bornylene-3- carboxylate 67 with hydroxylamine in the presence of sodium meth- oxide. On heating, the product decomposes forming epicamphor, the reactions involved probably being as follows, C.C(OH) z=N.OH C-N:C = C.NH.CO,H C 8 H 14 < || C 8 H 14 < || -- > C 8 H 14 < CH CH CH bornylene-3-hydroxamic acid C NH 2 C C 8 H 14 < || - > C 8 H 14 < | > C 8 H 14 < | CH CH 2 CH 2 epicamphor Epicamphor has an odor similar to that of ordinary camphor, it melts at 182, boils at 213; its oxime melts at 103-104 and the * 2 Compt. rend. IJfl, 697 (1905). 83 Her. 36, 4602 (1903). M J. prakt. Chem. (2) 84, 783 (1911). 68 Proc. Chem. Soc. 27, 167 (1911). M Chem. Ztg. 35, 765 (1911). "Bredt, Ann. 348, 200 (1906) ; 366, 1 (1909) ; Bredt & Perkin, J. Chem. Soc. 103, 2182 (1913) ; Furness & Perkin, J. Chem. Soc. 105, 2025 (1914). BICYCLIC NON-BENZENOID HYDROCARBONS 475 semicarbazone melts at 237-238. Sodium and alcohol reduce epi- camphor to the corresponding epiborneol, melting-point 181-182.5. Like ordinary camphor, the new ketone does not react with hydrogen cyanide and is not reduced by zinc dust in acetic acid. The chemi- cal properties of epicamphor do not differ markedly from ordinary camphor but "favorable action of epicamphor on the beat of the heart does not become apparent until the solution administered is about four times stronger than that which produces the same effect in the case of camphor." In connection with the discussion of the constitution of camphor and camphoric acid it will be convenient to review briefly several related derivatives. The nomenclature in this series of acids has been very much confused and the molecular rearrangements which some of them undergo made the determination of their constitution a matter of considerable difficulty. An extension of our knowledge of the pinacone-pinacoline rearrangement has assisted materially in clearing up the relationships of this group of substances. Bredt 68 has reviewed the nomenclature of the camphonene and laurolene series and suggests abolishing the designations "lauronolic acid" and "campholactone." Two series of unibasic unsaturated acids are known which are derived from camphoric acid. To one series belong camphonenic acid and lauronolic acid and since the latter acid is unsaturated it may more appropriately be called laurolenic acid. Both of these acids contain a carbonyl group which is attached to the tertiary carbon atom of camphoric acid. In the other series the carbonyl group is attached to the secondary position and includes campholytic and isocampholytic acids (|3-campholytic acid). It is now known that the substances formerly known as lauronolic and iso- lauronolic acids, bihydrolaurolactone and isobihydrolaurolactone are not merely differentiated by the different positions of the double bond, as was formerly considered to be the case, but possess different carbon structures since camphonenic acid (below) (iso or y-lauronolic acid) still contains the 0era.dimethyl group of camphoric acid; lauro- lenic acid (formerly lauronolic acid) does not possess this group. For the purpose of a key for reference, the revised nomenclature suggested by Bredt is given, as follows. n J. prakt. Chem. (2) 87, 1 (1913). 476 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS I. Camphonene and Camphonone Series. CH 3 CH 3 CH 2 - -C- -C0 2 H. CH 2 - -C- -C0 2 H CH 3 C CH 3 CH 3 C CH 3 5H 2 =r=z= C C0 2 H. CH 2 =i==r CH dehydrocamphoric acid camphonenic acid. CH 3 CH 3 CH 2 C C0 2 H . CH 2 C C0,li CH 3 C CH 3 CH 3 C CH 3 . CH 2 CH 2 CH.NH 2 camphonanic acid. aminocamphonanic acid CH 3 CH 3 CH 2 - -C- C0 2 H CH 2 - -C- -CO CH 3 C CH 3 | CH 3 C CH 3 , CH.OH CH 2 CH^ camphonolic acid camphonololactone CH 3 CH 2 C C0 2 H. CH 3 C CH 3 camphononic acid. II. Laurolene and Laurolane Series. CH 3 CH 3 CH 2 - -CH CH 2 - -C C0 2 H C CH 8 C CH 3 C CH 3 CHo C CH, L 2 laurolene laurolenic acid BICYCLIC NON-BENZENOID HYDROCARBONS 477 CH 3 CH 3 CH 2 - -C- -C0 2 H CH 2 - -C- -C0 2 H H C CH 3 CH.CH 3 CH 2 CH CH 3 CH 2 C CH, laurolanic acid CH, laurololic acid CH 3 CH, C CO CH.CEL ^b laurololactone Some of the synonyms of the above terms are as follows: Synonym Bredt's nomenclature Camphonolic acid Laurololic acid Camphonololactone Laurolanic acid Laurololactone Laurolenic acid Camphonenic acid Hydroxylauronic acid Hydroxy acid of campholactone 69 Isocampholactone 70 Dihydrolauronolic acid fCampholactone iBihydrolaurolactone Lauronolic acid y-lauronolic acid Bredt 71 has also suggested that substituents in the single methyl group of camphoric acid be designated as Q. Bredt follows Kipping's pro- posal that substituents in the 0era.dimethyl group be designated by the letter it. Thus the four known monobromocamphoric acids are, Noyes, J. Am. Chem. Soc. 34, 182 (1912). Noyes. J. Am. Chem. Soc. Sl t 278 (1909). "Ann. 335, 26 (1913). 478 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CH 3 C(XH C0 9 H H CH C CH >C -CO a H Br S-Bromocamphoric acid CH 3 C H 9 C C0 2 H 4-Bromocamphoric acid CH 2 Br CH 2 C C0 2 H CH 3 C CH 2 Br C C0 2 H H n-Bromocamphoric acid 72 CH, CH 3 C CH 3 C H Q-Bromocamphoric acid C0 9 H From chlorocamphoric phenyl ester by heating with quinoline, saponifying the resulting unsaturated ester and heating the free acid, Bredt 74 obtained an acid whose structure he represented as follows, CH 3 CH, CH, CH 3 C -i- -A- C0 9 R CH, C0 2 R CH, \ C0 2 R CH 3 C CH 3 CH C0 2 R Cl CH ===CH camphonenic acid "Kipping, J. Chem. 8oc. 69, 918 (1896). "Armstrong & Lowry, J. Chem. Soc. 81, 1467 (1902). 7 *5er. 35, 1286 (1902). This acid was originally termed "lauronolic acid" by Bredt, but later changed to camphonenic acid to avoid confusion with Fittig and Woringer's lauronolic acid. Cf. W. A. Noyes & Burke, J. Am. Chem. Soc. 3li, 177 (1912). In a later paper by Bredt [Ann. 895, 26 (1913)] d.dehydrocarnphoric acid, melting-point 202-203, and tU.dehydrocamphoric acid, melting-point 228, is described. BICYCLIC NON-BENZENOID HYDROCARBONS 479 Bredt arriveji at this structure from the fact that the acid gives camphoronic acid on oxidation. CH 3 CH 3 CH 2 - -C- -C0 2 H CH 2 - -C- -C0 2 H CH CH C0 2 H C0 2 H Camphoronic acid Camphononic acid is one of the important members of this series. On further oxidation it yields camphoronic acid 75 and its constitution is as shown in the following, CH 3 CH 3 C C0 9 H CH 2 - -C C0 2 H CH 3 C CH 3 :0 2 H C0 2 H camphononic acid By oxidizing dibromocamphor by dilute nitric acid and silver nitrate Lapworth and Chapman 75 obtained homocamphoronic acid, whose anhydride loses C0 2 yielding camphononic acid. CH a CH 3 C0 2 H CH 2 - -C- -C0 2 H CH 3 C CH 3 +C0 2 +H 2 COOH OH Lapworth and Lenton 76 also made camphononic acid in two other ways which also indicate the constitution shown. The amide of cam- phanic acid was converted by dehydration to the nitrile and this, on treating with concentrated alkali loses HCN to give camphononic acid. Their second method also starts with camphanic acid amide; by treat- ing with bromine and caustic soda the CONH 2 group is replaced by "Lapworth and Chapman, J. Chem. Soc. 75. 986 (1899). n J. Chem. Soc. 79, 1284 (1901). 480 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS NH 2 and the resulting product decomposes yielding ammonia and camphononic acid. CH 3 CH 3 C CO CH 2 C \ CO a K CH q C CH 3 O -*nitrile CH, C CH Q C CH CONH, CH C< OH CN CH 3 CH 2 C C0 2 H CH 3 C CH 3 camphanic acid amide * CH 3 CH 2 C- -C0 2 H CH 3 C CH 3 i< H NH 2 When camphononic acid is reduced electrolytically the ketone group is reduced without structural change, to give camphonolic acid or its lactone. 77 CH 3 CH 3 , C COCH CH 2 C COOH CH 3 C CH 3 -* | CH 3 C CH 3 > CH 2 C==0 CH 2 CH.OH camphonolic acid cis. trans. M.-P. 2^9 CH 3 C CH. CH, CH 2 CO CH, C CH 3 / H lactone M.-P. 160 "Bredt, J. praU. Chem. (2) 84, 786 (1911) ; Ann. 366, I (1909). BICYCLIC NON-BENZENOID HYDROCARBONS 481 It will have been noted that the above substances retain the carbon structure of camphoric acid. Fittig and Woringer 78 had obtained an acid by the decomposition of bromocamphoric anhydride which they had termed lauronolic acid but since it did not give the oxidation products described above many chemists refused to accept Bredt's pro- posed constitution of camphoric acid. Fittig and Woringer's lauro- nolic acid does not give camphoronic acid on oxidation. What ap- pears to be the correct explanation of the structure of this acid was given by Lapworth and Lenton 79 and also confirmed by other evi- dence. In the preparation of Fittig and Woringer's lauronolic acid by the decomposition of camphanic acid, Lapworth and Lenton assume a structural rearrangement similar to the change of position of a methyl group in the pinacone pinacoline rearrangement. Accord- ingly, Fittig and Woringer's lauronolic acid has the structure shown in the following. CH 3 CH 3 C CO CH 9 C CH, -CO \ CO CH 3 C CH 3 \ CH C C0H camphanic acid CH 3 CH 2 - -C- -C0 2 H CH, C H Laurolenic acid (Bredt). (lauronolic acid) \ CH 3 This lactone proves, in fact, to be different from the lactone of camphonolic acid later made by Bredt. In the light of the fore- going, other facts become clear, for example the oxidation of lauro- nolic acid by potassium permanganate 80 to laurenone. "Ann. 227, 6 (1885). " Loc. cit. Tiemann & Tigges, Ber. S3, 2950 (1900). 482 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CH 2 CO CH 3 CH 5 or GIL CH, CHu AH C-CIL CO -i H Also the nitro derivative and its reaction products, obtained by Schryver 81 are, according to Bredt, also in harmony with the other known facts, the nitro group displacing the tertiary hydrogen, as is customary, Schryver's nitro derivative probably having the struc- ture, CH 8 CH 2 - Oxidation to camphoronic acid is therefore the criterion as to whether or not the C (CH 3 ) 2 group of camphoric acid is retained, and as sur- mised by Lapworth and Lenton, Fittig and Woringer's lauronolic acid and its derived lactone (bihydrolaurolactone) and the hydrocarbon laurolene, do not possess this structure. A similar rearrangement of a methyl group occurs in the conver- sion of campholytic acid to isocampholytic acid. 82 . J. Chem. Soc. 73, 559 (1898). M This name has been agreed to by Noyes, Perkin, Aschan and Bredt to replace the various other names by which it has been known, e. g., isolaurolonic, camphothetic, and ^-campholytic acid. BICYCLIC NON-BENZENOID HYDROCARBONS 483 CH 3 CH 3 CH _ C CH 2 ; C CH 3 CH 3 -C-CH 3 I C-CH 3 CH 2 - -CH- -C0 2 H CH 2 C C0 2 H. campholytic acid. isocampholytic acid. In connection with the relationships just discussed it will be con- venient to mention the work indicating the structure of laurolene and isolaurolene C 8 H 14 . Isolaurolene has been obtained by heating copper camphorate and also from isocampholytic acid ((3-campholytic acid). Its structure has been determined by Blanc 83 by a study of its oxida- tion products and confirmed by it* synthesis, to be 1 . 1 . 2-trimethyl- A 2 -cyclopentene. isolaurolene CH 2 Noyes and Derick 84 prepared laurolene by treating aminolauronic hydrochloride with sodium nitrite, also by boiling the nitroso deriva- tive of aminolauronic anhydride with caustic soda. They found experimental evidence which they considered as supporting the struc- ture which Eijkmann 85 had proposed on the ground of refractometric considerations. Noyes and Kyriakides 86 made the hydrocarbon by simple methods of synthesis which confirm the structure proposed by Eijkmann, i.e., CH, laurolene M BuU. sac. chim. (3) 19, 703. "J. Am. Chem. 8oc. SI, 669 (1909). M Chem. Zentr. 1907, II, 1208. "J. Am. Chem. Soc. 52, 1064 (1910). 484 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The physical properties of the laurolenes are as follows: dlaurolene, [a] 26.2^ , 2815 o. d ^ = 0.8030; boiling-point 120.3-121. Iso- D 4 laurolene boils at 108. Derivatives of Camphor. Camphor Oxime 87 is readily made in good yields by adding the calculated quantity of concentrated caustic soda to an alcoholic solu- tion of camphor and hydroxylamine hydrochloride and heating on a water bath for about one hour. The oxime crystallizes well from dilute alcohol, melting-point 120. It boils at 249-250 with very slight decomposition. Camphor oxime is reduced by hydrogen and catalytic nickel at 180-200 to bornylamine, dibornylamine and camphylamine, the second being the principal product. 88 C = N.OH Isonitrosocamphor, C 8 H 14 <| exists in two forms, melt- CO ing at 114 and 153 . 89 For its preparation 90 102 grams of camphor are dissolved in 500 cc. of dry ether and 15.2 grams of sodium as sodium wire, are added. Amyl nitrite is then added in small portions with cooling, until 78 grams have been added. After standing about three hours, add cracked ice and ice water; the sodium salt of isonitroso camphor is in the aqueous phase and unchanged camphor and borneol in the ether. Acetic acid precipitates the free isonitroso derivative, which after recrystallizing from dilute methyl alcohol, or petroleum ether, melts at 152-154. Zinc and dilute acids readily reduce it to amido camphor CH.NH 2 C 8 H 14 <| boiling-point 244, whose hydrochloride has a physiological action similar to curare, but feebler. Two molecules of amidocamphor may be condensed to dihydrocamphenepyrazine. 91 Amidocamphor and potassium cyanate yield camphorylcarbamide CH.NHCONH 2 C 8 H 14 <| which may be converted to camphoryliso- cyanate by the action of nitrous acid. Like the well-known reagent, "Auwers, Ber. 22, 605 (1889). "Aloy & Brustier, Bull. soc. chim. (4) 9, 733 (1911). "Chem. Zentr. 1908, I, 1270. Claisen & Manasse, Ann. 271, 73 (1893). Ann. 313, 25 (1900). BICYCL1C NON-BENZENOID HYDROCARBONS 485 phenylisocyanate, the camphoryl derivative is very reactive and a large number of camphorylurethanes and other derivatives have been prepared from it. 92 When amidocamphor hydrochloride is treated C=N a with nitrous acid, azocamphor, C 8 H 14 <| is produced. 93 CO CO Camphor Quinone, C 8 H 14 <| . When isonitrosocamphor is heated CO with dilute sulfuric acid, the diketone is formed, as in the hydrolytic decomposition of oximes, C = N.OH CO (a) C 8 H 14 < | +H,0-^C 8 H 14 < | +H 2 N.OH CO CO C = N.OH CO (b) C 8 H 14 < | +NO.OH -> C 8 H 14 < | +N 2 + H,0 CO Nitrous acid also converts isonitrosocamphor to camphorquinone. 94 About 9 parts of camphor are dissolved in 15 parts acetic acid and 4 parts sodium nitrite (dissolved in minimum of water) are carefully added. After completion of the reaction the diketone is precipitated by diluting with cold water. The diketone is easily volatile, crystal- lizes well in yellow needles melting at 198, and is markedly soluble in hot water. The effect of the two contiguous CO groups upon the stability of the ring is noteworthy, the diketone being easily converted to cam- phoric acid or its derivatives under the influence of a wide variety of reagents. 95 The dioxime, C 8 H 14 <| is best made by the C = N.OH action of hydroxylamine on isonitroso-camphor. All of the eight pos- sible oximino derivatives of camphorquinone are known. The dis- covery of the two modifications of isonitrosoepzcamphor, constituting the third and fourth monoximes of camphor quinone, completes the list of theoretically possible oximes and Forster 96 has shown the prob- able configuration of these derivatives. Their physical properties are as follows, "Chem. Zentr. 1908, I, 257. "Angeli, Ber. 26, 1718 (1893). *Claisen & Manasse, Ann. 27-k 83 (1893) ; Lapworth, J. CTiem. Soc. 69, 322 (1896) ; Bredt, Rochussen & Monheim, Ann. S14, 388 (1900). 88 Aschan, Ber. SO, 657, 659 (1897). * J. Chem. Soc. 103, 662 (1913). 486 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS Melting- Point 114 in Chloroform 172.9 in 2% NaOH 275.3 Isonitrosocamphor (stable) 152 197.0 288.0 Isonitrosoepicamphor (unstable) 137 -179.4 -278.5 Isonitrosoepicaniphor (stable) . . .. 170 -200.1 -422.0 CainpliorQuiiioDe (x~dioxiiii6 201 -51.7 -103.8 CamphorQuinonG 3-dioxim6 248 -24.5 136 16.4 14.3 Camohorauinone. fi-dioxime . 194 52.8 87.0 Reduction of the diketone by zinc dust and acetic acid gives a-oxy- CH.OH camphor C 8 H 14 <| melting-point 203-205. Sodium and alco- CO CH.OH hoi causes further reduction to camphorglycol, C 8 H 14 <| CH.OH melting at 231. This glycol may be regarded as bornyleneglycol and is not identical with the glycol of camphene. Camphor-quinone undergoes condensation with nitromethane very readily and nitrome- thylenecamphor and the intermediate product, nitromethylhydroxy- camphor, have been isolated. 97 CO C 8 H 14 < | CO C(OH).CH 2 N0 2 C 8 H 14 < | CO C 8 H 14 < Condensation with ethyl cyanoacetate also readily takes place yield- C0 2 Eth ing ethyl camphorylidene-cyanoacetate, C 8 H 14 < | CO CN from which the corresponding camphorylidenemalonic acid was ob- tained. Para-diketocamphane: When a mixture of bornyl and isobornyl acetates are oxidized, in glacial acetic acid, by chromic acid, an acetoxy camphor is produced, 98 and since pure .isobornyl acetate does not give this result, this derivative must be a product resulting from 'Forster & Withers, J. Chem. Soc. 101, 1328 (1912). "SchrOtter, Monatsh. 1881, 224. BICYCLIC NON-BENZENOID HYDROCARBONS 487 the oxidation of bornyl acetate. Hydrolysis of the acetate gives hydroxy camphor, melting at 238-246, this product really consisting of two stereoisomerides. Oxidation of hydroxycamphor by chromic acid gives para-diketocamphane, melting-point 206.5-207 and 13 5 Bredt " finds that the diketone is optically active, [a] _ + 103.42. Since this substance is not identical with camphorquinone and since a substance in which both CO groups were attached to the same bridge carbon atom would be optically inactive, Bredt concludes that the constitution of the two substances are as indicated below. = CH 2 - C - C = H Para- dik e tocamphane The relative ease with which camphor reacts with metallic sodium or sodium amide to form a sodium derivative, has been made use of extensively for the preparation of other derivatives. Thus, sodium camphor in benzene solution reacts with C0 2 to give d.camphocarbonic CH.C0 2 H acid, C 8 H 14 <| , melting at 128. A recent synthesis of cam- phocarbonic acid by Ruzicka 10 is worth noting since a well-known reaction was successfully applied to this synthesis by the simple expedient of employing an autoclave to obtain a temperature of 200. The diethyl ester of homocamphoric acid was condensed by sodium ethylate in alcohol at 200. CO^ CO C 8 H 14 < ' C 8 H 14 CH 2 C0R CH.C0R | CH. Camphocarbonic acid has been employed for the preparation of pure bornylene. The a-hydrogen atom in camphocarbonic esters is readily displaced by sodium, by which means alkylation is easily effected; alkyl halides give C-derivatives but acid chlorides give o-acylated W J. prakt. Chem. (2) 101, 273 (1920). 100 Helv. Chim. Acta. S, 748 (1920). 488 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS C.C0 2 R CH.CN products, C 8 H 14 < 1 1 . The closely related nitrile C 8 H 14 <| C.OAc CO also forms a monosodium derivative which, by the action of alkyl iodides, yields a mixture of and C alkyl derivatives. Camphocarbonic acid and its alkyl derivatives readily decompose on heating, a molecule of carbon dioxide and camphor or a derivative CHR C 8 H 14 <| being formed. The introduction of a methyl group in CO this case has a very marked effect upon the melting-point, methyl- CH.CH 3 camphor C 8 H 14 <| melting at 38; ethylcamphor and dime- thylcamphor are liquids, their odor being suggestive of menthone rather than camphor. C = CH.OH Oxymethylenecamphor, C 8 H 14 <| is of interest on C = account of its strongly acidic character. It is prepared by the action of methyl or ethyl formate on sodium camphor or magnesium-camphpr bromide, or by the action of sodium methoxide on a-mono-halogen or a-dihalogen camphor. 101 As in similar "formyl" derivatives some reactions indicate the structure indicated above and other reactions CH.CHO point to the desmotropic form, C 8 H 14 <| '. It readily forms an acetate and a series of ethers; it combines with nascent hydro- cyanic acid to give a cyanhydrine and is reduced by sodium and CH.CH 2 OH alcohol to camphylglycol, C 8 H 14 <| , which is known in two CHOH forms, cis melting at 87 and trans melting at 118. The trans- glycol is oxidized by potassium permanganate to rans-borneolcarbonic CH.C0 2 H acid C 8 H 14 <| but cis-borneolcarbonic acid is unstable and CHOH oxidation in this case proceeds to camphoric acid. 102 CH.CH 2 OH Reduction of camphylcarbinol C 8 H 14 <| by sodium /^TT <^i 2 101 Brtihl, Ber. 37, 2069 (1904). 1M Bredt, Ann. S66, 62 (1909). BICYCLIC NON-BENZENOID HYDROCARBONS 489 in moist benzene or condensation of camphylbromomethane by CHCH 2 CH 2 CH sodium, 103 gives dicamphylethane, C 8 H 14 <| >C 8 H 14 , CH 2 CH 2 melting-point 209-211. Reduction of camphor by passing over catalytic nickel and alumina at 200 gives isocamphane, 104 melting-point 64.5, boil- ing-point 164-165. Camphor condenses with oxalic ester, 105 under the influence of sodium ethylate, to camphoroxalic acid CH.COC0 2 H C 8 H 14 <| melting-point 88, which has yielded a series CO of derivatives. The above reactions will serve to make clear the very marked reactivity of the CH 2 group contiguous to the ketone group in camphor. ^ Camphoric Acid has been discussed above on account of the im- portance of its constitution to that of camphor itself. Its preparation is not difficult and the original method of Wreden 106 gives quite satis- factory yields. To 300 grams of camphor 3 liters of nitric acid, Sp. Gr. 1.27, are added and the mixture warmed on a water bath for several days. When cold the crude crystals are taken up in about 1 liter of water and milk of lime made from 50 grams of lime are added, which forms the freely soluble acid salt. The soluble salt is separated from unchanged camphor and the camphoric acid is then precipitated with more milk of lime as the sparingly soluble neutral salt, from which the acid may be liberated by hydrochloric acid; yield about 250 grams, melting-point 178. It is soluble in 160 parts of water at 12 but is soluble in 10 to 12 parts of water at 100. In- j active (d.l.) or para-camphoric acid melts at 204. On heating calcium camphorate the expected ketone formation \ takes place but the bridged ring is also broken, the constitution of the resulting product, camphorphorone, having been shown by a study of its oxidation products and its synthesis from 2-methyl cyclopen- tanone and acetone (by condensing by sodium ethoxide) and the hydrolytic decomposition of camphorone by caustic alkali to this ketone and acetone. 107 The latter reaction will recall the similar 103 Rupe & Ackermann, Helv. Chim. Act'a. 2, 221 (1919). 10 *Ipatiev, Ber. tf, 3205 (1912). ios Tingle, J. Am. Cnem. Soc. 2J, 363 (1901) : 29. 277 (1907), 106 Ann. 163, 323 (1872). ""Wallach, Ann. 331, 322 (1904), 490 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS behavior of citral, pulegone and other substances which possess the group (CH 8 ) 2 C = c ^TT ( ;H S CO ? H 'H CH C ^ CH 3 ( 1 ^H C }_CH 3 Ca - H CO -> CO CH C C ~^H C % ^ CH - CH 3 C CH 3 CH 3 c- H ^TT C, CO 5 CH CO+(CH 3 ) 2 CO ..ii H, CH 3 C CH 3 When camphorphorone is reduced by hydrogen over catalytic nickel 108 at 130 the saturated ketone, dihydrocamphorphorone, is produced, which on reduction at a higher temperature, 280, is further reduced to l-methyl-3-isopropylcyclopentane (boiling-point 132- 134). CH, CH C 3 H 7 Camphorimide, melting-point 243, is readily made when dry am- monia is passed into boiling camphoric acid. 109 Camphor reacts normally with magnesium-allyl bromide 110 to C(OH).CH 2 CH = CH 2 give allyl borneol, C 8 H 14 <| which on oxida- CH 2 : 8 Godchot & Taboury, Bull. soc. cMm. (4) IS, 599 (1913). 109 Evans, J. Chem. Soc. 97, 2237 (1910). ll Khoin, J. Ruaa. Phys.-Chem. Soc. kk, 1844 (1912). BICYCLIC NON-BENZENOID HYDROCARBONS 491 C(OH).CH 2 C0 2 H. tion by permanganate yields the acid C 8 H 14 <| CH 2 Haller has shown that by the action of alkyl halides and sodium amide, camphor may be alkylated, the substitution in such ketones replacing one or more hydrogen atoms adj acent to the carbonyl group. Haller 111 has thus prepared dimethyl, methylethyl, propyl, dipropyl, benzyl, dibenzyl and ethylbenzyl camphors. By reduction of these ketones the diethyl, methylethyl, propyl and dibenzylborneols were obtained. The quantitative determination of camphor in commercial prod- ucts such as celluloid or spirits of camphor is somewhat difficult on account of its volatile character. It can be precipitated from alco- holic solutions by concentrated aqueous calcium chloride, 112 taken up in light petroleum ether and finally determined gravimetrically. In the case of celluloid, distillation of the finely rasped product with steam gives fairly satisfactory results, 113 if no camphor substitutes are present. Extraction of rasped celluloid for 10 hours with petroleum ether also gives good results. The use of an immersion refractometer on solutions of camphor in methyl alcohol has also been employed. 11 * Homocamphor: This ketone closely resembles ordinary camphor in its physical properties and chemical reactions. It is a white crys- talline substance melting at 189-190, sublimes easily and has an odor closely resembling ordinary camphor. It has one more CH 2 group, in the ring containing the CO group, than ordinary camphor. It has recently been made 115 from camphoric acid anhydride by con- densing with diethyl sodio-malonate, reducing the product thus ob- tained and on distilling the resulting acid hydrocamphorylmalonic acid is obtained, CO C = C(C0 2 C 2 H 5 ) 2 C 8 H 14 < > > C 8 H 14 < > CO CO C0 2 H CH 2 CH< CH 2 .CH 2 OXH C.H 14 < C0 2 H C 8 H 14 < > C0 2 H C0 2 H hydrocamphorylacetic acid. " Haller & Bauer, Compt. rend. 158, 754 (1914) ; Haller & Louvrier, Compt. rend. 148f 1643 (1909) . na Penniman & Randall, J. Jnd. d Eng. Chem. 6, 926. 118 Barthelemy, Kunstoffe, 3, 46 (1913). >Utz, Chem. Abs. I, 1467 (1907) ; Arnost, Z. Nahr. Oenussm. 12, 532. "Lapworth & Royle, J. Chem. Soc. 117, 744 (1920). 492 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS CH 2 CH 2 > C 8 H 14 < / CO Heating the lead salt of the last named acid or prolonged heating with acetic anhydride yields homocamphor. Synthetic Camphor The economic balance between the cost of manufacturing camphor synthetically and obtaining it from natural sources is at present com- paratively even, and success in the manufacture of synthetic camphor is very largely determined by factors over which the manufacturing chemist has no control. The development of this industry was coin- cident with the very great rise in price of natural camphor after Japan had succeeded in practically monopolizing the production, of; natural camphor. The large tree Cinnamomum camphora is the only commercial source of natural camphor and the production of camphor, by steam distilling the chipped wood of large mature trees of sixty years or more in age, has been carried out in China and Japan for several hundred years. True camphor was known in Europe at least as early as 1583, but owing to the custom of cutting down the trees for distillation of the wood, together with the increased demands for camphor resulting from the development of the celluloid industry, large mature trees became more and more scarce in Japan and North Central China, 116 with the result that in 1903 the industry in Japan was made a Government monopoly. Camphor had been produced in Formosa, coming into the market prior to 1895 as Chinese camphor, but after this island was acquired by Japan, camphor production was vigorously pushed by the Japanese and Formosa soon became the principal producing locality. With the elucidation of the constitution of camphor and related substances, its artificial production was prac- tically certain to be undertaken, but this was greatly stimulated by the attempted price manipulation of the Japanese monopoly. Cellu- loid, the manufacture of which was first developed by John W. Hyatt of Newark, N. J., is the chief industrial use of camphor, this industry consuming seventy to eighty per cent of the world's total camphor production. The recent rapid development of the moving picture industry has added to the consumption of camphor for the manu- facture of films and a further consumption has been brought about " Foochow was formerly the center of the Chinese camphor market. During the period of high prices in 1919 about 930,000 Ib. of camphor were shipped from Foochow. B1CYCL1C NON-BENZEN01D HYDROCARBONS 493 by the manufacture of transparent films and sheets for automobile curtains. Camphor was used at one time in the manufacture of smokeless powder and it is still so used to a limited extent in some sporting powders. The United States imports the largest share of the annual production of natural camphor but with the development of the celluloid industry by the Japanese, the Japanese Monopoly Board has seen fit to allot certain proportions of the output to the various consuming countries, a situation which is having the natural result of stimulating the production of natural camphor in the United States. All of the successful processes for the manufacture of artificial camphor employ pinene or turpentine as a raw material, and while the primeval camphor forests in Formosa and the interior of China are being rapidly destroyed, the manufacture of artificial camphor is dependent upon a raw material the supply of which is likewise rapidly diminishing with the destruction of the American turpentine forests. Turpentine is the largest item of cost in the manufacture of artificial camphor. However, the use of light petroleum fractions and other turpentine substitutes, particularly in the paint and varnish industry, should enable scientific forestry to keep pace with the consumption of turpentine in those industries in which it is indispensable. Another factor in the situation is the planting of camphor trees and the distillation of camphor from the twigs and leaves. (The cam- phor tree does not exude an oleoresin which can be collected and sepa- rately distilled, as in the case of turpentine.) The cultivation of camphor trees and distillation of the leaves has been carried out ex- perimentally in numerous subtropical localities, the Bahamas, 117 Florida, Ceylon, Java and Formosa, and according to a recent Bul- letin of the Imperial Institute (1920) the Japanese Monopoly Board are stated to have planted 3,000,000 trees between the years 1900 and 1906 and 11,000,000 trees in the three years following. In 1913 the Board adopted the plan of planting 3,000 acres annually in camphor trees. However, leaf distillation has not proven economical and the total production of Formosan camphor has declined steadily since 1916. In 1919 the Japanese Monopoly Board estimated that the production of Formosan camphor from old trees would average about 6,500,000 Ib. annually and that the trees set out previously, as before mentioned, would be ready for working about 1930. In the United 117 Emerson & Weidlein, J. Ind. & Eng. Chem. k, 33 (1912) ; Eaton, U. 8. Devi Agric. Bull. 15 (1912) ; Beille & Lemaire, Butt, de Pharmaeie Bordeaux 191Z, 521. 194 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS States one celluloid company has about 3,000 acres planted in cam- phor trees near Satsuma, Florida, and another company is reported to have about 12,000 acres planted in camphor near Waller, Florida. The United States Department of Agriculture has a station at Orange City, California, engaged in the study of camphor cultivation. The cultivation of camphor and distillation of the leaves has also been studied in the Federated Malay States and at the Hakgala Gardens in Ceylon. In the former tests the yield of crude camphor varied from 1.1 to 2.6 per cent; a yield of about 180 pounds per acre per year was estimated. Old wood of mature trees yields on an average about 4 per cent of crude camphor oil, and air dry leaves of cultivated trees average 1.5 to 2 per cent of camphor oil (75 per cent of which is camphor). R. T. Baker 118 has reported a high yield of camphor from the -Australian species Cinnamomum olivieri and C. laubatii. Formerly, crystalline bornyl chloride (usually miscalled pinene hydrochloride) was manufactured and sold under the name of "arti- ficial camphor." This product has been known for more than a cen- tury but its usefulness in the manufacture of true camphor was not appreciated until the development, since 1906, of the processes now employed in the making of synthetic camphor. Bornyl chloride can- not be substituted for camphor in the manufacture of celluloid and it contains unstable hydrochlorides, which liberate free hydrochloric acid. It is no longer a common commercial article. All known processes for the industrial manufacture of synthetic camphor involve the oxidation of borneol or isoborneol. Borneol occurs in nature as "Borneo camphor" in the wood of one of the Dipterocarpacece and in a large number of essential oils, including most of the pine needle and cedar leaf oils, ginger oil, et cetera, but from none of these natural sources can it be produced cheaply or in quantity. The borneols are obtained industrially from bornyl chlo" ride, and this explains the use of pinene or turpentine as a raw material. No other material is known from which the borneols 01 camphor can be manufactured cheaply and in quantity. Turpentines suitable for the production of bornyl chloride and synthetic camphor are derived mostly from the long leaf pine, Pinus palustris, of the southern United States, the Cuban pine, Pinus hetero- phylla, and the Pinus pinaster of France. The turpentines from these species consist almost exclusively of a and (3-pinenes. Small propor- tions of limonene and phellandrene might occasionally be found in "Schimmel & Co. Semi-Ann. Rep. 1911 (1), 38. BICYCLIC NON-BENZENOID HYDROCARBONS 495 American turpentine since, as Herty and Dickson 119 have shown Pinus serotina yields a so-called turpentine consisting chiefly of limo- nene, but these trees are scattered and relatively unimportant. Syl- vestrene, one of the principal constituents of Russian and Finnish turpentine, has never been found in the oil from American species. The various "process turpentines," made by solvent extraction of pine wood or from stumps, is not suitable for the manufacture of bornyl chloride since such turpentines commonly contain liberal proportions of the solvent employed for its extraction, and other constituents which have been noted in such oils are limonene or dipentene, cineol, terpi- neols, terpinene and fenchyl alcohol. Turpentine is considerably modified by air oxidation, forming alcohols, terpineols, sobrerol, formic and acetic acids and resinous substances, and since moisture must be rigidly excluded from the preparation of bornyl chloride, the presence of very small traces of alcoholic or other oxidation products, which can form water by the interaction of hydrogen chloride, very materially decreases the yield of bornyl chloride and the use of old turpentine which has been exposed to air oxidation should accordingly be avoided. Testing of Turpentine: (a) Specific Gravity: This should be within the limits 0.862 to 0.870 at 20C. Lower specific gravity would indicate the presence of petroleum naphtha. A higher specific gravity would indicate the presence of wood turpentine, "pine oil" (terpineols), or that the tur- pentine has become oxidized by long storage. (b) Boiling-point: Nothing should distill below 154C. (except a drop or two of water), and 75% should distill below 160. Some specifications require that 95% should distill below 170. Petroleum naphthas are sometimes very closely cut so as to boil within this range (154 to 170), but ordinarily will show some distillate below 154. Limonene and dipentene boil at 176 and any considerable amount will be thus indicated. The terpineols boil at 210-218. Rosin spirit has a wide range of boiling-point, like petroleum naphtha, and also contains considerable dipentene. (c) Optical Rotation: It is not known whether pinenss of low optical rotation give better yields of bornyl chloride, as is the case with crystalline nitrosyl chlorides, or not. " J. Am. Chen*. 8oc. SO, 872 (1908). 496 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS (d) Refractive Index: At 20C. this should be within the limits 1.4680 to 1.4760. Low values would indicate petroleum naphtha. (e) Bromine and Iodine Numbers: These are chiefly useful in detecting petroleum naphtha but are hardly necessary when the other tests are made. It is difficult to carry out these determinations and get accurate, concordant results as with fatty oils, owing to substitu- tion reactions taking place with formation of halogen acid. Effect of Other Constituents on Bornyl Chloride Preparation: (1) (3-pinene yields the same hydrochloride as a-pinene: (2) Camphene, a minor constituent of turpentine oil, yields a low melting, unstable hydrochloride which probably yields camphene readily in the autoclave process. Its presence is not objectionable but is partly responsible for the partial decomposition of the crude bornyl chloride, free HC1 being given off. (3) Water: The presence of moisture in the reaction mixture causes the pinene to be converted chiefly into dipentene dihydro- chloride. which melts, when pure, at 50, but the impure mixture remains oily and retains considerable bornyl chloride in solution. Dipentene dihy- drochloride is much less stable than bornyl chloride and when such an oily mixture is heated, it is readily decomposed to free HC1 and dipentene. ... .; (4) Limonene, the optically active form of dipentene, forms the dihydrochloride and diminishes the yield of bornyl chloride in the manner indicated above. (5) Terpineols and other Alcohols yield water when treated with HC1; cf. item (3). B1CYCL1C NON-BENZENOID HYDROCARBONS 497 (6) Organic acids also cause the reaction with hydrogen chloride to go too far, with rupture of the C 5 ring to form dipentene dihydro- chloride. Distillation and Drying of the Turpentine: The secret of obtain- ing good yields of bornyl chloride is effective drying and purification of the turpentine. Mere drying is not sufficient as it is necessary to remove or destroy alcohols or other substances which yield water when treated with HC1, and metallic sodium is therefore best for this dual purpose. The still, which may be of iron or copper, should be pro- vided with a stirrer so that after the sodium is melted it will be thor- oughly emulsified in the oil. A small fractionating column is advis- able, in which case 90 per cent of American turpentine can be used for the preparation of bornyl chloride. Turpentine and Hydrogen Chloride: When the pinene has been well dried and purified from oxidation products and the hydrogen chloride is carefully dried, preferably by sulfuric acid Sp. Gr. 1.84, a yield of bornyl chloride corresponding to about 75 per cent of the theory can be obtained. Lead-lined or glass-enameled mixing vessels should be employed; iron, or alloys or other material which can yield iron chloride give liquid chlorides. Reaction temperatures above 30 yield increasing proportions of liquid chlorides but the reaction with hydrogen>chloride is slow below 15. Dry neutral solvents such as petroleum ether or carbon tetrachloride can be employed without diminishing the yield of bornyl chloride and the solidification of the reaction mixture can thus be prevented, but alcohol, ether or glacial acetic acid cause liquid chlorides to be formed. Bornyl chloride is remarkably stable for a chlorine derivative of the non-benzenoid hydrocarbon series. The statements in the earlier literature that it is decomposed slowly at room temperature and fairly rapidly at 100 probably refers to the decomposition of the crude product containing unstable impurities such as dipentene dihydro- chloride and camphene hydrochloride. Wallach 12 states that bornyl chloride distills practically without decomposition at 207-208, a fact which is familiar to those experienced in this field. Bornyl chlo- ride, once formed, is not changed by further treatment with hydrogen chloride, either dry or in the presence of moisture. True pinene hy- drochloride is readily converted to dipentene dihydrochloride by HC1. Pinene hydrochloride has never been isolated from the products of 110 Ann, 39, 4 (1887). 498 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS the reaction of pinene and hydrogen chloride but was made by Wal- lach 121 from nopinone by means of magnesium-methyl iodide. CH 3 -C-CH 3 xlx nopinone tCH 3 M Q I Ci pinene hydro chloride The mixture of oily chlorides accompanying the crude bornyl chlo- ride contains nearly 50 per cent of bornyl chloride in solution; thus, equal quantities of bornyl chloride melting at 131 and dipentene dihydrochloride melting at 50, melt down to an oil at room tempera- ture 20 to 22, and fenchyl chloride, which is an oil, has a similar solvent effect. When the oily chloride mixture is heated to about 180, the unstable chlorides are decomposed and a fairly brisk libera- tion of hydrogen chloride results. The resulting terpenes, chiefly dipentene, may then be distilled and the subsequent fractions boiling from 185.-215 yield an' additional quantity of crystalline bornyl chloride. About 10 per cent of the original oily chloride mixture remains behind as a heavy viscous mixture of polymers?. The amount of crystalline bornyl chloride which is recoverable in this way is equivalent to about 35 to 38 per cent of the original oily chlorides, when these chlorides are separated originally at 15. The formation of bornyl chloride from pinene and hydrogen chlo- ride is exothermic. 122 On account of the extraordinary stability of bornyl chloride many attempts have been made to employ catalysts to facilitate either the formation of camphene or conversion to bornyl esters. Anhydrous aluminum chloride reacts energetically with bornyl chloride, evolving hydrogen chloride and causing further decomposition and polymeriza- tion. Anhydrous ferric chloride is markedly less active and fused zinc chloride is still less active. Stannic chloride and titanium chlo- ride are much like zinc chloride in their effect on bornyl chloride. Cuprous chloride, or finely divided copper, is claimed to have a cata- 121 Ann. 356, 227 (1907). 1M Guiselin, Chem. Ztg. SI, 1299 (1910). Large scale work ihowed 119,000 calorie* liberated on treating 100 kilos of turpentine. are BICYCLIC NON-BENZEN01D HYDROCARBONS 499 lytic effect upon a wide variety of reactions of both alkyl and aryl chlorides, as in the manufacture of glycol, 123 or the conversion of chlorobenzene to phenol, 12 * but appears to be of no value in reactions of bornyl chloride. Barium chloride markedly catalyzes the decom- position of simple alkyl chlorides to olefines 125 and calcium chloride causes rapid condensation of benzyl chloride. But zinc chloride ap- pears to be the only catalyst appearing in the patent literature of the bornyl chloride reactions. 126 The purpose of this catalyst is to avoid the higher temperatures and pressures usually necessary for the com- plete conversion of bornyl chloride to camphene and bornyl acetate, when acetic acid and sodium acetate are used. However, considerable polymerization invariably takes place and one patentee 127 seeks to avoid this by introducing sodium acetate at intervals which has the effect of converting the zinc chloride into zinc acetate and sodium chloride, the latter separating on account of its slight solubility in acetic acid. Another patentee refluxes a solution of bornyl chloride in formic or acetic acids and adds zinc formate or acetate. 128 These reactions are quite analogous to the conversion of chloropentanes to amyl acetates by heating with sodium acetate in acetic acid solutions. In both cases zinc salts cause the formation of 10 to 25 per cent of heavy viscous polymerized hydrocarbons. Conversion of Bornyl Chloride to Camphene and Bornyl Acetate. In the following discussion no attempt is made to distinguish be- tween camphene and bornylene. The difficulty with most of the processes for making camphene from bornyl chloride by heating with alkalies, is chiefly a mechanical one, i. e., the insolubility of bornyl chloride in alkalies and inorganic alkaline mixtures. Naturally vigorous agitation affords better con- tact of the reacting substances and the presence of a fine solid sus- pension, milk of lime, assists in the emulsification. 129 The addition of fatty acid soaps has been proposed 130 and molten alkali pheno- lates 131 also have been suggested. Complete miscibility is obtained "Matter, U. S. Pat. 1,237,076. 124 Meyer & Bergius, U. S. Pat. 1,062,351; Ber. 47, 3155 (1914) 12 *Braun & Deutsch, Ber. 45, 1271 (1912). 126 Bergs, U. S. Pat. 903,047; Weizman, U. S. Pat. 910,978; von Hcyden, U. S. Pat. 919,762. 127 Ruder, U. S. Pat. 1,105,378. 128 Philipp, U. S. Pat. 919,762. 128 Schmitz & Stalman, U. S. Pat. 1,030,334. 18 Stephan, U. S. Pat. 725,890. 111 Koch, U. S. Pat. 970,829 ; Bergs, U. S. Pat. 833,666. 500 CHEMISTRY OF THE NON-BtfNZENOID HYDROCARBONS when bornyl chloride is heated with organic bases such as aniline, 132 naphthylamine, 133 pyridine, 134 or alcoholic ammonia. 135 The aniline process gives very good yields, about 90 per cent of the theory but an excess of aniline is necessary, as otherwise, bornyl aniline hydrochlo- ride is formed and this substance is not easily decomposed. When bornyl chloride is heated with sodium acetate in acetic acid in an autoclave to 180 to 200 the bornyl chloride is almost quan- titatively converted into camphene 138 and bornyl acetate. The cam- phene and acetic acid may be distilled together from the resulting reaction mixture and converted to bornyl acetate by the addition of a small quantity of sulfuric acid according to the well-known method of Bertram and Wahlbaum. (In order to separate the bornyl acetate thus formed from the excess acetic acid without diluting with water, a slight excess of sodium acetate may be added to form sodium sulfate and acetic acid, followed by fractional distillation in vacua.) Bertram and Wahlbaum 137 originally recommended acetylating camphene at 50, using a mixture such as the following: 2000 cc. acetic acid, 1000 cc. camphene, 50 cc. water and 50 g. sulfuric acid. Verley 138 recommends much more water, as indicated by the follow- ing: 450 parts sulfuric acid diluted to 60 to 66 per cent, 100 parts camphene, 100 parts acetic acid, the mixture being vigorously agitated at 30. Still better results, according to the writer's experience, are obtained by the method of Behal, 139 according to which the Bertram- Wahlbaum mixture is allowed to stand at room temperature for 24 hours. The formation of polymers is much reduced by operating at the lower temperatures. With pure camphene the yield of bornyl acetate is 92 to 94 per cent of the theory. When the resulting bornyl acetate is fractioned in vacuo, unchanged hydrocarbons pass over with the acetic acid fractions. Several other modifications of the Bertram- Wahlbaum reaction are obviously mere patent word play. It is possible that sodium formate and formic acid can be sub- stituted for acetic acid and acetate; in fact, such a process is described by Dubosc. 140 Henry has shown that sodium formate in methyl alcohol, and an alkyl halide, gives excellent yields of the corresponding 182 German Pat. 205,850 (1907) ; Bruhl, Ber. 25. 146 (1892) ; Ullmann & Schrnid, Ber. 43, 3202 (1910). 183 German Pat. 206,386 (1907). 13 *Weizmann, U. S. Pat. 896,962. ""German Pat. 264,246 (1912). 188 Wallach, Ann. 252, 6. 187 J. prakt. Chem. (2) $9, 1 (1894) ; German Pat. 67,255. 188 U. S. Pat. 907,428 (1908). "Austrian Pat. 38,203 (1908). 140 Brit. Pat. 14,379 (1907). BICYCLIC NON-BENZENOID HYDROCARBONS 501 alcohol, 141 and ethylene chloride can be smoothly converted to the glycol by this reaction. 142 Bornyl chloride is so remarkably stable, however, that, when using methanol, the reaction is slow at 180 and 330 Ibs. pressure. Heating bornyl chloride with alkali oxalates has also been tried. 143 Other Processes for Manufacturing Borneol or Bornyl Esters. The first attempt to manufacture artificial camphor on an indus- trial scale was in 1900 at Niagara Falls, where the Thurlow process 14 * was operated by the Ampere Electrochemical Co. At that time, tur- pentine could be had for about 35 cents per gallon but the yields of borneol were so low that the cost of artificial camphor by this method was considerably greater than the market price of natural camphor and the process was accordingly soon abandoned. In the Thurlow process anhydrous oxalic acid was added to dry turpentine at 120- 130. The reaction is energetic and much material was lost by the reaction becoming too violent. Dipentene was separated from borneol esters by distilling with steam, the esters saponified and the borneol oxidized to camphor by chromic acid mixture. It was found most expedient to purify the borneol before oxidation rather than to purify the camphor made from impure borneol. The Thurlow process had quite a few European modifications. Zeitschel 145 heated pinene and glacial acetic acid to 200 for five hours and reported a yield of 10 to 15 per cent camphene, about 40 per cent bornyl acetate and the remainder was dipentene. According to the writer's experience the yields of camphene and bornyl acetate are not improved by the addition of acetic anhydride. Fenchyl alco- hol is also formed and Bouchardat and Lafont observed 146 the forma- tion df fenchyl alcohol when using benzoic acid under similar condi- tions. Bischler and Baselli 147 treated camphene with anhydrous oxalic acid at 110-115; Seifert 148 used salicyclic acid and pinene at 110 for 50 hours; Austerweil 149 used "poly-substituted acids"; Hert- korn 150 heated turpentine with boric acid and absolute alcohol, etc. 141 Bull. acad. roy. Belg. 1902, 445. 142 Brooks and Humphrey, U. S. Pat. 1,215,903; J. Ind. d Eng. Chem. 9 t 750 143 Charles, Eng. Pat. 5.549 (1904). 144 U. S. Pat. 698,761 ; 833,095. 146 U. S. Pat. 907,941. 1M Compt. rend. 113, 551. 147 U. S. Pat. 876,310. 148 U. S. Pat. 779,377. 149 U. S. Pat. 986,038. "U. S. Pat. 901,293. 502 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS It is claimed 151 that by the action of tetrachlorophthalic acid on turpentine at 106-108 (12 hours) and finally at 140 (six hours) that esters of borneol are formed and that the borneol thus obtained, after saponification of the ester, is quite free from isoborneol. This method should therefore be operated to advantage in conjunction with the method of catalytically oxidizing the borneol to camphor by pass- ing over heated copper, since isoborneol is chiefly decomposed to camphene by this treatment. Hesse 152 has described the reaction of bornyl chloride, with mag- nesium in ether, as in the well-known Grignard reaction, and oxidizing the magnesium-bornyl chloride by air or oxygen to obtain borneol. This reaction, although patented, is of little interest since, like all the more complex alkyl halides, the reaction with magnesium is very slow and the main reaction is one of condensation, (a) C 10 H 17 C1 + Mg - -> C 10 H 17 MgCl (b) C 10 H 17 MgCl + C 10 H 17 C1 - - MgCl 2 + C 20 H M Regardless of its high cost, this method is not even a good laboratory method. By reacting upon bornyl chloride with milk of lime vigorously stirred, at a comparatively low temperature, 135 to 150 for about three days, an alcohol isomeric with borneol is obtained. 153 This alco- hol, camphene hydrate, is much less stable than borneol, melts at 149-150, boils at 206 and on heating with dilute acids is readily converted to camphene. This instability would indicate the struc- ture of a tertiary alcohol but its constitution is not yet definitely known. The treatment of pinene with ozone has also been described in a patented process 154 but hydrolysis of pinene ozonide does not really give borneol or camphor but pinonic acids (q.v.) and a series of other products. Bornylene ozonide might be expected to give camphor on hydrolysis. The oxidation of borneols to camphor by ozone has also been patented 155 but the industrial value of all oxidation methods depending upon ozone is questionable. Pat * aqUeS de produits chimi( l ues de Tfa an et de Mulhouse. ' 168 Sobering, 'German Pat. 219,243 (1908); Ber. 41, 1092 (1908). 1M Knox, U. S. Pat. 1,086,372 (1914). 1M Stephan & Hunsalz, U. S. Pat 801,483 (1905). BICYCLIC NON-BENZENOID HYDROCARBONS 503 Borneol and Isoborneol The relation of camphor to borneol is shown by the formation of borneol from camphor by reduction by sodium and alcohol. CH 3 CH 3 I I H CH 2 C C = CH 2 - -C C< CH. C CH 3 OH CH 2 C CH 5 H In addition to borneol, the closely related isoborneol is also formed in this reaction. The two borneols are commonly believed to be stereoisomers, i. e., CH 3 CH 3 H OH CH 2 C C< OH CH 3 C CH, CH 2 C CH 2 H Both yield camphor on oxidation and their behaviors on oxidation are nearly identical and for the purpose of manufacturing synthetic cam- phor need not be separately considered. Isoborneol is the principal product of the hydration of camphene in the Bertram-Walbaum reac- tion. Isoborneol is somewhat less stable than borneol and yields a "camphene," melting-point 50, when decomposed by the action of zinc chloride or dilute sulfuric acid. Isoborneol Borneol Crystal form hexagonal hexagonal Melting-point 212 203-204 Solubility in benzene at 1:2% 1:6% Solubility in ligroin at 20 1 :2% 1 :6 Phenylurethane, M.-P 138-139 138-139 Chloral compound, M.-P liquid 55- 56 Bromal compound, M.-P : . 72 98- 99 Zinc chloride, conversion to f camphene unchanged Dil. sulfuric acid, conversion to \ M.-P. 50 unchanged Sulfuric acid + CHaOH CIL ether unchanged Oxidation by CrO a camphor camphor Oxime of camphor produced, M.-P 118 118 Para-nitrobenzoate " 129 137 156 Henderson & Heilbron, Proc. Chem. Soc. t9, 381 (1913). The nitrobenzoate is conveniently prepared by treating with p-nitrobenzoyl chloride in pyridine. 504 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS When borneol or isoborneol is decomposed with loss of water, two hydrocarbons are produced, camphene being the principal product. The hydrocarbon formed* in smaller proportions is bornylene and this hydrocarbon retains the structure of the parent alcohols. CH CH A great deal of work has been done upon the structure of camphene and bornylene, which is reviewed elsewhere (q.v.) but in the earlier literature no distinction is made between these two hydrocarbons. Tschugaeff's method of preparing olefines by decomposing the methyl xanthate esters gives a fairly pure bornylene when applied to the decomposition of borneol. 157 Tschugaeff's bornylene melted at 109 to 109.5 and boiled at 146.5. Dezfro-bornylmethyl xanthate yields Zce?;o-bornylene and vice versa, an interesting example of the Walden inversion. Bornylene is noteworthy for its high melting-point, as compared with all other hydrocarbons, i. e., 113, and its boiling-point 146. Bornylene is less readily acetylated than camphene, by the Bertram-Walbaum method. A less pure bornylene may be made by treating bornyl iodide with alcoholic caustic potash. Bornylene is also more resistant to oxidation than camphene and Henderson and Caw 15S accordingly purified bornylene by oxidation by hydrogen peroxide and obtained a specimen showing the melting-point 113 and boiled at 146. A very pure bornylene made through camphocarbonic acid 159 also showed a melting-point of 113, and a boiling-point of 146. CH.C0 2 H CH.C0 2 H. C.C0 2 H CH C 8 H 14 < | * C 8 H 14 < 7 -+ C 8 H 14 < || -> C 8 H 14 < || c = o CHOH CH CH It is still generally believed that "camphene" may be a mixture of hydrocarbons, or that camphenes of different origin are not identical. The camphenes from various natural sources differ widely in physical 167 Ann. 388, 260 (1912). 188 J. Chem. Soc. 101, 1416 (1912). "Bredt, Ann. 366, 11 (1909) ; ,7. prakt. CJiem. (2) 81,, 778 (1911). NON-BENZENOID HYDROCARBONS 505 properties. Wallach 160 isolated a specimen of camphene from a Siberian pine-needle oil which showed a low melting-point, 39, a boil- 4ft ing-point of 160-161, d 4QO 0.8555, [a] D 84.9 and n_1.46207. Camphene made from bornylamine 161 melts at 50 and showed the high rotation of [<*]T\ 103.89. Ordinary camphene hydrochloride, melting at 155, is identical with the chloride of isoborneol. Oxidation of Borneol and Isoborneol As stated above, the old Thurlow process, practiced at Niagara Falls about 1900, employed chromic acid for oxidation of the borneols to camphor. Various special modifications of the chromic acid oxida- tion method have been described in the patent literature, and the processes of Verley, 162 Florizoone, 163 Jluder 16 * and Weizmann 165 men- tion the use of a solvent added to insure thorough exposure of the borneol to the oxidizing solution. Carbon tetrachloride, benzene and acetone 165 are useful for this purpose, but acetic acid forms appre- ciable proportions of bornyl acetate which resist oxidation to cam- phor. Verley recommends 50 parts of sodium dichromate, 68 parts of sulfuric acid and 600 parts of water but Ruder employs solutions of about one third this concentration. Free sulfuric acid should be avoided as much as possible on account of the decomposition of iso- borneol to camphene, which is more resistant to oxidation, by heating with dilute sulfuric acid, as noted above. Gradually acidifying the reaction mixture as the oxidation proceeds is therefore advantageous. The oxidation of camphene itself by chromic acid has been de- scribed 166 but the yields are lower than when borneols are employed. Another patentee 167 proposed to employ potassium persulfate for the oxidation of camphene. The use of sodium dichromate or chromic acid for this purpose, on a tonnage scale, involves the electrolytic regeneration* 68 of this oxidizing material or its utilization as basic chromium salt solutions in tanning or the mordanting of textile goods, otherwise the method would be too costly. "Ann. S57 f 79 (1907). "Wallach, Ann. 357, 84 (1907). 82 U. S. Pat. 908,171 (1908). 88 Brit. Pat. 5,513 (1908). M U. S. Pat. 1,066,758 (1913). "Brit. Pat. 21,946 (1907). "Dubosc, Brit. Pat. 8260-A (1906) ; 8356-A (1906). 67 Sauvage, French Pat. 389,092. M Ges. Chem. Ind. Basel, French Pat. 387,539; LeBlanc, Z. Elektrochem. 7. 290 (1900) ; McKee & Leo, J. Ind. & Eng. Chem. 12 f 16 (1920). 506 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The yields of camphor by oxidizing the borneols by air, in the absence of catalysts, are very poor, 169 but catalytic dehydrogenation of borneol to camphor, by jneans of finely divided copper 170 at 175- 180 or reduced nickel 171 at 200-240 is said to have been practiced industrially. Aloy and Brustier 172 state that when borneol is passed over copper at 300 the yield of camphor is quantitative but that above 320 the yield of camphor is progressively diminished until at 420 hydrocarbons only are produced. Camphor is not reduced to borneol by hydrogen and catalytic nickel at 180-200, either alone or in solution in cyclohexanol. Neave 173 states that borneol yields camphor in nearly quantitative yields by passing over finely divided copper at 300 but that isoborneol under the same conditions gives chiefly camphene. Thorium oxide at 350 yields a terpene mixture boiling at 150 to 180, the constituents of which were not definitely characterized. 17 * Small proportions of unchanged bornyl chloride or other chlorides poison the catalyst unless the material is previously purified to remove such chlorides, as, for example, by digesting with a little inert solvent over metallic sodium. Quite a number of processes for the oxidation of the borneols by nitric acid or oxides of nitrogen have been described. Hesse 175 used pure concentrated nitric acid; another process 176 prescribes nitric acid containing oxides of nitrogen, at 10 to 15, and nitrous acid itself is said to give excellent yields. 177 The addition of small amounts of vanadium pentoxide to the nitric acid is claimed to be advantageous and several patents have recently been granted to Andreau, 178 who employs a mixture of about 339 parts of sulfuric acid 66 Be, and 253 parts of nitric acid, 26 Be, and who notes that once the oxidation has been initiated by raising the temperature to about 40, the reac- tion may then be carried out smoothly with cooling so that the tem- perature does not rise above 40. In the nitric acid process the cam- phor forms a liquid double compound with the nitric acid, which floats on the acid mixture as a sparingly soluble oil 'layer. This obviates the use of a solvent to insure complete oxidation of the 1M Cf. Stephan & Rehlander, TL S. Pat. 801,485. 170 Sobering, German Pat. 161,523; Goldsmith, Brit. Pat. 17,573 (1906). 171 Aschan & Kempe, U. S. Pat. 994,437 (1911) ; Zimmerman, Brit. Pat. 26,708 (1904) . . i 172 Bull. soc. chim. (4) 9, 733 (1911). 178 J. Chem. Soc. 101, 513 (1912). 174 Aloy & Brustier, J. pharm. chim. (7) 10, 49 (1914). Ber. 39, 1144, (1906). Ges - Chem - Ind - Basel, Brit. Pat. 9,857 (1907); Philip, Austrian Pat. 33,720 . 177 Boehringer & Son, U. S. Pat. 802,793 (1904). 178 U. S. Pat. 1,347,071 (1920). BICYCLIC NON-BENZENOID HYDROCARBONS 507 borneol, enclosure of borneol particles by solid camphor being avoided. The oily nitric acid compound is decomposed by water, precipitating the camphor. Camphoric acid and nitrocompounds are also formed, the latter coloring the crude camphor light yellow, and imparting to it a peculiar "nitro" odor. Practically every known method of oxidizing organic compounds has been proposed for the oxidation of the borneols, or camphene, to camphor, including chlorine, 179 hypochlorites, 180 potassium perman- ganate both in acid 181 and alkaline 182 solution, etc. When perman- ganate is employed the camphor formed is removed from the spent mixtures by distilling with steam. Camphene, in dilute acetone, has also been oxidized by potassium permanganate, to camphor. 183 All of these methods using permanganate are relatively very costly, except where methods for its regeneration have been perfected. Impurities of Crude Synthetic Camphor If the borneol or isoborneol is not purified before oxidation, the resulting camphor will contain small proportions of the fenchones, which, like camphor, are quite resistant to further oxidation and form very stable double compounds with nitric acid. The behavior of the fenchenes in the Bertram- Walbaum reaction follows the general esterification behavior of unsaturated terpenes. Komppa and Hinticka 184 share Quist 's view that isof enchene, boiling- point 152-155 has the constitution. CH 3 CH CH C< CH. H, CH c CH, CH 3 As noted above the chief impurity in bornyl chloride is dipentene dihydrochloride but fenchyl chloride is present in the oily part of the 178 Boehringer & Son, TJ. S. Pat. 802,792; Brit. Pat. 28,035 (1904). 180 Hertkorn, U. S. Pat. 901,708 (proposes the addition of salts such as CuCla and FeCl 3 ) ; Glaser, U. S. Pat. 875,062; 864,162 (1907). 181 Semmler, Ber. 33, 3430 (1900). 182 Sobering, German Pat. 157,590 (1903) ; Stephan and Hunsalz, U. S. Pat. 770,940 (1904) . 188 Behal, Austrian Pat. 38,203 (1908). 1M Chem. Abs. 13, 2864 (1919). 508 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS hydrochloride mixture since fenchene has been found in the crude camphene made from these chlorides. Aschan 185 represents the for- mation of bornyl chloride and fenchyl chloride as follows, CH 3 CH-( :-CH 3 bornyl chloride CR CH, - N CH, chloride of fenchyl alcohol Crude camphene also contains a very small amount of p-pinolene or tricyclene. It will be noted that the fenchyl chloride or the cor- responding alcohol whose structure is shown above cannot lose HC1 or water to form a double bond with either of the adjacent carbon atoms. But Quist made tricyclene by decomposing the methyl xan- thate ester of fenchyl alcohol and therefore shares Aschan's views as to the nature of tricyclene and its method of formation, H Tricyclene is stable toward alkaline permanganate but, as with most cyclopropane structures, acids rupture the 3 carbon ring and the Bertram- Walbaum reaction accordingly gives the acetate of iso- fenchyl alcohol. Hydrogen chloride at 10 yields a hydrochloride Ann. 887, 24 (1912). BIG YC LIC NON-BENZENOID HYDROCARBONS 509 melting at 27.5 to 29, which on decomposition forms a fenchene boiling at 154. Tricyclene itself, as purified by Aschan by oxidizing the accompanying fenchenes by permanganate, boils at 141.5-143.5. Fenchenes or tricyclene contained in the crude camphene, employed for the manufacture of artificial camphor, will accordingly be con- verted to fenchyl and isofenchyl alcohols which will in turn be oxi- dized to the corresponding ketones. The relations of these substances are probably as follows, CH, fenchyl alcohol CH, p-pinolene (tricyclene) CH 3 isofenchyl alcohol CH 3 isofenchone As regards the purification of synthetic camphor for industrial purposes, it should be noted that manufacturers of celluloid usually specify that the chlorine 186 content shall not exceed 0.1 % and borneol should not be present in excess of 0.5 per cent. For some grades of celluloid a melting-point of 165 is sufficient but for high-grade ma- terial the melting-point should not be lower than 174. A saturated solution in 95 per cent alcohol should show no yellow color and when kept in ordinary diffused daylight in a colorless transparent bottle the 186 For the quantitative determination of chlorine in synthetic camphor the method of Drogin and Rosanoff (J. Am. Chem. Soc. 38, 711 [1916]), or that of Van Winkle and Smith (J. Am. Chem. Soc. 42. 333 [1920]) is recommended. The per cent, of borneol may be determined by acetylating with acetic anhydride, in the usual manner, and determining the saponification number of the product ; borneol or isoborneol mny also be separated from camphor by the phthalic anhydride method. 510 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS camphor should not become visibly discolored in 30 days. Nitro derivatives are very apt to cause the development of a yellow color. The presence of nitro derivatives also causes the formation of a slight tarry or resinous residue when a few grams of the camphor are sub- limed slowly from a watch glass. The chief impurities encountered in commercial natural camphor are camphor oil, water and mineral matter. As regards the yields of synthetic camphor obtained in industrial practice there are naturally no reliable published data. Schmidt 187 gives the following yields: solid bornyl chloride 43 per cent, camphene from bornyl chloride 95 per cent, isobornyl acetate from camphene 86 per cent, saponification to isoborneol 98 per cent, oxidation of iso- borneol to camphor about 80 per cent, or a net yield from the original turpentine of 24 per cent of the theory. Austerweil 188 gives the yield of crystalline bornyl chloride as 55 to 60 per cent and Ullman 189 gives 55.3 per cent of the theory as the yield of the chloride. Accord- ing to the writer's experience these yields are too low, particularly as regards bornyl chloride, which with reasonable skill can be obtained in yields of 75 to 78 per cent of the theory, and the acetylation of camphene can be relief upon to give a yield of isobornyl acetate cor- responding to 92 to 94 per cent of the theory. The net yield of cam- phor should be 45 to 50 per cent of the theory. Methods for the synthesis of camphor which are of theoretical interest are discussed in connection with the constitution of camphor. ""Chem. Ind. 29, 241 (1906). "'German Pat. 211,799 (1908). "Enzykl. techn. chem. Ill, 257. Chapter XIV. Cyclic Non-benzenoid Hydrocarbons. Cycloheptane, Cyclooctane, Cyclononane and Polynaphthenes. Cycloheptane and its derivatives are difficult to prepare and have been comparatively little studied. The ketone, cycloheptanone (suberone) , is the material most frequently employed for the prepa- ration of other cycloheptane derivatives and Willstatter has used the cycloheptatriene (tropilidene) formed by the exhaustive methylation and decomposition of tropidine, and also cycloheptatriene from anhy- dro-ecgonine. Eucarvone has also been employed in the preparation of other cycloheptane derivatives. The physical properties of cycloheptane, cycloheptene, cyclohep- tadiene (hydrotropilidine) and cycloheptatriene (tropilidine) are as follows, 1 Boiling-Pomt d n D Cycloheptane 117. -117.5 0.8253 Cycloheptene 114.5-115. 0.8407 A 1-3 -cycloheptadiene 120. -121. 0.8810 1.495997 A 1 ''''-cycloheptatriene 116. (corr.) 0.9083 1.5175 Cycloheptane was made by Markownikow 2 from cycloheptanone by reducing the ketone to cycloheptanol (suberyl alcohol) and reduc- ing the corresponding bromide by zinc dust and alcohol. Willstatter and Kametaka 3 reduced cycloheptadiene (hydrotropilidene) by Saba- tier and Senderens' method, at 180. The cycloheptane made under these conditions is quite pure but at 235 further hydrogenation to normal heptane occurs and at 250 this change is quite rapid. Cyclo- heptanol cannot be reduced to cycloheptane by heating .with hydri- odic acid, methylcyclohexane being formed.* The formation of a hydrocarbon, C 7 H 8 , by distilling methyltropine- 1 Willstatter, Ann. 317, 204 (1901). 'Ann. 327, 59 (1903). Ber. 41, 1480 (1908). Markownikow, J. prakt. Chem. (2) 49, 430 (1894). 511 512 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS methyl iodide was observed by Ladenburg, and Merling 5 obtained the hydrocarbon by exhaustive methylation of tropidine and decom- posing the tertiary ammonium hydroxide by heat. The conversion of tropinic acid to normal pimelic acid led to the view that the tropine bases and their nitrogen free decomposition products possessed a cycloheptane ring and that tropilidine and hydrotropilidine were cycloheptatriene and cycloheptadiene respectively. Willstatter con- firmed this by synthesizing both hydrocarbons. Cycloheptene can readily be prepared by decomposing cyclo- heptyl iodide in the usual manner, and the addition of bromine gives 1 . 2-dibromocycloheptane, but when the dibromide is heated with quinoline two bromine atoms are removed, not two molecules of hydro- bromic acid, the resulting product being cycloheptene, not the expected diene. Alcoholic, caustic potash converts the dibromide into the un- saturated ether, and similarly, heating the dibromide with dimethyl amine forms a dimethyl amino derivative, CH 2 CH 2 CHBr CH 2 CH 2 CH . N (CH 3 ) 2 HBr \ + 2NH(CH 3 ) CHBr \ CH I 2 _ CH 2 CH 2 CH 2 CH 2 CH By adding methyl iodide to the resulting base and decomposing the tertiary ammonium hydroxide by heat, Willstatter made A 1 - 3 -cyclo- heptadiene, which proved to be identical with hydrotropilidene H CH 2 CH 2 CN(CH 3 ) 3 .OH CH 2 CH = CH CH -> CH. CH 2 CH 2 CH CH 2 CH 2 CH Willstatter also made A 1 - 3 -cycloheptadiene in another manner. 6 From the decomposition products of cocain A 1 -cycloheptenecarboxylic acid was obtained, which was treated with hydrogen bromide to form 2 bromocycloheptanecarboxylic acid, which was decomposed, losing "Ber. 24, 3109 (1891). The cycloheptane ring is bridged in the following manner, CH 2 CH CH 2 NH > CHa (tropane). f*TT , f*TT - I^TT Einhorn & Willstatter, Ann. 280, 136 2 (1894). BICYCLIC NON-BENZENOID HYDROCARBONS 513 HBr to form a mixture of the A 1 and A 2 acids. The A 2 acid was sepa- rated by fractional crystallization; converted to the amide, and the latter~"treated with bromine and alkali (Hofmann's reaction) to form the amine, from which, by the method of exhaustive methylation, the conjugated diene was made. CH, CH; CH: The addition of bromine to A 1 - 3 -cycloheptadiene takes place in accordance with the general rule of the addition of bromine to con- jugated dienes, to form 1 . 4-dibromo-A 2 -cycloheptene, which by heat- ing with quinoline, loses 2 molecules of hydrogen bromide to form cycloheptatriene CH 2 CH =CH CH Br, CH CH 2 CH CH Br CH CH \ CH )H 2 CH 2 CH.Br CH =CH CH \ / CH Cycloheptatriene resinifies rapidly in contact with the air and follows generally the behavior of conjugated dienes. Tetramethylcycloheptatriene was made by treating eucarvone with magnesium-methyl iodide. 7 It is not definitely known whether the 'Rupe & Kerkcvius, Ber. U, 2702 (1911). 514 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS third double bond is in the ring or semicyclic. The physical proper- ties of the hydrocarbon are as follows, boiling-point 67-67.5 (11 mm.), d 20 o 0.8687, n D 1.5066, M D 50.70, EM D 1.33. Its constitu- tion may be inferred from the constitution of eucarvone (q.v.). It is very little changed by boiling with 10 per cent sulfuric acid. Reduction by sodium and ethyl alcohol yields a diene, CuE^, distilling at 64.5- 65.5 (12mm.). Diazoacetic ester combines with toluene and the xylenes to form derivatives of norcaradienecarboxylic acid. Thus when toluene and diazoacetic ester are boiled together (copper powder as a catalyst is not necessary) nitrogen is rapidly evolved and 3-methylnorcara- dienene-7-carboxylic ester is formed. Para-xylene, treated in the same way, yields the bicyclic ester, and this ester can be treated in several ways to break the 3-carbon ring. The first condensation product is regarded by Buchner and Schulz 8 as the ethyl ester of 2.5-dimethyl-A 2 - 4 -norcaradienenecarboxylic acid. By heating the amide to 160-170, or heating the crude condensation product with 15 per cent sulfuric acid, or by heating with water at 160-170, the 3-carbon ring is broken, forming chiefly 2.5-dimethyl-A 2 - 5 - 7 -cyclohep- tatriene-7-carboxylic acid, melting-point 136-137. CH 3 CH=:C CH ..CH.C0 2 C 2 H 5 = C CH CH 3 --jo When the A 2 - 5 - 7 acid is reduced by sodium amalgam two atoms of hydrogen are added forming what Buchner regards as the A- 5 acid, melting-point of the crude acid 38-40, but too unstable to purify. Obviously a number of isomeric acids containing two double bonds are possible, and by adding hydrogen bromide to the A 2 - 5 acid and then removing HBr by the action of alkali, Buchner obtained an isomeric acid melting at 82, which he regards as the A 2 - 6 acid. Reduction by hydrogen and platinum black yields 2 . 5-dimethylcycloheptanecar- boxylic acid, an oil at ordinary temperatures (amide melting at 185- 186). Ann. 378, 259 (1910). BICYCLIC NON-BENZENOID HYDROCARBONS 515 Goldsworthy and Perkin 9 made trans. 1 . 2 . 4 . -cycloheptanetri- carboxylic acid by the latter's well-known method of synthesis, using sodium ethylate as a condensing agent, C(XC 2 H 5 CH 2 CH(C0 2 C 2 H 5 ) 2 CH 2 CH< CH 9 Br CH 2 C HBr CH CO,C 2 H 6 CH 2 + | * CH CHBr CH 2 -CH(C0 2 C 2 H 5 ) 2 C0CH C H 2 CH C0 2 C 2 H 5 The ester was saponified by alcoholic caustic potash in the usual manner, the free acid melting at 198-200. Cycloheptanone, the raw material most frequently employed for preparing cycloheptane derivatives, may be prepared by heating the calcium salt of suberic acid. 10 When purified by means of the semi- carbazone or the bisulfite compound and regenerating the ketone, it 21 5 has the following physical properties, boiling-point 180, d 10.9498, n D 1.46027, M D 32.35 (calculated 32.34) ." Cycloheptanone forms a dibenzylidene derivative, C 7 H 8 0.(CH.C 6 H 5 ) 2 , melting at 107-108, and like cyclopentanone and cyclohexanone, forms a series of well crystallized compounds with other aldehydes (with anisaldehyde, melting-point 128-129 ; with cinnamic aldehyde, melting-point 198; with piperonal, melting-point 137). It con- denses with acetone but with much greater difficulty than the lower cyclic homologues. 12 Like other cyclic ketones the oxime is rear- ranged by sulfuric acid to the so-called isooxime, CH 2 CH 2 CH 2 CO CH 2 CH 2 CH 2 NH This isooxime is readily split by heating with hydrochloric acid to amido-n-heptylic acid. The ketone reacts normally in the Grignard reaction, for example, with magnesium-methyl iodide to form 1-meth- 9 J. Chem. Soc. 105, 2675 (1914). 10 Wislicenus & Mager, Ann. 275, 357 (1893). "Auwers, Ann. 410, 283 (1915). "Wallach, Ann. 394, 366 (1913). 516 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS ylcycloheptanol-(l), which in turn readily decomposes to k l -methyl- cycloheptene, boiling-point 137.5-138.5, d 1ft _ 0.824, n^ 1.4581 iy.o D The reactions of this hydrocarbon are parallel to those already dis cussed; for example, it forms a nitrosochloride (melting-point 106 C which on heating with dimethylaniline yields the unsaturated oxime which in turn may then be hydrolysed by dilute acids to the unsatu rated ketone A 1 -2-methylcycloheptene-3-one (boiling-point 200- 205, d 0.9695). CH C u t CH, XH , CH,r CH, CH t CM VCH / CH CH, -CH, C^ -CH, CH, -CH, C; .NOH \ ;C-CH, CH, CH CH Z CO -CH 2 CH C-CH, It also condenses with bromo-acetic ester in the presence of zinc to give cycloheptanol acetic acid, from which Wallach 13 obtained meth enecycloheptane in the usual manner. Methenecycloheptane dis- tills at 138-140, d 0.824, n 1.4611. This hydrocarbon under- goes reactions strictly parallel with those which have already been discussed in connection with other hydrocarbons having the methene >C = CH 2 group, for example, on oxidation it forms a glycol (melt- ing-point 50-51) which is converted to cycloheptane aldehyde by heating with dilute acids. Kotz 14 has studied the chlorination and bromination of cyclohep- tanone and finds that, like cyclohexanone, the halogen enters the ring in the CH 2 group adjacent to the carbonyl group, these facts harmonizing with the view that the ketone reacts with the halogen in the enol form, adding C1 2 or Br 2 and subsequently splitting off a mole- cule of halogen acid. The chloroketone is much more stable than the bromoketone. The chloroketone is not hydrolyzed by aqueous caustic potash at ordinary temperatures but, on warming, the corresponding oxyketone is formed (yield poor) . Oxyketones of this type show most interesting properties; neither the oxyketone nor its methyl ether forms an oxime and the methyl ether may readily be prepared by Ann. Slit, 158 ; 3^5, 146. "Ann. 400, 47 (1913). B1CYCLIC NON-BENZENOID HYDROCARBONS 517 saturating the methyl alcohol solution by hydrogen chloride, like the esterification of a carboxylic acid. The unsaturated ketone, A 2 -cyclo- heptenone, was reduced by Kotz, by Paal's method, to cycloheptanone, confirming Willstatter's 15 constitution for this ketone (tropilene). Eucarvone: When carvone combines with one molecule of hydro- bromic acid and is then treated with alkali to remove HBr, the result- ing ketone proves to be an isomer of carvone. Baeyer, the discoverer of the reaction, regarded eucarvone as bicyclic having a cyclopropane ring although he himself pointed out several objections to such a structure. Dihydroeucarvone and tetrahydroeucarvone he regarded as derivatives of cycloheptanone. Further objections to Baeyer 's con- stitution for eucarvone CH 3 H 2 C CH eucarvone (Baeyer) C H\ C(CH 3 ) were pointed out by Wallach, who prepared a condensation product with benzaldehyde, clearly indicating the presence of the CH 2 CO group. Also, when prepared from optically active carvone, eucarvone, according to Baeyer's constitution, should be capable of optical activity, but, as Baeyer himself observed, it is inactive. Reduction of eucarvone by sodium gives dihydroeucarvone and a little tetra- hydroeucarvone, but by catalytic hydrogenation in the presence of palladium tetrahydroeucarvone is readily produced, 16 a method of reduction which practically precludes rearrangements. Baeyer had already shown that tetrahydroeucarvone was a derivative of cyclo- heptanone, being oxidized in the following manner, CH 2 CH 2 CH 2 CH 2 CH 2 CH 2 \ CH-CH \ CO-CH CH, C CH, C = CH, - C OH, C0 H CH 3 CH 3 keto acid. 18 Per. 44, 465 (1911). Others considered tropilene to be tetrahydrobenzaldehyde. " Wallach, Ann. SS9 f 107 ; S81, 67. 518 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS by Ca salt CH 2 CH 2 CH 2 CH 2 CH 2 CH 2 CH 3 C0 9 H. CH a - CH, CO CH 2 C0 2 H fifi-dimethylpimelic acid dimethylcyclohexanone. Wallach therefore explains the behavior of eucarvone by the foH lowing formula, the bicyclic ketone assumed by Baeyer evidently be- ing an unstable intermediate product, :H, CH hydrobromocarvone unstable intermediate eucarvone product (Wallach) The reduction by sodium indicates that the two double bonds in eucarvone are in the conjugated position, but the constitution of the dihydro derivative is not definitely known, although the reduction probably follows the partial valence rule of Thiele, leading to the following constitution for dihydroeucarvone. HC CH 3 C HC CH 2 HC - C(CH 3 ) OH, HC C = HC CH 2 H 2 C C(CH 3 ) dihydroeucarvone ? BICYCLIC NON-BENZENOID HYDROCARBONS 519 Eucarvone has the following physical properties, boiling-point 85- 20 87 (12 mm.); d 0.952, n -1.5048. 18 Its semicarbazone melts at 183-185, the oxime at 106 and the benzylidene compound, fprmed by its reaction with benzaldehyde, melts at 112-113. By partially hydrogenating the oxime of eucarvone Wallach 19 discovered the oxime of a dihydroeucarvone, which is not identical with that previously known, and Wallach accordingly distinguishes the two known dihydroeucarvones as a and |3, the latter being the newly discovered one. Oxime Semicarbazone B.P. d M.P. M.P. 21 205 0.9215 liquid 189-191 213-214 0.9325 122-123 195-197 Tetrahydroeucarvone has the following physical properties, boiling- point 207, d 0.906, n.^ 1.4553; it yields a semicarbazone, obtained in two forms one melting at 201 and the other at 161-163; the oxime is an oil when made from the saturated ketone but when made by the catalytic hydrogenation of eucarvoxime, melts at 56-57, the further reduction of which yields tetrahydroeucarvylamine. Cyclooctane: Just as tropin may be oxidized to tropinic acid and to normal pimelic acid, pseudopelletierin may be oxidized to suberic acid. The alkaloid from pomegranate root therefore contains a cyclo- octane ring, in fact, as Ciamician and Silber, 20 and Piccinini 21 have shown, the alkaloid contains the bridged ring CH 2 - -CH- -CH 2 CH 2 N.CH 3 CO CH 2 - -CH- -CH 2 By exhaustive methylation of the base, adding methyl iodide to cfes-dimethylgranatanine and decomposing the free tertiary ammo- nium hydroxide, Willstatter and Veraguth 22 obtained a cycloocta- diene. The octadiene polymerizes so readily that when distillation was attempted polymerization occurred at 130-150, with almost 18 Wallach, loc. cit. 19 Ooett. Nachr. 1913, 246. Ber. 26, 156, 2738 (1893) ; 27, 2850 (1894) ; 29, 490, 2970 (1896). 21 Gazz. chim. ItcU. 29 (2), 104 (1899). M Ber. S8 t 1976 (1905). 520 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS explosive violence, to a resinous mass and resinification takes place rapidly on standing in contact with air. At 16.5 mm. it distills at ^ 39.5, d 0.889. It has a very penetrating, unpleasant odor and inhalation of the vapors readily produces headache. It reacts readily with nitric acid, dissolving completely. On standing the hydrocarbon is converted to a dimeride which may be crystallized out by strongly cooling, the crystals melting at 114. Bromine reacts with the hydro- carbon very energetically even at 10 in chloroform solution, some hydrobromic acid being evolved. When the dihydrobromide is de- composed by heating with caustic alkali or quinoline, a new octadiene is formed, distilling at 143-144 and characterized by a very pleas- ant odor, and almost no tendency to polymerization. The former cyclooctadiene has properties very strikingly similar to those of cyclopentadiene and Willstatter accordingly regards the first diene as having the double bonds in conjugated positions, i. e., A^-cyclo- octadiene. The more stable cyclooctadiene, evidently A 1 - 4 or A 1 - 5 , is smoothly reduced by Sabatier and Senderen's method to cyclo- octane, boiling-point 146.3 -148. Cyclooctane is stable to per- manganate but is oxidized by nitric acid to suberic acid. When the unstable diene, a-cyclooctadiene, is treated with hydro- bromic acid, the dihydrobromide formed is always accompanied by a saturated monohydrobromide, from which a bicyclic octene is formed by heating with quinoline. This bicyclic hydrocarbon distills at 138-139, d 0.9097. Its constitution is not definitely known; by carefully oxidizing with permanganate a crystalline a-oxyketone CO is formed, C 6 H 10 <| CHOH. The yield of cyclooctanone obtained by heating the calcium salt of azelaic acid is very poor. The ketone distills at 195-197, melt- ing-point 25-26. By distilling the barium salt of p-vinylacrylic acid Doebner 23 has obtained a cyclooctadiene boiling at 50-52 (17mm.) which he regards as A 1 - 5 -cyclooctadiene. Sorbic acid treated in the same way yields S^-dimethyl-A^-cyclooctadiene, boiling-point 68-71 (15mm.). By decomposing the tetrabromide of the more stable or |3-cyclo- "Ber. 35, 2129 (1902) ; Ber. 40, 146 (1907). BICYCLIC NON-BENZENOID HYDROCARBONS 521 octadiene, Willstatter 24 obtained cyclooctatetrene. Cyclooctatetrene has been of interest because of the fact that it possesses none of the properties characteristic of the benzenoid hydrocarbons. It readily combines with 4 molecules of hydrogen, reduces permanganate and absorbs bromine. With nitric and sulfuric acid nitrating mixtures, it yields resin but no nitro- derivatives. Its structure is therefore to be represented as follows: / \ HC CH H e Willstatter, who discovered the substance, accordingly states that benzene cannot have the structures indicated by the Kekule constitu- tion. No indication of the existence of a more stable form was ob- tained and Willstatter, favoring the Armstrong and Baeyer constitu- tion for benzene, considers that in the case of an eight carbon ring centric equilibrium of the fourth valence of each carbon atom is not established because the distance of the carbon atoms from the center, or from opposite carbon atoms, is greater than in the case of benzene. As might be surmised from the properties of the fulvenes, noted above, cyclooctatetrene is a yellow oil of very powerful odor, oxidizes rapidly in contact with air and readily polymerizes. Cyclononane : Calvi 25 attempted to prepare cyclononanone by heating the calcium salt of sebacinic acid but without success, and later Petersen 26 attempted the same preparation but noted a very complex decomposition, identifying benzene, propionic aldehyde and heptane aldehyde among the products. Dale and Schorlemmer 2T noted the formation of a hydrocarbon C 16 H 32 distilling at 283-285, under similar conditions. Zelinsky 28 obtained only 20 grams of a ketonic substance from 2 kilos of sebacinic acid, from which he pre- pared the semicarbazone and regenerated the ketone in the usual oo 5 manner. The ketone distills at 95-97 (17-18 mm.), d _ 0.8665. Zelinsky reduced the ketone, with sodium and moist ether, to the alcohol which he converted to the iodide and reduced the latter by 14 Willstatter & Waser, Ber. 44, 3423 (1911). "Ann. 91, 110 (1854). "Ann. 103, 184 (1857). "Ann. 199, 149 (1879). Bvr. 40 , 3278 (1907). 522 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS zinc dust in acetic acid. The quantity of the hydrocarbon made by Zelinsky was evidently too small for great reliance to be placed upon its physical constants and inferences drawn therefrom. According to Windaus 29 cholesterol contains a bicyclic nine carbon atom ring, having obtained evidence of a ring of the character indi- cated by the following partially elucidated formula, CH 9 CH 9 CH CH 2 CH 2 CH.CH = Polynaphthenes: Lubricating oils of the general empirical for- mula C n H 2n = 2 and C n H 2n= 4 are generally believed to be poly cyclic hydrocarbons although there is no direct evidence of their constitu- tion other than the empirical formula and inferences drawn from their physical properties such as viscosity and their not crystallizing at low temperatures. Such oils, even when repeatedly distilled and fractions separated within very narrow temperature limits, are un- doubtedly very complex mixtures and the great difficulty of separat- ing pure substances from even simpler mixtures has already been pointed out. The task of determining the constitution of such hydro- carbons, assuming it to be worth while, would be an almost impossible one and probably not worth the enormous effort required. When formaldehyde or para-formaldehyde is heated with tetra- hydronaphthalene and phosphorus pentoxide, a highly viscous oil it? obtained, boiling at 257-258 (15mm.). A brittle resin is also formed. The oil has been proposed as a lubricant and its process ot manufacture patented. 30 Heusler and Engler 31 observed the poly- merization of light, low-boiling, unsaturated hydrocarbons to poly- naphthenes of the lubricating oil type, by heating the former hydro- carbons under pressure. It is quite possible that the lighter unsaturated hydrocarbons found in shale oil distillates could be polymerized by such a process, 32 or by treating with anhydrous aluminum chloride, to more stable polymers of the lubricating oil type; in fact, such u "Ber. 53, 488 (1920). 80 German Pat. 333,060; 319,799. sl Ber. 28, 490 (1895) ; 30, 2358, 2365 (1897). 12 Phenols and organic acids present in crude shale oil naturally destroy the efficacy of the aluminum chloride, unless removed. BICYCLIC NON-BENZENOID HYDROCARBONS 523 process, using aluminum chloride, has already been carried out indus- trially in the United States, polymerizing the low-boiling olefines made by cracking (such olefines being largely lost by the usual refining process), to lubricating oils of good quality. In the absence of scientific information regarding the chemical constitution of lubricating oils, a great many claims are sometimes made for the supposed excellence of certain oils, or rather commercial brands. As pointed out by Dunstan and Thole 33 "It appears beyond doubt that the high boiling fractions of petroleum, irrespective of their place of origin, are complex mixtures containing a very small percentage of paraffine hydrocarbons of the formula C n H 2n 2 , and con- sisting chiefly of compounds whose formula} range from C n H 2n to C n H 2n - 8 ." This is true even of so-called paraffine base oils of the light Pennsylvania type. The nature of the unsaturated hydrocarbons in lubricating oils is an open question. They show large losses on treating with con- centrated sulfuric acid, these losses amounting to 20 to 40 per cent; they show indefinite iodine numbers but attempts to hydrogenate such oils have been negative. Treatment with liquid sulfur dioxide results in a partial separation. Thus, an oil showing an iodine number of 46 gave, after extraction, a residue of iodine number 33 and an ex- tracted portion having an iodine number of 73. Somewhat similar results are effected by fullers' earth. As previously pointed out, the usual methods of determining iodine numbers are of little value for lubricating oils. Thus Dunstan and Thole state that the reaction of mineral oils toward iodine differs profoundly from that of fatty oils. According to their experience with the Wijs reagent, by varying the time and proportion of iodine chloride, a given mineral oil may yield widely varying values. For example, a California mineral oil gave a value of 20 in 2 hours, 40 in 4 hours, 60 in 64 hours and 80 in 266 hours, whereas rape oil reached a steady value in three minutes. Again, the iodine value of rape oil was found to be practically independent of the amount of Wijs' solu- tion used (provided a fair excess was employed) but with a mineral lubricating oil an increase in the proportion of reagent to oil invariably augments the iodine value. Highly refined oils such as white pharmaceutical oil are inferior in viscosity and lubricating value to those oils which are less highly refined and which contain a certain proportion of so-called unsatu- rated hydrocarbons. M J. Inat. Petr. Techn. 7, 417 (1921). Chapter XV. Rearrangements Rearrangement of carbocyclic structures to substances having a different number of carbon atoms in the ring is occasionally observed in the case of hydrocarbons but such rearrangements are much more frequently noted with derivatives. On strongly heating cyclohexane under pressure, conversion to methylcyclopentane occurs and Markownikow x showed that when cyclohexyl chloride or iodide is heated with concentrated hydriodic acid methyl cyclopentane is formed. Pure cyclohexane, however, is not effected by heating with hydriodic acid. The conversion of cyclo- heptyl iodide into methylcyclohexane and dimethylcyclopentane is a reaction of very much the same kind. Cyclobutylcarbinol and hydro- gen bromide results in conversion of the four carbon to the five carbon ring, namely, to cyclopentylbromide. Change of cyclobutane derivatives to cyclopentane derivatives was noted by Kishner 2 on heating dimethyl or diethylcyclobutyl carbinol with oxalic acid. CH 2 J " 1 -5 C.H. C 2 H 5 / CH 2 C< CH 2 CH C OH / C,H \ + HI -H> iodide + KOH -nCH 2 C 2 H 6 \ [ 2 -CH 2 CH = *Ann. 302, 1 (1898). '/. Rugs. Phys.-Uhem. Soc. J t 2, 1211 (1910) ; 45, 1149 (1911). 524 REARRANGEMENTS 525 CH 2 C< CKL CH 2 OH HBr. CH 2 CH 2 The change of the four carbon ring in pinene to a five carbon ring by the action of dry hydrogen chloride forming bornyl chloride is a reaction the mechanism of which is obscure, but is, nevertheless, an illustration of the tendency of four carbon atom rings to rupture or to change to five carbon atom rings. Demjanow 3 discovered that cyclobutylmethylamine is converted into cyclopentanol by the action of nitrous acid: H C CHNH.CH, H 2 C CH 2 HONO H 2 C CH, H,C C HOH \ C H Cyclobutylamine and nitrous acid, however, yield a mixture of eyglobutanol and cyclopropylcarbinol. H 2 C - CHNH 2 + HONO - > HC H. CH, \ CH 2 OH. \ H. Cyclopentylmethylamine and nitrous acid readily yields cyclo- hexanol and in the same way cyclohexylmethylamine is converted into cyoloheptanol.* Conversion of the cyclopentane ring to the cyclobutane ring has been noted by Rosanov, cyclopentyl nitrite being converted by the action of concentrated alkali into nitromethylcyclobutane. 5 CH, CH, H CH 2 C< CH 3 H, CH N(X J. Chem. Soc. 1910, I, 838. Wallach, Ann. 353, 331 (1907). J. Chem. Soc. Aba. 1915, I, 657. 526 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Dehydration of cyclopentylcarbinol by oxalic acid yields cyclo- hexene, which Rosanov considers is formed with the intermediate formation of the bicylic hydrocarbon, K -CH CH Probably influenced by his studies of the reactions of sabinene, which contains a three carbon ring, Wallach has offered an explana- tion of these reactions based upon the intermediate formation of a cyclopropane ring 'structure as indicated by the following : CH 2 CH, CH 9 CH CH.CH 9 NK /H 2 H0 CH CH, CH 2 CHOH CH, CH, It is curious that when treated with hydrogen bromide, the cyclo- propane ring in sabinane is broken in such a way that the five carbon ring, not the six carbon ring, is preserved. 6 HB. Tiffeneau 7 observed change of the cyclohexane ring to the cyclo- pentane ring, when 2-iodocyclohexanol is treated with silver nitrate. Kisbner, J. Russ. Phys.-Cliem. Soc. 43. 1157 (1911). "'Cornet, rend. 159, 771 (1914J. REARRANGEMENTS 627 CH 3 HO When cyclic a-monochloroketones are treated with alcoholic caus- tic potash, cyclic acids result in which the number of carbon atoms in the ring is reduced by one. 8 Thus 2-chlorocylohexanone gives cyclopentanecarboxylic acid, and 4-chloro-l-methylcyclohexane-3- one yields methylcyclopentane-3-carboxylic acid. Favorski's experi- mental work does not show that the carbon atom to which the chlo- rine is attached is the one which becomes the carboxyl group. Wal- lach's theory of the intermediate formation of a bicyclic compound is applicable to this case, and explains the function of the alkali which is necessary to effect the change, CH CH CHC1 CH CH C = CH 2 CH 2 CH CH 2 CH "C = O H0 CH 2 H 2 ^ 1 CH, CH CH, COOH. Wallach has reviewed 9 the rearrangement of dibromocyclic ketones, particularly cyclohexanones, by alkali to hydroxy acids of one less number of carbon atoms in the ring. As a rule the two halogen atoms are substituted not on the same carbon atom as > CBr 2 but each halogen replaces a hydrogen atom of the two adjacent carbon atoms. Wallach assumes the intermediate formation of a three- carbon ring derivative leading to the formation of an ortho-diketone. Such diketones when isolated appear to have changed to the ketol, like buchu camphor. Both menthone and carvomenthone give dibro- mides which yield the same cyclopentane carboxylic acid derivative, which Wallach explains as follows, 'Favorski & Boshovski, J. Bus*. PJiys.-Chem. 8oc. 46, 1097 (1914). Ann. 414, 296 (1918). 528 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS In a study of the Wagner rearrangement Ruzicka 10 obtained evi- dence which strongly supports the theory of the intermediate forma- tion of cyclopropane derivatives in such rearrangements. As regards the intermediate cyclopropane theory versus the theory of dissociation to bivalent carbon Ruzicka finds that rearrangement takes place with tertiary alcohols, such as methyl borneol and methyl fenchyl alcohol, which could not yield bivalent carbon directly by loss of water. The products of dehydration of methylborneol and methylfenchyl alcohol are identical, which fact Ruzicka considers to be in confirmation of the tricyclene or cyclopropane theory. Meerwein 11 showed that l-isopropylcyclopentane-1.6-diol is con- verted by the pinacoline rearrangement to the six carbon ring 2.2- dimethy 1 cyclohexanone : CH 2 CH 2 CH 2 CH 2 >c-c< CH a CH, H CH 2 CH 2 CO I I CH 3 CH 2 CH 2 C< CH 3 Further work showed that apparently, (1) By pinacoline rearrangement no intermediate products of the trimethylene or ethylene oxide type could be detected or isolated. "Helv. Chim. Acta. I. 110 (1918). "Ann. S76, 152 (1910). REARRANGEMENTS 529 (2) The behavior of the cyclic pinacones on rearrangement is essentially a special case of the general rules, holding good also with acyclic pinacones. (3) The course of the pinacoline rearrangements is determined by different factors according to the structure of the pinacone. In the symmetrical type \ / C C /I |\ R OH OH R x the rearrangement is determined by the ease of "migration" of the groups R and R x . With those of the unsymmetrical type, R R " -' ""'' \-c R OH OH "R 1 the relative stabilities of the two hydroxyl groups are more im- portant. Meerwein made the diethyl and diphenyl derivatives correspond- ing to the above from a-oxycyclopentane carbonic ester, CH 2 CH 2 CH 2 CH 2 R \C C0 2 CH 3 + 2RMgBr -* \C C < CH 2 CH 2j CH -CH 2()H OH The diethyl derivative yields the two pinacolines, 2 . 2-diethylcy clo- hexanone and 1 . 1-ethylpropionylcyclopentane, CH 2 the latter in largest amount. n 2 L/xlj (u O _ o CH 2 C C CA H 2 C 2 H 5 CH 2 CH 2 1 > .CH 2 CH 2 C 2 H 5 ^" ^. 530 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The diphenyl derivative, however, yields the six ring ketone quan- titatively, C*o m T c< OH OH -/ to The as-diphenylpentamethylene glycol, however, does not rear- range to the cycloheptanone derivative (as might be expected from Baeyer's theory) but yields chiefly the oxide. \ OH OH The dimethylpentamethylene glycol gives the two ketones, CH OH OH This heptanone derivative was overlooked by Tarbouriech. 12 Rearrangement of five carbon rings to six carbon rings by pina- coline rearrangement has been observed previously. Klinger and Lonnes 13 reduced fluorenone with zinc dust and acetyl chloride ob- taining not the expected glycol or pinacone but the ketone, C.H CO c f/C.H 4 \< "Compt. rend. 11$, 605, 863 (1909) ; 150, 1606 (1910). er. 29, 2154 (1896). REARRANGEMENTS 531 Meerwein shows that the similarly constituted os-dimethyl and as-diethyldiphenylene glycols yield exclusively the normal pinacoline, without change of the five carbon ring, e. g., The diphenyl derivative is converted into the six carbon ring of phenanthrene. The same explanation, pinacoline rearrangement, explains the result noted by Klinger and Lonnes on oxidizing diphenyldiphenylene- ethylene. CTT f^ TT 6^4 ^6^5 C 6 H 5 C 6 H 4 COC 6 H a V^gO-A^ | C 6 H 4 Theory of the Pinacoline Rearrangement: The assumption of in- termediate formation of a triatomic ring was made first by Erlen- meyer, 14 who supposed the formation of substituted ethylene oxides. Under special conditions, the formation of ethylene oxide derivatives can be shown in the case of benz-pinacone and sym.-diphenylditolyl- glycol: these particular oxides are converted into ketones on heating with dilute mineral acids. Against the general theory, Meerwein cites the well-known fact that the oxides are usually converted, by addition of water, to glycols under much milder conditions than the latter are converted into pinacones, recalling particularly the case of tetra- metfiylethyleneoxide which takes up water to form the glycol even in the absence of acids. The results of Tiffeneau 15 on the properties of substituted ethylene oxides also support, the views of Meer- wein. Meerwein assumes that in diethyltetramethyleneglycol both hydroxyls have approximately equal tendencies to split off as water, therefore, leading to the formation of the two ketones, thus, "Ber. 14. 322 (1881). (Cf. also Nef, Ann. 198, 148 [1879]). chim. phys. (8) 10, 346, 375 (1905). 532 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS In the conversion of borneol or isoborneol to camphene with loss of water the 1,2,3,4,8 pentamethylene ring is converted into a six carbon ring, and the original six carbon ring, 1 to 6, is converted into a five carbon ring. =CH, H, y H H Wagner 16 in his original camphene-borneol article called attention to changes of the type, CH 3 CH 3 C CH CH 3 CH 3 CH 3 CH 3 OH CH 3 CH 3 and for which Tiffeneau suggests the name "retropinacoline rear- rangement." 17 For comparison with borneol it is necessary to ascertain the dehy- dration behavior of, (1) Alcohols of the cyclohexane series, hydroxyl being in the ring. (2) Alcohols of the cyclopentane ring in which the hydroxyl group is in the side chain, for example, 18 J. Ruaa. Phys.-Chem. Soc. SI, 680 (1899). "Rev. gen. sci. 18, 583 (1907). REARRANGEMENTS 533 2 .2-dimethylcyclo- 1 .1-methyl-a- hexanol-1 oxethylcyclo- pentane In splitting off water from I, two products result, as indicated, about 75% ^-isopropylcyclopentene about 25% Instead of 1 . 1-methyl-ct-oxethylcyclopentane, Meerwein employed its derivative 3-isopropyl-l . 1-methyl-a-oxethylcyclopentane. 18 As in the first instance, the normal product of dehydration was not formed but a mixture as follows: ,CH, "Ann. Ifl5 t 129 (1914) . 2-dimethyl-4-isopropyl fc-cyclohexene principal product 534 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The rearrangement of 2.2-dimethylcyclohexanol-l to A 1 -isopro- pylcyclopentene, with decrease in the carbon ring, C 6 to C 5 , and also likewise the ring enlargement by rearrangement of 3-isopropyl-l . 1- methyl-a-oxethylcyclopentane to 1 . 2-dimethyl-4-isopropyl-A 1 -cyclo- hexene, C 5 to C 6 , is completely analogous to the rearrangement of borneol to camphene. This is made clearer by writing the change as follows: jG /H CH. II rl or or The conversion of borneol to camphene is practically a summary of the two reactions above. Clear explanation will be possible probably only when the mecha- nism of the change of pinacoline alcohol to tetramethylethylene is clear. Zelinsky and Zelikov, 19 like Nef, suppose the formation of a trimethylene ring. CH 3 CH 3 CH 3 >C CH CH 3 > >C = C< CH 3 \/ CH 3 CH 3 CH 2 This explanation has been given for the borneol-camphene change but as pointed out by Semmler 20 the hypothetical intermediate hydro- carbon 18 Ber. 84, 3251 (1901). 20 Ber. 85, 1018 (1902) ; Lipp, Ber. 53, 769 (1920), considers that the reactions of tricyclenic acid support Semmler's constitution for tricyclene. REARRANGEMENTS 535 is completely symmetrical and the resulting camphene should there- fore be optically inactive, which generally is not the case. Tiffeneau 21 has suggested that water is split off from pinacoline alcohols as follows: CH 3 H OH CH 3 \ \/ \ \/ CH 3 CH 3 CH 3 C C CH 3 - CH 3 C-C-CH 3 - >C = C< / / CH 3 CH 3 CH 3 CH 3 Meerwein 22 succeeded in making the desired 1 . 1-methyloxethyl- cyclopentane as follows: CH CH, CH, CH, Ci CH 3 acid chloride CH(OH)CH 3 By warming with ZnCl 2 water is readily removed. Of the three possibilities, CH 3 II would undoubtedly be rearranged to IV. III and IV are known. The substance obtained proved to be very 21 Rev. gen. act. 18, 583 (1907). 28 Ann. 417, 255 (1918). 536 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS nearly pure 1 . 2-dimethyl-A 1 -cyclohexene, or III, boiling-point 135- 20 20 137; d 0.8234; n 1.4566. Oxidation, gave 2,7-diketooctane, D CH 2 H, CR CH, COCH 3 COCH, Meerwein has obtained additional results, parallel to those pre- viously noted, in the case of 1,2,2,3-tetramethyl-l.a-oxethylcyclopen- tane, which gives, partly with ring enlargement and partly with a rearrangement of a methyl group, l^S^-pentamethyl-A^cyclo- hexene and l,2,2-trimethyl-3-isopropyl-A 3 -cyclopentene. CH 3 CH 3 \/ CH, CH C C CH, CH, CH, CH 3 CH C CH (OH) . CH, I >C< CH 2 CH \ CH CH 2 - CH CH 3 J! CH, \ CH 3 CH C CH 3 C CH CH 2 -CH CH 3 Wallach has described the conversion of a series of cyclohexanone derivatives into cyclopentanones. 23 When cyclohexanone is bromi- nated in acetic acid l,3-dibromohexane-2-one is formed, which, on treating with dilute aqueous caustic potash at room temperature, yields an acid derivative of cyclopentanone. On treating the latter with lead peroxide and sulphuric acid cyclopentanone itself is formed. 28 J. Chem. Soc. Ala. J916, I, 487. REARRANGEMENTS 537 Similar reactions were carried out with methylcyclohexanones and in the case of menthone, l-methyl-3-isopropylcyclopentane-2-one was formed. CH 3 CH CH 3 H 2 C CH 2 HoC CH H 9 C H,C = Chapter XVI. Physical Properties With the exception of a comparatively small number of hydro- carbons of the terpene series, the physical properties recorded in the literature of the non-benzenoid hydrocarbons differ so widely in each case that it is very difficult to draw any general conclusions of value from data at present available. Simple derivatives of the hydro- carbons are frequently known in much purer condition and the physi- cal properties determined with much greater accuracy than in the case of the parent hydrocarbons themselves. The tables of physical properties of the simpler paraffine hydrocarbons shown in the accom- panying tables show the wide disagreement in physical constants of these simple hydrocarbons. The figure given in the literature for the melting-point of normal octane is 98.2 but recent determinations by Forcrand * show that the melting-point of normal octane is 57.4. Very few of the unsaturated hydrocarbons derived from paraffine hydrocarbons are known in a state of purity and the constitution of many of them are still in doubt. Most of the individuals of this series described thus far are very evidently mixtures of two or more hydrocarbons. Density and Molecular Volume: The density of individual hydro- carbons is frequently given at temperatures other than zero degrees, 4, 15, 17% or 20 and it is frequently desirable to recalculate the density from the temperature stated to some standard temperature, usually zero or 20, for the purpose of comparison. Walden has shown that the increase in molecular volume V M for each degree Centigrade is about 0.11 per cent 2 the molecular volume V M being equal to the molecular weight divided by the density. The molecular volumes of the normal paraffines at show an average increment for CH 2 of 15.9. 3 l Compt. rend. 172, 31 (1921). 2 Z. physik. Chem. 65, 158 (1909). * Kauffmann, Beziehungen zwischen physikalische Eigenschaften u. chemische Con- stitution, 1920, p. 60. 538 PHYSICAL PROPERTIES Hydrocarbon V A Pentane ............................... 15.4 Hexane ................................ 1275 15.8 Heptane ............................... 143.0 Octane ................................ 158.9 16.0 Nonane ............................... 174.9 15.9 Decane ................................ 1905 16.0 Undecane ............................. 2065 15.6 Dodecane 4 ............................ 222.4 16.5 Tridecane .. ........................... 238.9 15.8 Tetradecane 4 .......................... 254.7 165 Pentadecane ....... . ................... 27L2 Hexadecane ........................... 2875 The molecular volumes of isomeric hydrocarbons show slight dif- ferences, for example, v *l V M C 5 Hu n.pentane ' .............................. 115.2 " isopentane 8 ............................. 116.4 CaHu n.hexane 6 ............................... 130.5 " 2-methylpentane 7 ........................ 131.1 " 2.2-dimethylbutane T ..................... 132.8 " 2.3-dimethylbutane 7 ..................... " 3-methylpentane 8 ....................... 129.1 The octanes show only very slight differences in molecular volume, the maximum being that of 2 . 5-dimethy Ihexane and the minimum that of 3 . 4-dimethy Ihexane, 9 V 15 V M~ maximum ........................ 163.5 minimum ........................ 157.2 n. octane ......................... 161.7 20 The monochlorohexanes also show only slight differences for V - Apparently the observed specific gravities of these two hydrocarbons at are too low by about 0.0009. 'Timmennans. Chem. Zentr. 1912 (2), 472. Auwers & Eisenlohr, Z. physik. Client. 83, 431 (1913). 'Kishner, J. Russ. Phys.-Chem. 8oc. tf, 595 (1911) ; 47, 1111 (1915). "Kishner, J. Rus*. Phys.-Chem. Soc. 45, 973 (1913). Clarke, 7. Am. Chem. Soc. S3, 520 (1911) ; 34, 170, 674 (1912). 40 CHEMISTRY OF THE NON-BENZENOlD HYDROCARBONS Maximum, 2-chloro-2-methylpentane ................... 139.7 Minimum, 3-chloro-3-methylpentane, ................... 136.4 n.hexyl chloride ........................... 137.7 The molecular volumes and densities of the hydrocarbons of the ethane and propane series have been determined by Maas and Wright. 10 R p Difference D.-r. A T/ Hydrocarbon C d R V V M V f V' M $ Experimental Calcul. Ethane 883 05459 5495 55 0044 Propane ... 44.5 0.5853 75.2 77 18 0033 Ethylene ... 103.9 0.5699 491 44 + 51 0045 Propylene ... _47.0 0.6095 68.9 66 + 2.9 .0034 Acetylene ... 83.6 0.6208 41.9 33 + 8.9 .0046 Allylene . 27.5 0.6785 59.0 55 + 4.0 .0027 dg densities at the boiling-points. V,, molecular volumes calculated from d R V'-.. molecular volumes calculated on the basis C = 2H = 11. AV ___r=the tempertaure coefficients of the specific volume at the AV boiling-point. The values -- are dependent upon the critical tem- peratures of the hydrocarbons, which fact makes it possible to calculate the specific volumes of any of the hydrocarbons at any temperature, provided the specific volume at any one temperature is known and the specific volumes for one of the other hydrocarbons is known at all temperatures. This was one of the deductions made by van der Waals from his equation of corresponding states, namely that where Vj. and V 2 are the specific volumes of one liquid and Vj and v 2 are the specific volumes of another liquid where V v and V 2 v 2 are measured at the same corresponding temperatures. Taking propylene as a standard Maas and Wright calculated the specific volumes of ethane, ethylene, acetylene, propane and allylene at a temperature 30 higher. The greatest observed discrepancy between the calculated and experimental values was only 0.3%, in the case of propane, the other cases being within the experimental error of 0.1%. 10 J. Am. Chem. 8oc. 1$, 1105 (1921). PHYSICAL PROPERTIES 541 Ring closing usually has a greater effect on the molecular volume than a double bond or branched chain structure, as compared with normal carbon chain structures. 121 V M CaHw, n.hexane ...................................... 130.5 CeH 12 , a-hexene 11 .................................... 126.0 ft-hexene ..................................... 123.3 cyclohexane ................................... 108.1 propylallyl ether ............................... 125.1 -hexylene oxide 1 Clt-CH^CH-CHal ....................... 1146 CEk CH, allyl ether ..................................... 119.3 cyclohexanone ................................. 101.9 methylbutyl ether ............................. 118.4 pentamethylene oxide ......................... 97.4 Cyclohexane is thus seen to have a greater density than n.hexene but the true effect of ring closing on the molecular volume is realized by correcting for the volume of the two hydrogen atoms difference between C 6 H 14 and C 6 H 12 . When the value 32.05 for these two hy- drogen atoms is substracted from the molecular volume of hexane, 130.5 32.05 = 98.4, the result is lower than the observed value of cyclohexane by 9.7 units. The molecular volume of the saturated cyclic hydrocarbons may be calculated in another manner, i. e., by 20 multiplying the value for CH 2 at 20, V -= 16.27, by the number of such groups in the hydrocarbon. The results show that in all cases the molecular volume of the cyclic hydrocarbons is materially greater than the values corresponding to the number of CH 2 groups present. The latter method is designated as II and the former, deducting 32.05 20 from the V-^-oi the normal hydrocarbons, is method I in the follow- M ing table. T ,20 , Excess of observed HYDROCARBON" V^-Cobs.) oyer ca i cu i ate d values / 11 Cyclobutane, C 4 H ..................... 81.7 16.6 Cyclopentane, C 8 H W .................... 94.1 10.9 12.7 Ccyclohexane, CeH^ ........... . ........ 108.1 9.7 10.5 Cycloheptane, C 7 H 14 .................... 121.1 6.6 72 Cyclooctane, C 8 H 16 ...................... 133.7 3.1 3.5 Cyclononane, C,^ ..................... 163.9 17.4 17.5 "v. Braun, Ann. 382, 22 (1911). 12 Cf. values given by Willstatter, Ber. 40, 3988 (1907) ; p, 1483 (1908) ; 4S f 1182 (1910). 542 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS It might have been expected from Baeyer's strain theory that the closest agreement between the calculated and observed molecular volumes would be in the case of cyclopentane and cyclohexane, instead of cyclooctane. The same order of differences are observed in the case of the ethyl derivatives. Effect of ring closing; excess of observed 20 20 Hydrocarbon V -^- over calc.V jj- Ethyl cyclopropane 13 103.3 20.2 Ethyl cyclobutane " 115.5 17.1 Ethyl cyclohexane 15 143.1 12.5 Ethyl cycloheptane 16 154.8 8.2 The ketone derivatives show the same order of differences, the molecular volume at 20 having a maximum difference from the cal- culated value in the case of cyclobutanone, minimum in the case of cyclooctanone and again a large difference in the case of cyclonona- none. The effect of ring closing in the case of 3 . 3-dimethylbicy clohex- ane, 17 considered as derived from 0em.dimethylhexane is indicated by 20 a difference of 26.9 units. V 138.3. M T. W. Richards regards the relations between boiling-point and density as a natural corollary of the theory of atomic compressibility or deformability. "Thus, as regards two substances otherwise similar, the less volatile one would be less compressible, denser and possess greater surface tension. These outcomes of the theory correspond with the facts in a majority of cases thus far studied; for example, o-xylene is denser, less volatile, less compressible and possesses a greater surface tension than either m-xylene or p-xylene." 18 Tyrer has re-examined Trouton's rule and states that the relation between the molecular volume and boiling-point may be better ex- pressed by a modification of Trouton's rule, which Tyrer 19 formu- lates as follows, "Zelinsky, B&r. 46, 170 (1913). "Kishner, /. Russ. Phys.-Chcm. Soc. 45, 973 (1913). '"Lebedew, J. Russ. Phys.-Chem. Soc. 43, 1124 (1911). 18 Markownikow, Ann. 327, 73 (1903). "Zelinsky, Ber. 46, 1466 (1913). In his monograph, Kauffmann makes use of Zelmskys supposed spirocyclane, which Philipow, J. prakt. Chem. 93, 162 (1916), has shown to be a mixture of methylcyclobutene and methenecyclobutane ; Kauffmann's inferences are accordingly incorrect "/. Ohem. Soc. 99, 1211 (1911). *PW. Mag. (6) W, 522 (1910). PHYSICAL PROPERTIES 543 in which the constant K is 68 for the aliphatic hydrocarbons and ethers, 70 for alkyl chlorides and amines, 74 for the fatty acid esters and bromides and 79 for aliphatic iodides and aromatic hydrocarbons. Like most such rules the constant is subject to considerable variation, for example, Substance T V M K Methyl bromide . 286. 58.2 73.8 Ethyl bromide 312.7 77.7 73.0 Propyl bromide 344. 972 74.8 Isopropyl bromide 333. 97.2 71.9 Ethyl chloride 285.2 71.2 68.8 Propyl chloride 319.2 91.4 70.9 Isopropyl chloride 309.5 93.6 68.2 Chloroform 334.1 &4.5 76.1 Carbon tetrachloride 349.7 103.7 74.3 Melting-Point. Ring closing has a much more pronounced effect upon the melting- point. Langmuir comments upon the fact that the physical prop- erties of nitrous oxide are practically identical with those of carbon dioxide at a temperature 3 lower, but that the freezing points of these two substances are in marked contrast to the general agreement, being 102 for nitrous oxide and 56 for carbon dioxide. He states that "This fact may be taken as an indication that the freezing- point is a property which is abnormally sensitive to even slight dif- ferences in structure. The evidences seem to indicate that the mole- cule of carbon dioxide is more symmetrical, and has a slightly weaker external field of force than that of nitrous oxide." Organic chemists, however, in studying the relations between structure and physical properties, have paid much greater attention to boiling-points, specific gravities and optical properties. Probably on account of the fact that, as Langmuir points out, the freezing-point is so sensitive to dif- ferences in molecular structure, it is very difficult to trace simple rela- tionships or make any useful generalizations. Also, since most of the non-benzenoid hydrocarbons and their simple derivatives, of which we have fairly complete knowledge, are oils at ordinary temperatures, data as to their freezing-points are usually lacking. Most of the non-benzenoid hydrocarbons which are solid at ordi- nary temperatures are normal paraffines, and of these none of their many possible isomers are known, so that no data exist from which 544 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS to draw any conclusions as to the effect of variations in structure upon the melting-point. The melting-points of the normal paraffines are given in the fol- lowing table, in which it will be noticed that the increment for CH 2 is irregular in the lower part of the series but becomes more regular and smaller with increase in molecular weight. The melting-point of n.heptane given is probably too low. Melting-Point Hydrocarbon 20 CH 4 184. C 2 H a 172. v^/sxis " " lot/. C 4 H 10 135. 5^ 130.8 C 6 Hi 4 94.3 C T H M 97.1 CsHw 57.4 /~l TT __ r-i CioH^ '. '.'.'.'.V.'.'.'.V.'.V .'.'.'.'.'. 32! CnH 2 26.5 Ci2.H2a CisHzs C 1B H 32 Cl6ll3 4 CniM CasHw C24H> 12. 6.2 + 5.5 10. 18. 22.5 28. 32. 36.7 40.4 44.4 47.7 51.1 68.1 70.5 Boiling-Point C 164. 84.1 44.5 0.1 + 36.2 68.9 98.4 125.8 149.5 173. 194.5 214.5 234. 252.5 270.5 287.5 303. 317. 330. According to Tsakalotos 21 the curve connecting the melting-points of the normal paraffines is fairly regular and smooth from about C 16 H 34 and upwards in the series, and agree well with the values calculated from the formula An = 85-0.01882 (n-1) where An is the difference between the melting-point of one member and the next highest in the series and n is the number of carbon atoms in the molecule of the hydrocarbon the lower in the series (of the pair) . 20 Methane, Moissan & Chavanne, Compt. rend. 1$, 409 (1905) ; ethane, Ladenburg, Ber. 33, 638 (1900) ; propane, Maas & Wright, J. Am. Chem. Soc. $, 1100 (1921) ; C 4 Hi to CgHig, Timmermans, Chem. Zentr. 1911 (2), 1015; CgHao et. sea. Krafft. Ber. 15 1687 (1882) ; 19, 2218 (1886) ; 21, 2256 (1888). 21 Compt. rend. 1*3, 1235 (1906); Forcrand [Compt. rend. 172, 31 (1921] states that the simpler normal paraffines, and the cyclic hydrocarbons C 8 to C follow the rule of the alternance of melting-points. PHYSICAL PROPERTIES 545 The substitution of chlorine causes marked rise in the melting- point. 22 Melting-Point Boiling-Point Substance C C CH 4 ....................... 184. 164. CHaCl ..................... 103.6 23.4 CH 2 Cla .................... 96.7 +41.6 CHCU ..................... 63.3 61.2 CC1 4 ....................... 22.95 76.7 CH 3 CH 3 ................... 172.1 84.1 CHaCaCl ................. 138.7 +12.5 CHaCHCl, ............. .... 96.7 +57.3 CH 2 C1.CH.C1 2 ............. 35.5 CHC1 2 .CC1 3 ................ 29.0 +161.9 +187. +189.6 Unsaturation usually causes a marked rise in melting-point, 23 in the case of hydrocarbons. Ethane .............. 172.1 Propane .............. 189.9 Ethylene ............. 169.4 Propylene ............ 1855 Acetylene ............ 81.8 Allylene ............. 104.7 It is of interest to note the very great differences in the melting- points of the following pairs of isomers, noting that the differences in boiling-point are by no means so large. Melting- Boiling- Point Point n.pentane, CH 3 (CH 2 ) 3 CH 3 24 ............... 130.8 +365 tetramethylmethane, (CHa^C ............ 20. +9.5 n. octane, CH^CH^CIk 26 ................. 57.4 125.8 2.2.3.3. tetramethylbutane M ............... + 103.-104. 106.-107. (CH 3 ) 3 C.C(CH 3 ) 3 The effect of ring closing upon the melting-point is to raise it and, as will be noticed above, the boiling-point is also raised. Melting-Point Melting-Point n.hexane ............ 93.5 n. octane ............. 985 cyclohexane .......... + 6.4 cyclooctane .......... + 11.5 22 Timmermans, loc. cit. 23 Maas & Wright, J. Am. Chem. Soc. 43, 1100 (1921). 2 * Timmermans, Chem. Zentr. 1911 (2), 1015. 26 Forcrand, Compt. rend. 172, 31 (1921). CH, 26 The similarly constituted undecane, (CH 3 ) 3 C C C)CH 3 ) 3 , and tetradecane, CH 8 CH 3 CH 3 (CH 8 ) 3 C C C C(CH 3 ) S are not known but, by analogy, their melting-points CH 8 CH 3 would be much higher than n.undecane and n.tetradecane. CH 3 S>r CH.CH 3 p*^ CH - L Br CH 3 ^>O CH.CH 2 CH ^^> CH 3 | L CH 3 CH 3 546 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The very marked effect of structural differences on the melting- point is well illustrated by the dibromohexanes. Increase in the num- ber of methyl groups, and proximity of these methyl groups to the bromine or other negative substituent raises the melting-point as compared with isomeric derivatives, for example, CH 3 CH . CH 2 CH 2 CH CH 3 melting-point 38.2 Br Br (n.isomers are liquid) melting-point 7 liquid >C CH< melting-point 24-25 CH 3 | CH 3 CH 3 CH 3 >C - C< melting-point 169-170 CH 3 | | CH 3 corresponding dichloro, Br Br melting-pointlGO CH 3 \ melting-point 187 CH 3 C.CBr 2 CH 3 corresponding dichloro, / melting-point 151 CH 3 Nitro groups similarly placed result in crystalline derivatives, for example, CH 8 N0 2 CH 8 C C CH 3 melting-point 173-174 CH 3 N0 2 CH 3 CH 3 >C - C< melting-point 213-214 CH 3 | | CH 3 N0 2 N0 2 PHYSICAL PROPERTIES 547 The melting-point of the latter substance is higher than that of any of the dinitro or trinitro benzenes. Dinitroheptane and octane of sim- ilar structures are also well crystallized substances, CH 3 CH 3 > C CH 2 C< melting-point 81-82 CH 3 | | CH 3 N0 2 N0 2 CH 3 CH 3 >C CH 2 . CH 2 C< melting-point 124-125 CH 3 | | CH 3 N0 2 N0 2 Closing of the ring raises the boiling-point slightly, as a com- parison of butane and cyclobutane and a number of their derivatives indicates, Boiling- Boiling- Difjer- n. butane series Point C Cyclobutane series Point C enceC n. butane 0.1 c. butane +11, +11- C 4 H 9 C1 +77. C 4 H T .C1 85. +8. CJLNH, 76. C^T.NH, 81. +5. C 4 H 8 .Br 100. C.HT.Br 104. +4. C 4 H 9 .OH 116. C 4 H 7 .OH 122.5 +6.5 C 4 HJ 131. C 4 H T .I 138. +7. 186. CiHi.COaH 195. +9. When the two series of hydrocarbons of three to eight carbon atoms are compared, the cyclic series is seen to have consistently higher boiling-points. Boiling-points of cyclic and normal hydrocarbons. Boiling- Boiling- Differ- Normal Point C Cyclic Point C enceC Propane 44.5 Cyclopropane 35. + 9.5 Butane 0.1 Cyclobutane +11- + 10-5 Pentane 36J2 Cyclopentane 49. + 12.8 Hexane 68.9 Cyclohexane 81. +12.1 Heptane" 98.8 Cycloheptane 117.2+.2 Octane 1253 Cyclo-octane 145. 2 +19.4 Among observations of the boiling-point a few qualitative gen- eralizations can be made, for example, Wallach 28 has noted that in a series of isomeric ketones the widest separation of the ketone group and the alkyl side chain gives the highest boiling-point. "Forcrand, Compt. rend, m, 31 (1921). 38 Ann. 391, 183 (1913). 548 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The effect of unsaturation on the boiling-point and critical tem- perature in the ethane and propane series has recently been shown by Maas and Wright. 29 The figures given in the fourth column show that the boiling-points are approximately equal fractions of the critical temperatures, when figured as degrees absolute. Mol. Latent Boiling- Hydrocarbons Point C fj ea t Boiling-Point t Evaporation t C C. Ethane . ... 88.3 35.0 3800 0.60 Ethylene 103.9 9.9 3510 0.60 Acetylene . 83.6 36.5 5150 0.61 Propane 44.5 95.6 4500 0.62 Propylene 47.0 92.1 4600 0.62 Allylene 27.5 127.9 5230 0.61 The critical values of a number of hydrocarbons of the methane series are given by Young. 33 Crit. Pressure Grit. Temp. Grit. Density Hydrocarbon mm.Hg. C Density atO/4 n Pentane . 25100 197.2 0.2323 0.64536 Isopentane 25018 187.8 0.2343 0.63927 n Hexane 22510 234.8 0.2344 0.67703 Diisopropyl 23360 227.3 0.2411 0.67948 n Heptane . 20430 266.8 0.2341 0.70048 n' Octane . 18730 296.2 0.2327 0.71854 Diisobutyl 18660 276.8 0.2366 0.71021 Cyclohexane 30278 280. 0.2735 0.79675 Benzene 36395 288.5 0.3045 0.90006 Absorption of Light; Color: All saturated hydrocarbons are color- less but show selective absorption in the infra red. 30 Examination of hexane, cyclohexane and camphane shows no well defined absorption band in the ultraviolet part of the spectrum, of wave lengths greater than 185 \I\L, but ring formation seems to cause a shift of general absorption toward the longer waves. Certain recent works 31 state that unsaturated hydrocarbons show no selective absorption in the ultraviolet but, on the contrary, ethylenic hydrocarbons show definite absorption bands and, in the case of the doubly conjugated fulvenes, absorption bands occur in the visible part of the spectrum, these hydrocarbons being colored yellow to orange. The aliphatic defines isobutylene, trimethylethylene, hexylene and octylene show two ab- sorption bands, at A230-X205 and at about A180. 32 The presence of 29 Loc. cit. * a 8ri. Proc. Roy. Soc. Dublin 12, 374 (1910). " Coblentz, Jahrb. Radioakt. k, 7 (1908). 81 Watson, Color in Relation to Chemical Composition, 1918, p. 66. 82 Stark, Steubing, Enklaar & Lipp, J. Chem. Soc. Abs. 1913, II, 363. PHYSICAL PROPERTIES 549 two double bonds, not in conjugated positions, causes an intensifica- tion of the two absorption bands observed in the case of the singly unsaturated hydrocarbons, as in diallyl and geraniolene. Two con- jugated double bonds as in isoprene, 2-3-dimethylbutadiene, and hexa- diene-(1.4) show a shift in the position of the two bands of about 20 to 30 jxji, toward the visible spectrum, and an intensification of both bands. Camphene and a-pinene show an absorption band at X204-U98. In limonene and sylvestrene the head of the absorption band is about X185 but in the case of a and (3-phellandrene two bands are clearly developed. Fulvene is a yellow oil and dimethylfulvene, CH=CH CH 3 >C = C< = CH CH 3 WJLJ. in shows three well developed absorption bands with heads at X370, X258 and X207 respectively. A hydrocarbon of the empirical formula C 15 H 22 and containing four double bonds has been described by Sherndall 34 as having an intense blue color and accordingly named azulene. The hydrocarbon combines with eight atoms of hydrogen, in the presence of colloidal palladium, forming C 15 H 26 . The hydrocarbon is probably a tricyclic sesquiterpene perhaps identical with a-gurjunene. It readily com- bines with picric acid, forming black needles melting at 118. The intensity of the color is indicated by the fact that 0.064 g. of azulene in 1 liter of gasoline is matched in color by an ammoniacal copper sulfate solution containing 0.24 g. copper sulfate per liter. The very- exceptional color of azulene as compared with the fulvenes, renders the confirmation of Sherndall's work, and particularly the purity of the material employed, very desirable. The absorption spectra of cyclohexene and cyclohexadiene are of interest on account of the fact that they are very different from the absorption spectrum of benzene. The first two hydrocarbons show broad bands differing greatly from the groups of narrow bands of ben- zene and naphthalene. 35 Fluorescence: No non-benzenoid hydrocarbons are known which exhibit fluorescence. The marked fluorescence of petroleum distillates is undoubtedly due to traces of substances of the nature of chrysene, M J. Am. Chem. Soc. 37, 1537 (1915). "Stark & Levy, Jahrb. Radioakt. 10, 179 (1913) ; J. Chem. Soc. Abs. 1913, II, 366. 550 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS fluorene or pyrene, although the substances which cause fluorescence in these oils have never been isolated and identified. They are easily sulfonated and are, therefore, removed to a large extent on refining such oils with concentrated sulfuric acid, appearing as water-soluble sulfonic acids. It has been suggested by Schulz 36 and others that the blue fluorescence of petroleum distillates is due to colloidally dis- persed particles of carbon, sulfur or other material but oils which are carefully dried and filtered are optically homogenous under the ultramicroscope and the fluorescence is in no way affected by an electrostatic field. 37 The formation of such fluorescent material is almost universally observed when organic material is partially car- bonized by heat, even in the heating or "boiling" of linseed oil. The manner in which "deblooming" agents, such as nitrobenzene or nitro- naphthalene, suppress this property is not known but it is probably a purely physical phenomenon. 38 Ether, benzene and amyl alcohol intensify the fluorescence and aniline and carbon bisulfide suppress it, changing the blue fluorescence of petroleum oils to a dull faint green. Refractivity : Refractivity has been frequently employed as an aid in determining the constitution of organic compounds and the constants of a great many substances have been studied and correlated. Although the refractive indices of most organic substances fall within the range 1.30 to 1.70, most instruments of reputable make are accu- rate to the fourth decimal place and only a few drops of liquid sub- stance are required for the determination. However, in order that the molecular refractivity may be calculated, it is necessary to know the density. It is assumed that most of the readers of this volume are familiar with the refractometer and it will suffice to recall the two formulae for molecular refractivity which are in common use, the n 2 formula proposed by H. A. Lorentz and L. Lorenz 39 in 1880 being the one most frequently employed. (1) Derived from Gladstone and Dale M = - m. d (2) Lorenz and Lorentz, M= 2 .. The Lorenz and Lorentz formula is practically independent of tem- Petroleum, 5, 205. "Brooke & Bacon, J. Ind. d Eng. CTiem. 6, 623 (1914). "Kauffmann, Ann. 393, 1 (1912). "Wied. Annalen. 9, 641; u, 70 (1880). PHYSICAL PROPERTIES 551 perature and pressure. When the refractive index is accurately deter- mined to the fourth decimal place, the molecular refraction of sub- stances having molecular weights of about 100 will be accurate within zb 0.2. Gladstone and Dale 40 discovered that the refractivity of organic substances is modified by the manner of combination of their constituent atoms ; in other words, refractivity is a constitutive prop- erty. The study of refractivity with reference to chemical constitu- tion has been developed particularly by Briihl and Auwers. Although the values for the group CH 2 , obtained by Briihl and Conradi and Landolt, agree fairly close, Eisenlohr 41 has recalculated this value from data of 503 carefully purified substances, with the results shown in the following, for the sodium D line. Hydrocarbons Number of Substances 66 Mj, 4624 Aldehydes and ketones 92 4626 Acids . . 74 4613 Alcohols . 81 4634 Esters . ... 190 4605 mean 4.618 The values obtained by Conradi 42 and by Eisenlohr are as fol- lows: Element Na^ Na ^ (Conradi) (Eisenlohr) Carbon . 2.50 2.418 CH, 4.60 4.618 H 1.05 1.100 0, as in >C = 2.28 2511 O, as in ethers 1.68 1.643 O, as in OH 1.52 1.525 Cl 5.99 5.967 Br 8.92 8.865 I 14.12 13.900 Ethylene bond > C = C <, increment 1.733 Acetylene bond C = C , increment 2.398 The exaltation caused by an ethylene bond was computed from observations on the following hydrocarbons, "PMJ. Trans. 153, 323 (1863). 41 Z. physik. Chem. 75, 585 (1910). 42 Z. physik. Chem. 3, 210 (1889). 552 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Mol.Wt. Boiling-Point D M D I _ on Amylene" C 5 H 10 / ....... 70 ' 08 34 ' ^ ' 6476 1F 70 ' 08 367 ' 6664 24 ' 83 _ Amylene" / = ...... \ Hexylene" = ....... 84 ' 10 Hexylene" Octylene" C 8 H ia / = 40TT /=.. 140.2 154. 0.7720-^r 47.17 Decylene 44 doH 2 o/ Diallyl, d.limonene and sylvestrene were also used by Eisenlohr. Eijkman 46 estimates that the exaltation in refractivity of ethylene bonds increases with the number of radicals attached to the doubly linked carbon atoms, as follows, 47 One radical, RCH CH 2 ....................... 1.60 Two radicals, RCH = CHR ................... 1.75 Three radicals, R 2 C = CHR .................... 1.88 Four radicals R 2 C = CR 2 ....................... 2.00 Le Bas 48 has noted that usually introduction of methyl groups into ring structures produces exaltation, (1) In trimethylene ring =0.39 (2) In tetramethylene ring = 0.15 (3) In cyclopentane and cyclohexane ring = 0.15 In cases where two methyl groups are in the 1.1 or 1.2 positions this exaltation disappears. The values for nitrogen and sulfur also vary according to the manner of their combination. Nitrogen 49 Na D Sulfur n Na D Hydroxylamines .............. 2.48 Mercaptans ............... 7.69 Amines, primary .............. 2.45 Sulfides .................... 7.97 Amines, secondary ........... 2.65 Thiocyanates .............. 7.91 Amines, tertiary .............. 3.00 Bisulfides .................. 8.11 Nitrites, aliphatic ............. 3.05 Oximes, aliphatic ........ . ____ 3.93 Nitro Group," nitroparaffines ............ 6.72 Benzenoid derivatives ..... 7.30 Brtihl, Ann. 200, 181 (1880). "Landolt & Jahn, Z. physik. Chem. 10, 302 (1892). B Briihl, J. prakt. Chem. (2) 49, 241 (1894). M Chem. WeekUad. 3, 706 (1906). 47 This is by no means an infallible rule as cases are known in which substitution of a methyl group causes a decrease in refractivity ; for example, styrene and methyl- sty rene, cf. Auwers & Eisenlohr, J. prakt. Uhem. (2) 82, 65 (1910). " Trans. Faraday Soc. 13, 53 (1917). 'Briihl, Z. physik. Chem. 79, 1 (1912). >Bruhl, Z. physik. Chem. 25, 647 (1898). "Price and Twiss, J. Chem. Soc. 101, 1259 (1912). PHYSICAL PROPERTIES 553 Briihl 52 concludes that ring closing does not, of itself, affect the molecular refraction, except in the two types to which attention has already been called, i. e., ethylene and cyclopropane. The agreement between the calculated and observed molecular refractivities in the cyclic series is fairly close. 53 M M Observed Calculated Cyclobutane 18.22 18.41 Cyclopentane 23.09 23.01 Cyclohexane 27.67 27.62 Cycloheptane" 32.18 32.22 Cyclooctane 36.58 36.82 Cyclononane 42.36 . 41.61 Attention has already been called to the instability, or condition of stress, of cyclopropane. Ostling 55 has examined a large number of cyclopropane and cyclobutane derivatives. The following values are typical, and indicate that ring closing in the case of cyclopropane produces an exaltation of approximately 0.70 or a little less than one half that produced by a single ethylene bond. EXALTATION OF MOLECULAR REFRACTTVITY CAUSED BY RING CLOSING; CYCLOPROPANES. Formula Boiling-Point D l ~ x_/ A / CH . CHs nno CH a | 32.6- 33.2 0.6755^- 68 \CH.CH 3 2 >C CH.CHa 66 , 37.5 0.7052^ 0.92 H 2 4 CH 2 -17 co >CH.CH 2 OH 5T 123.3 0.8995^- 0.71 H 2 CH 2 1QO >CH.CO 2 H 56 183.2M84. C 1.0897 -^ 0.68 H 2 CH 2 170 >CHCHO 5T 98. (737mm.) 0.9294-nr 0.90 H a CH 2 CO 2 CH 3 15 7 >C< 196.6 1.1509-r^ 0.71 H 3 C0 2 CH 3 Sabinane 156.2-156.8 0.8142^r 0.70 Sabinene 58 163.-165. 0.8422^- 1.36 Carane 59 49.-50.(9mm.) 0.8381 |p 0.93 82 Cf. Ber. 25, 1952 (1892) ; 27, 1065 (1897). 53 Cf. Auwers, J. Chem. Soc. A6. 1918, II, 343; Ann. 422, 133 (1921). M WillstHtter & Kametaka, Ber. $J, 1483 (1908). 65 J. Chem. Soc. 101, 457 (1912). "Gustavson & Popper, J. prakt. CJicm. (2) 58, 458 (1898). "Demjanov & Fortunov, Ber. kO, 4397 (1907). "Auwers, Roth & Eisenlohr, Ann. S7S, 275 (1910). M Semmler & Feldstein, Ber. Jft, 384 (1914). 554 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The effect of a double bond in a conjugated position to the cyclo- propane ring produces increased exaltation, as shown by sabinene, and an aldehyde or ketone group adjacent to the cyclopropane ring also produces a definite, though slight increase in exaltation, as is shown by cyclopropyl formaldehyde, exaltation for M = 0.90. Of the following three ketones I and II show increased exaltation while III, in which the carbonyl group is not conjugated with reference to the cyclopropane group, shows the average exaltation. increment M = 0.89 D increment M = 1.03 D increment M = 0.70 or 0.76 D The hydrocarbon 1,2,3, nme%cyclopropane shows the abnor- mally large increment for M^ of 1.37 which harmonizes with the observations of Le Bas and Eijkman, noted above, as to the effect of methyl groups. Closing of the ring as in cyclopropane has no effect upon the molecular dispersion but the conjugation of a cyclo- propane ring and an ethylene bond causes an increase of approxi- mately 10 per cent. The increment in molecular refractivity produced by the cyclo- butane group is smaller and is influenced somewhat by substituent groups, as in the case of cyclopropane derivatives. PHYSICAL PROPERTIES 555 EXALTATION OF MOLECULAR REFRACTIVITY CAUSED BY RING CLOSING; CYCLOBUTANE SERIES. Formula Boiling-Point Incre- ment for M CH a -CH a TO 10.-11. 0.703fr CH 2 CH, CH a CH a w 98.5-99. 0.9381-jp- CH a C = CH a -CH,- [0.91592: CH2 "~ C \OH 122.5 084 15 o 0.9226^ CH 2 -CH 2 195. 1.0570^ PTT r/ H U11 3 v-^c0 2 H in Q CH a -CH, 157. 0.9525^ CH a CH.C0 2 C 2 H5 20 CH a -CH, 104.-105. 1.0456^- I /C0 2 C 2 H 5 12mm ' CHa ~ U \C0 2 C 2 H 8 17 3 CH 2 -CH.C0 2 C 2 H 5 " 114.5 1.1191-^- 20mm. CH a CH.C0 2 C 2 H 8 d.a pinene 156.4-156.6 0.8594^1 nopinone . 118.2 0.9827 43mm. 0.49 0.42 0.43 0.46 0.43 0.37 0.54 0.46 0.57 > WillstStter, Ber. 40, 3982 (1907). 81 Kishner, J. Russ. Phys.-CJiem. Soc. STt, 106 (1905). Briihl, Ber. S2, 1222 (1899). Zelinsky & Gutt, Ber. 40, 4744 (1907). "Demjanov & Doyarenko, J. Rues. Phys.-Chem. Soc. 4S, 835 (1911) ; Chem. Alts. 6, 478 (1912). 5 Auwers, Ann. S73, 274 (1910). 556 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS EXALTATION OF MOLECULAR REFRACTION CAUSED BY CONJUGATION OF Two ETHYLENE LINKINGS. M D EM D Substance Calc. Obs. Reference A^-dicyclohexene, C 12 H 16 /= 2 52.34 52.65 0.31 66 A'. 8 < 9 )-p-menthadiene, C 10 H 16 ,/= 2 45.24 46.0 0.76 67 A a - 8 ( 9 >-m-menthadiene, C 10 H 10 /= a 45.24 46.6 1.36 67 AW)- m -menthadiene, CioHie,/ 2 45.24 46.3 1.06 67 AVOO-O-menthadiene, CioHi ft / 2 45.24 46.0 0.76 68 V>H = C 3 C 10 Hi6>/= 2 45.24 45.69 0.45 69 CH, / < ^CH 2 C > CH a C CulW 49.86 50.13 0.27 69 N W \CH 3 CH, I /~ ^-CH^C CuH 18 /= 2 .. 49.86 50.39 0.54 69 X r \CH 3 iJW .... 49.86 49.97 0.11 69 \CH 3 In a recent paper Auwers 70 carefully reviews the effect of ring closing on the molecular refraction, particularly in the cyclohexane series on account of the evidence of refractivity as to the constitution of benzene. Although the refractivities of the saturated cyclopen- tanes, cyclohexanes and cycloheptanes are practically normal, it is noted that the expected exaltation of the molecular refraction nor- mally caused by conjugated double bonds is not observed in the case of cyclopentadiene, cyclohexadiene, and cycloheptadiene. Still greater differences are observed between the cyclic and acyclic conjugated Wallach, Goettingcn Nachr. October, 1910. 87 Haworth, Prekin & Wallach, J. Chcm. Soc. 99, 123 (1911). "sperkin, 8th Int. Cong. Appl. Ghem. VI, 244. 89 Haworth & Fyfe, J. Ghem. Soc. 105, 1662 (1914), ^5, 98 (1918), PHYSICAL PROPERTIES 557 trienes, the exaltation being particularly great for the acyclic hydrocar- bons but very slight in the case of the cyclic conjugated trienes. It was this fact which caused so much doubt and controversy over the con- stitution of A 1 - 3 -cyclohexadiene. The chemical evidence leaves no room for reasonable doubt regarding the constitution of this hydro- carbon and that it is by no means an exception will be seen from the following table, E M a and E M^ being the difference between the observed and calculated values for the a hydrogen and sodium D lines respectively. E M E M CH 2 CH 3 iJH_CH = i H +1.81 +2.10 CH - CH 2 -CH = C CH CHr=CH 0.45 0.47 72 CH CH 3 CH 3 CH CH = CH + 1.96 + 2.03 CH CH 2 CH 2 CH CH = CH + 0.02 + 0.05 73 CH CH 3 CH 3 CH CH = CH CH 2 ..... 1.62 74 CH CH 2 - CH, II I CH CH = CH CH 2 + 0.50 ..... The introduction of alkyl groups causes a definite exaltation of the molecular refraction as compared with the unsubstituted hydro- carbon. The differences in the following series of trienes are particularly noteworthy, Auwers representing benzene as cyclohexatriene, 71 Ber. 49, 833 (1916). "Auwers, Ber. 45, 3077 (1912). Harries, Ber. J,5, 809 (1912) ; Willstatter & Hatt, Ber. 45, 1647 (1912). 74 J. prakt. CJim. (2) 82, 74 (1910). 558 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS 1 D E M a EM CH = CH 2 CH 2 75 i.. CH CH +2.90 CH = CH CH 76 0.18 0.18 CH = CH 2 CH CH 3 .8- - CH 3 + 4.07 + 4.J CH = CH C CH 3 76 CH = CH C CH 3 +0.20 +0.19 M E M^ a D CH = CH 2 CH C 2 H 5 CH = CH C CH 3 77 +3.89 +4.19 + 0.05 CH = CH C C 2 H 5 CH = CH C CH 3 +0.06 CH 3 CH 3 \/ CH = C CH -CH, CH CH C CH,, 77 -1- 3.99 4-4.29 (allo-ocimene) CH 3 CH CH 3 CH^C CH CH = CH C CH 3 78 + 0.30 + 0.31 Just as a ketone or aldehyde group, in conjugated position with reference to a cyclopropane group, produces abnormally high refrac- "Auwers & Eisenlohr, J. prakt. Chem. (20) 84, 40 (1911). "Landolt & Jahn, Z. Physik. Chem. 10, 303 (1892). "Enklaar, Rec. trav. cMm. 36, 215 (1917). 'Landolt & Jahn, Z. physik. Chem. 10, 303 (1892). PHYSICAL PROPERTIES 559 tivity, a similar effect is produced by ketone and ethylene groups in conjugated positions, for example, M a E M x> Observed Calc. H Crotonaldehyde, CH 3 CH = CH.C 21.29 20.24 1.05 CH 3 Mesityl oxide (CH 3 ) 2 C = CH.C 30.13 29.39 0.74 Carvenone, 46.52 45.82 0.70 Menthenone, 46.78 45.82 0.96 Aliphatic conjugated dienes usually show exaltation, as in 2.4- hexadiene E M a = 0.98 isoprene, E M-Q =1.03 79 and hexatriene EM a =2.06. As noted above, unsaturated cyclic hydrocarbons containing two or more alkyl side chains and conjugated double bonds usually show exaltation, as in a-phellandrene and a-terpinene. 80 CH 3 CH 3 calc.= a 44.97 A "D obs. =45.35 calc. = 45.24 obs. =46.15 ) 3 H T C 3 H 7 a-phellandrene a-terpinene Opinions differ as to whether a study of refractivity has really contributed much to the elucidation of the constitution of substances such as a-terpinene, and as to whether or not the evidence of such physical constants can be relied upon. Usually, as in the case of the terpinenes, our most trustworthy evidence has resulted from Calculated from recent data of Harries, Per. yt, 1999 (1914), using !^! 0.6867 gives M D 25.38 ; calcul Auwers, Ber. 42, 2404, 2424 (1909). an< l a 0.6867 gives M D 25.38 ; calculated 24.35. 560 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS chemical investigation. 81 In this connection it should be noted that certain substances in which considerable exaltation might be expected, show only the normal refractivity, as, for example, cyclooctatetrene discovered by Willstatter. 82 20 /=4 I I n - 1.53460 calc. C 8 H 8 / 35.07 HC=:CH.CH=:CH D obs. 35.20 This hydrocarbon has all the chemical properties which one would expect such a substance to have and yet benzene, having quite dif- ferent chemical properties also shows only very slight exaltation, M _ observed 25.93, calculated for C 6 H 6 /= 3 26.25. Wallach has urged caution in interpreting refractivity values and conclusions thus drawn are of doubtful value unless supported by other good evidence and assurance that the substance examined is of the highest purity. Thus, methyl heptenone on condensation yields a material which for a long time was considered, from its analysis and refractivity value, to be dihydroxylene. This has been shown, however, to be a mixture of xylene and tetrahydroxylene. 83 Early in the study of refractivity Bruhl stated that the refractive index showed that the terpenes, C 10 H 16 , in orange peel oil, lemon oil and bergamot were identical and that the chemical evidence was merely confirmatory but not neces- sary to prove this fact. Wallach 8 * pointed out that on such grounds Bruhl should have claimed the identity of limonene and sylvestrene, since their physical constants are practically identical. Limonene, CioHw Sylvestrene, Boiling-point ...... . ..... 175-176 174-176 d 20 " .................... 0.845 0.847 n D ..................... 1.4746 1.477 M D .................... 45.23 45.08 The close agreement of these constants is to be expected from the constitution of these hydrocarbons q. v. Wallach 85 has called attention to the fact that hydrocarbons hav- ing a semicyclic double bond show abnormally high refractivity. 81 Wallach, Awn. S50, 142 (1906) ; 374, 224 (1910) **Ber. 44, 3423 (1911). 81 Wallach, Ann. 395, 76 (1913) ; 396, 273 (1913). "Ann. 245 t 191 (1888). **Ann. 345 f 142 (1906) ; 360, 34 (1908). PHYSICAL PROPERTIES 561 MOLECULAR REFRACTION Calculated 31.83 31.83 36.43 41.03 Observed Observed CHCH 32.12 31.89 32.26 31.8 36.82 ^CH 2 CH 3 36.52 )>=C(CH 3 ) )>CH(CH 3 ) 2 41.56 _CH 2 CH 2 41.02 CH 2 CH 2 CH 36.43 CH 2 CH 2 36.64 CH 3 36.43 Similarly the terpenes, terpinolene, sabinene, d.l.fenchene, p-terpinene and (3-pinene have been shown by chemical investigation to have semi- cyclic double bonds, >C = CH 2 , and the exaltations of their specific refractions due to this group vary from 0.3 to 0.5. Auwers 86 there- fore argues that camphene must have the constitution proposed by Wagner since the exaltation of the molecular refraction (MD) of camphene is 0.51. M D 44.02 :calc. 43.51 camphene (Wagner) "Ann. 387, 240 (1912). 562 CHEMISTRY OF THE NON-BENZEN01D HYDROCARBONS Here also, however, the chemical evidence is much more convincing that camphene has the structure shown. 87 The refractive index has been of very little value in the examina- tion of commercial hydrocarbon oils. Rittman and Egloff, 88 give the results of the examination of corresponding fractions of seventeen different petroleums, five from California, four from Pennsylvania, five from Oklahoma, two from Russia and one from Mexico. As one would expect from what is known of the different types of hydro- carbons present in different petroleums, the refractive indices varied within wide limits. Fraction Refractive Indices To 100 1.375 to 1.423 100-150 1.407 to 1.434 150-200 1.425 to 1.448 200-250 1.437 to 1.465 250-300 1.449 to 1.493 The refractive indices were found to vary as the specific gravities. The refractive index is usually determined in the examination of essential oils but the so-called "constants" obtained are of no value as evidence of adulteration unless supported by other good evidence. 89 It is well known that the refractive index of a substance varies with the wave length of the light employed, light of short wave length giving the greater refraction. The dispersion or difference in the refraction of two different wave lengths, for example, the a and y hydrogen lines, may be measured and the specific and molecular dis- persivities calculated. 90 However such determinations can hardly be carried out except in a well equipped physical laboratory and the results show little more than the refractivity for a single wave length, for example, the sodium D lines, a strong, nearly monochromatic and satisfactory light available in any laboratory. Magnetic Rotation. When a beam of polarized light is passed through a transparent substance placed between the poles of an electro-magnet, so that the light travels in a direction parallel to the lines of the magnetic field, 8T Cf. Haworth & King, J. Chem. Soc. 105, 1342 (1914); Buchner. Ber. L6, 759, 2108 (1913) ; Lipp, B&r. 1,1, 871 (1919) ; Komppa, Ber. 47, 934 (1914). 88 7. Ind. & Eng. Chem. 1, 759 (1915). 89 Cf. Gildemeister & Hoffmann, Die Aetherischen Oele, Ed. II, Vol. I, 580 (1910). 80 Cf. Auwers & Eisenlohr, J. prakt. Chem. (2) 82, 70 (1910); Darmois, Compt. rend. Ill, 952 (1920) ; Falk, J. Am. Chem. Soc. 31, 86, 806 (1909) shows that disper- sion, M0 M O or M y M fl is not affected by temperature. PHYSICAL PROPERTIES 563 the plane of the polarized light is rotated. The original discovery was made by Faraday, but has been thoroughly investigated by W. H. Perkin, who has shown that the magnetic rotatory power of a sub- stance depends partly upon its constitution. Perkin has applied the method to the study of the constitution of certain hydrocarbons, but it has never been widely employed, probably on account of the fact that the apparatus required is rather costly and involved; that, as Faraday showed, the amount of rotation is proportional to the strength of the magnetic field ; that a "series constant" must be employed which is different for each slight difference in constitution, e. g., 0.508 for normal paraffines and 0.631 for iso paraffines and unknown for most of the other possible isomeric types; that the "constant" increment for the double bond varies from 0.578 to 1.112 in different types of substances; that the effect of substitution may vary with each suc- cessive substituent as when substituting halogens. In spite of the very admirable and painstaking work of Perkin, which is of the great- est interest from a physical standpoint, the method has not proved of great value in the investigation of hydrocarbons and the reader is therefore referred to Perkin's original papers, 91 or to other sources 92 for further information in regard to it. Optical Activity. The majority of the terpenes occurring in nature are optically active. The degree of rotation of a particular terpene may vary practically from the one extreme value to the other, or from extreme Isevo to dextro-rotatory power. Bacon 93 noticed that the rotatory power of specimens of phellandrene distilled from Manila elemi, collected from separate trees, varied from 60.6 to + 126.0. Turpentine, chiefly ex-pinene, varies in much the same way, though within smaller limits. The turpentine from American long leaf pine, Pinus palustris, is preponderatingly dextro rotatory, that from the Cuban pine, Pinus heterophylla, is usually Ia3vo rotatory, and French turpentine, from Pinus pinaster is usually highly Ia3vo-rotatory. In most cases, a par- ticular species yields one or more terpenes whose optical activity is characterized by being strongly dextro or la?vo-rotatory. Thus the Aleppo pine 94 of southern Europe yields a highly dextro a-pinene, 91 J. Chem. Soc. 69 (1896) ; 81. 315 (1902) ; 89, 849 (1906) ; 91, 835, 851 (1907). 92 Cohen, Org. chem. Vol. II, 44, Ed. II (1919). Smiles, Relations between Chemi- cal Constitution and Physical Properties, 1910. 93 Philippine J. Sci, 4, 96 (1909). *Vezes, Bull. Soc. chim. (4) 5, 931 (1909). 564 CHEMISTRY OF THE NON-BENZENOW HYDROCARBONS [ a \\ 4~ 48.4, and a-pinene of extreme Isevo rotation, [a] ^ 48.63, has been found in the essential oil of one of the species of eucalyptus. 95 Optically inactive pinene is much rarer but is found in American oil of peppermint, coriander and some lemon oils. In the case of limonene the dextro form is much the commoner variety and although the nature of racemic substances had long been understood, it was not until the discovery of 1. limonene that Wallach was able to show that equal portions of d . and 1 . limonene yield the derivatives charac- teristic of "dipentene." Thus dipentene tetrabromide, melting-point 120, is the racemic form of the d. and l.tetrabromides melting at 104-105. That heat tends to racemize optically active hydrocarbons is well known but in most cases distillation at atmospheric pressure, of the terpenes, does not appreciably affect the rotatory power. Bacon has noted that in the case of a-phellandrene exposure to direct sunlight causes comparatively rapid racemization. When it is attempted to prepare optically active hydrocarbons by decomposing optically active alcohols, racemization occurs simultane- ously and the resulting hydrocarbons are usually inactive. Perkin and K. Fisher decomposed terpineols by magnesium-methyl iodide in the cold, and also by heating with anhydrous oxalic acid but the resulting hydrocarbon was dipentene. 96 However, Perkin succeeded in preparing the isomeric hydrocarbon, A 3 8(9) -p-menthadiene, in highly optically active form by resolving one of the intermediate products into its~d and /. forms. The acid CH 2 CH MeCH 67.3 The isomeric m.menthadiene was also obtained in an optically ac- tive form by resolving the unsaturated acid CH 3 by means of \/C0 2 H /.menthylamine and subsequent reactions as in the case of the p.men- thadiene. The optical activity of petroleum, or more accurately, certain fractions of petroleum distillates, is one of the most significant facts bearing on the theory of the formation of petroleum from organic remains. According to Engler, 97 no oils which have been carefully examined are entirely without optical activity. Most petroleums are dextro-rotatory but a lvo oil has been reported from Borneo. Tschugaeff 98 called attention to the optical activity of a vaseline oil in 1904 and also stated the importance of this fact to the theory of organic origin. Rakusin " then reported dextro-rotatory fractions from American, Baku and Grossny oils. In 1835 Biot 10 had observed dextro-rotation in a "naphtha" of unknown origin but later observers all had regarded petroleum as optically inactive. Since 1904, how- ever, many observers have confirmed the fact that certain fractions of petroleums are optically active. Crude petroleum cannot be meas- ured for optical activity on account of the color and asphaltic matter which is frequently present. Cholesterol yields oily decomposition products when subjected to destructive distillation 101 and wool grease yields optically active oils on decomposition, which led Marcusson 102 to attribute the optical activity of these oils and petroleum oils to decomposition products of " Ber. 47, 3358 (1914). **J. Russ. Phys.-Chem. Soo. 36, 453 (1904) ; Chem. Ztg. 190$, 505. M J. Russ. Phys.-Chem. Soc. 36, 456 (1904) ; Chem. Ztg. 190$, 505. ^Mem. de I'acad. Sci. 13, 139 (1835). 1M Windaus, Ber. 37, 2027 (1904). 102 Chem, Ztg. 1906, 788, 566 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS cholesterol. Engler and his students have shown many points of sim- ilarity in the oils from cholesterol and optically active petroleum frac- tions. Thus when cholesterol is rapidly distilled the resulting oil is slightly Ia3vo-rotatory and when this ^.distillate is heated for several hours the rotatory power diminishes and finally becomes dextro rota- tory; a closely parallel behavior is shown by the two Isevo-rotatory Java oils. 103 The chemical character of the optically active substance in petro- leum has not been definitely shown. The naphthenic acids isolated from Baku oil are feebly active, 10 * but oils which have been entirely freed from these acids show practically undiminished optical activity. Engler and his students have concentrated the optically active sub- stance by repeated fractional distillation in vacuo until the most active fraction represented only 3 per cent of the original material but was unable to find indications of the presence of any substance other than hydrocarbons. Evidence that the optical activity of all petroleums is .derived from a common original material, perhaps choleresterol, is afforded by the fact that the fractions of greatest degree of rotation, of the various petroleums examined, have approximately the same range of boiling-point, as is shown by the following table, which also shows the magnitude of rotation of these fractions after concentra- tion by repeated distillation. MAXIMUM OPTICAL ACTIVITY OF PETROLEUM FRACTIONS. 105 P Petroleum Source Fraction B.-P. C mm. Saccharimetcr Hanover 235-275 12 +10.4 Baku 230-278 12-13 +17.0 Galicia 260-285 12 +22.8 Roumania 250-270 12 +22.0 Pechelbronn 265-281 12.5 +7.6 Pennsylvania 255-297 14 +1.0 Java 268-281 15.5 +4.1 Optical activity has not been observed in petroleum fractions boil- ing below 200 at atmospheric pressure. Specific Heat. In 1831 Neumann discovered that in a series of compounds of analogous composition the specific heat varies inversely as the mo- 103 See cholest&rylene. 'Bushong & Humphrey, 8th Int. Gong. Appl. Chem. 6, 57 (1912). 105 Engler, Das Erdol, Vol. I, 202, 1913. For Isevo-rotatory oil see Jones & Woot- ton, J. Chem. Soc. 91, 1146 (1907). Hydrocarbon Molecular Weight 86 Specific Heat 05272 No. of Atoms 20 C H , 100 05074 23 CsHis 114 05052 26 128 05034 29 r* 9 H! 142 05021 32 r* H 156 05013 35 C^HM 170 0.4997 38 C Ho 184 0.4986 41 C 4Hao 196 0.4973 44 C 15 H 32 , 210 0.4966 47 224 0.4957 50 PHYSICAL PROPERTIES 567 lecular weight. Mabery 106 determined the specific heats of a series of light fractions of Pennsylvania petroleum probably consisting of normal paraffine hydrocarbons. The uniform decrease in specific heat with increasing molecular weight suggests a constant relation analogous to the law of Neumann. In the following table the con- stant K is expressed in terms of the specific heat multiplied by the molecular weight and the product divided by the number of atoms in the molecule, Molecular K. 2.26 2.21 2.21 222 2.23 2.23 2.23 2.24 2.23 2.24 2.23 Mabery also gives the latent heat of vaporization of the following hydrocarbons, Heat of Vaporization Boiling-Point in Calories Hexane 68 79.4 Heptane 98 74.0 Octane 125 71.1 Cyclohexane 68- 70 87.3 Dimethylcyclopentane 90- 92 81.0 Methylcyclohexane 98 75.7 Dimethylcyclohexane 118-119 71.7 From an industrial point of view data on the specific heats and heat of vaporization of various petroleum fractions are of value in order properly to design stills and condensers. 107 According to Trou- tons' rule the molecular heat of vaporization divided by the absolute boiling-point equals a constant, which is approximately 20. Graefe 108 points out that this relation can be employed to calculate the mean heat of vaporization of petroleum fractions (which distill at atmos- 106 Am. Chem. J. 28, 66 (1902). 107 In the early days of the American petroleum industry m-.ich of the apparatus employed was perfected and developed purely by the cut and try method. Karawajeff [J. Soc. Chem. Ind. 32, 128 (1910)] states that the average specific heat of heavy petroleum oils at 100C is about 0.48, rising as a linear function of the temperature to about 0.60 at 400C. For the specific heats at low temperatures see Bushong and Knight, J. Ind. & Eng. Chem. 12, 1197 (1920). 108 Petroleum 5, 569 (1909) ; Chem. Abs. 4, 1362 (1910). 568 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS pheric pressure without decomposition). Graefe used a rather ingen- ious method for determining the mean molecular weight, i. e., lowering of the freezing-point of stearic acid. His results, calculated in this manner, are as follows: Heat to Average Average Heat of Raise to Total Material Sp.Gr. Mol.Wt. B.-P. Vaporize B.-P. Heat Light crude oil 0.883 1 13 216 86.5 82 168.5 Gas oil 0.890 158 273 69.2 107 176.2 Paraffine oil, light 0.920 190 328 63.3 130 193.3 " heavy ... 0.933 230 346 53.8 138 191.8 Values, for petroleum distillates, determined calorimetrically usually fall within the range 130 to 190 calories. Thermochemistry of the Non-benzenoid Hydrocarbons. There is little question but that organic chemistry is too largely a compendium of methods of preparation, and, considering the meager equipment of many laboratories and the ease with which most organic reactions may be carried out, it was perhaps inevitable that this should be so. Also the difficulties in the way of understanding the theory or mechanism of organic reactions are manifold, and many factors other than thermochemical relations play very important roles. Heat changes in organic reactions are often small and reac- tions frequently take place with the formation of substances which do not directly lead to increased entropy of the system; stereo chemi- cal relations play an important part, many phases of which, for exam- ple the Walden inversion, we are far from understanding. The modi- fication of the chemical properties of a given atom or element by neigh- boring substituents of pronounced chemical character is a factor which we know mostly in a qualitative way, in much the same way that a chef is familiar with the strength of his assortment of condiments. We know in a more or less quantitative way that a condition of stress in a molecule affects certain physical properties and endows the sub- stance with unusual chemical activity, as for example cyclopropane. Thermochemical data have, as yet, been of very little assistance to organic chemists. Thus, according to Thomsen, ethylene oxide must have the structure H 2 C.O.CH 2 "for the introduction of an atom of oxy- gen into the molecule of ethylene, in place of the double linkage, corre- sponds to a thermal effect of 93.98-73.47 = 20.51 Calories, since the PHYSICAL PROPERTIES 569 taking up of an atom of oxygen by the ethane molecule in place of the single linkage produces a heat effect of 124.95-104.51 = 20.44 Calories. The relation is, therefore, exactly the same, and, if dimethyl ether has the composition CH 3 .O.CH 3 ethylene oxide must be dimethylene ether, CH 2 .O.CH 2 . The view that ethylene oxide contained a single link- age between the carbon atoms (CH 2 CH 2 ) would necessitate a heat \' of formation greater by 14.71 Calories, that is to say about 15 per cent higher than the experimental value." 109 No organic chemist would accept Thomsen's proposed structure of this substance in view of its many chemical reactions which point clearly to the ethylene oxide formula. Early in the use of the refractometer, Wallach cau- tioned Briihl that the refractive index should not be relied upon to decide questions of constitution unless well supported by chemical evidence and the history of such disputed cases has amply borne out Wallach 's contention. Thus in the case of thermochemical evidence as to the constitution of benzene and other organic compounds, it should always be kept in mind that in the present state of our knowl- edge thermochemical data are evidence, not necessarily proof. Also in many organic systems the mere number of reactions which are observed to take place precludes quantitative predic- tion, at least in the light of our present understanding. For example, in the pyrolysis of the simpler paraffine hydrocarbons we can calcu- late the thermal changes involved in a large number of reactions, some exothermic and others endothermic. One author has been criticised for employing the Nernst formula to calculate the reaction velocities of a large number of possible and impossible reactions of hydrocar- bons at various temperatures, for example, K K K 600 750 900 C + H 2 > CH 4 0.077 0.012 0.003 C 2 H 6 C 2 H 4 + H 2 0.0027 0.094 1.28 Experimentally it has been shown that ethylene is produced to a less and less extent as the temperature rises within this range, the per- centage of ethylene after one minute being, at 675, 24 per cent; at 810, 11 per cent; at 1000, 7 per cent. 110 109 Thomsen-Burke, Thermochemistry, 1908, 453. 110 Bone & Coward, J. Chem. Soc. 93, 1197 (1908). 570 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS As pointed out by Thomsen, the determination of the thermal! effect which should result on formation of organic substances from their] elements is particularly difficult, for only a few such substances can! be so formed and these only under conditions which practically pre-1 elude such measurement. No satisfactory method is at present known 1 for the measurement of these values, except determination of the] heats of combustion. The heats of combustion of a number of hydro-] carbons, as found by Thomsen, are given in the table below. "The third column gives the heats of combustion, which in the! case of non-gaseous bodies is given in the state of gas or vapor at 18. i In each case it is assumed that the products of combustion are cooled to 18, and that consequently carbon dioxide and nitrogen appear as gases, water, on the other hand, as a liquid." "The fourth column gives the heats of formation of the products of the combustion; that is to say, the amount of heat which. is evolved by the elements of the compound when they are burned in the free state, as, for instance, carbon to the dioxide, and hydrogen to water. The heat of combustion of carbon is taken as 96,960 c. for each gram-atom of carbon, this being the heat of combustion of amorphous carbon. The heat of formation of water is 68,360 c. per gram-molecule." "The fifth and sixth columns contain the heats of formation of th substances in the state of gas or vapor at 18. This value is cal- culated from the heats of combustion according to the equation already given. (C a , H 2b , 0.) - - a(C,0,) + b(H,,0) f(C.H,b,0.) "The values calculated in this manner are the heats at formation at constant pressure. External conditions, however, exercise a certain influence on these values, since the products formed usually occupy a smaller volume than the sum of the volumes of the constituent ele- ments. Thus 2 gram-molecules of hydrogen are required for the formation of 1 gram-molecule of CH 4 ; this corresponds, therefore, to a decrease in volume of 1 gram-molecule of hydrogen, or of 22,340 cubic centimeters, at and 760 mm. pressure. Such a diminution of volume will result in the evolution of 543 c. at 0, which corresponds to 580 c. at 18. If now from the heat of formation of the compound we subtract 580 c. for each gram-molecular volume which has dis- appeared, we obtain the heat of formation at constant volume. It is this value, which is given in the sixth column of the following tables: PHYSICAL PROPERTIES HYDROCARBONS. 571 (2a + b)OaC0 2 + bH 2 O Heat Heat Heat of Formation of Com- of For- of the Compound Combustion : C a H 2b bustion motion at Con- at Con- Molecular of the of the stant slant Compound Formula gas at 18 products pressure volume PARAFFINS. Methane CH 4 211,930c. 233,680c. 21,750c. 21,170c. Ethane C 2 H 370,440 399,000 28,560 27,400 Propane C 3 Hs 529,210 564,320 35,110 33,370 Trimethylmethane .... CH(CH 3 ) 3 687,190 729,640 42,450 40,130 Tetramethylmethane. . C(CH 3 )4 847,110 894,960 47,850 44,950 Diisopropyl (CH) 2 .(CH 3 ) 4 .... 999,200 1,060,280 61,080 57,600 UN SATURATED HYDROCARBON S . Ethylene C 2 H 4 333,350 330,640 2,710 3 290 Propylene, normal . . . Trimethylene CH.rCH.CH. .... 492,740 499,430 495,960 +3,220 +2,060 495,960 3,470 4 630 Isobutylene Isoamylene CH 2 :C:(CH 3 ) 2 ... 650,620 807630 661,280 +10,660 +8,920 826 600 + 18 970 + 16 650 Diallyl C H C H- 932 820 923 560 9 260 1 1 580 Acetylene Allylene CH'CH . CH'C CH 3 310,050 467550 262,280 47,770 47,770 427600 39950 40530 Dipropargyl . 882,880 786,840 96,040 97,200 According to Thomsen isomeric organic substances may give iden- tical heats of combustion, for example, Allyl chloride, CH 2 = CH.CH 2 C1 454.68 Cal. 2-Chloropropylene CH 2 = CHC1 . CH 3 453.37 Cal. and the two isomeric dichloroethanes, Ethylene chloride CH C1.CH,C1 296.36 Cal. Ethylidene chloride CH 3 CH.C1 2 296.41 Cal. It should be pointed out that the chemical properties of these isomers differ widely, for example, the chlorine atom is much more stable or firmly bound in 2-chloropropylene than in allyl chloride, and ethyli- dene chloride is much more reactive to water and alkalies than ethylene chloride. However, T. W. Richards ni has developed ex- ceedingly accurate methods of calorimetry by which he has detected slight differences in the heats of combustion of five isomeric octanes, and it is, therefore, possible that Thomsen's conclusions regarding the identity of the heats of combustion of isomers such as the above may have to be modified. The differences in the heats of combustion for the octanes studied by Richards are considerably greater than the experimental error of his method of measurement. The values found by him are as follows, 111 Richards & Jesse, J. Am. Ghent. Soc. 32, 268 (1910) ; S6, 248 (1914). 572 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Normal octane, liquid 5448 Kilojoules 2, 5 dimethylhexane 5442 2 methylheptane 5454 3.4 dimethylhexane 5444 3 ethylhexane 5439 Richards and Davis 112 find that the increase in the heat of combus- tion by substituting CH 3 for a hydrogen atom in a side chain is 648 4 kilojoules, but when substituted for a hydrogen of the ben- zene nucleus the value is 638 kilojoules. It is greatly to be regretted that so few thermal measurements of organic substances have been determined with the accuracy of Rich- ards' measurements. It would be of interest to determine if the high degree of molecular symmetry of tetramethyl methane C(CH 3 ) 4 and 2,2,3,3,tetramethylbutane CH 3 CH 3 CH 3 C- -CCH 3 CH 3 CH 3 which are characterized by abnormally high melting-points, show heats of combustion appreciably different from their normal isomers. The effect of conjugation of double bonds upon the l^at of com- bustion has been carefully investigated by Auwers, Roth and Eisen- lohr. 113 A number of terpenes were used in the investigation and. since empirical expressions for the determination of the "calculated values" of heats of combustion are not trustworthy, the average of the experimental values determined for limonene, i-limonene and sylves- trene, 1464 Calories was taken as a normal value for this terpene series. The values found for the molecular heats of combustion were as follows, Calories d.-limonene [A^W-p-menthadiene] 1466 i-limonene 1462 sylvestrene [A 1-8 < 9 )-m-menthadiene] 1464 a-phellandrene 1434 a-terpinene 1428 d-a-pinene 1469 camphene (liquid) 1471 sabinene 1475 According to their results hydrocarbons having two conjugated double linkings have molecular heats of combustion about two per cent lower than the isomeric hydrocarbons containing two non- conjugated double 112 J. Am. Chem. Soc. W, 1599 (1920). 118 Ann. 373, 267 (1910)! PHYSICAL PROPERTIES 573 linkings. These "thermal depressions" are about of the same order of magnitude as the exaltations of molecular refractivity of such hydrocarbons. On account of the refined experimental technique re- quired to make such thermal measurements with great accuracy, the thermal method will never displace the optical methods but may be helpful as an auxiliary in certain cases. Thermal measurements made in the course of this work support the contention that Semmler's car- venene is identical with a-terpinene. The parallelism between the thermal and optical data disappears in the case of the bicyclic ter- penes. The heats of combustion of a number of cyclic hydrocarbons have been reported by Zuboff, 114 as follows, expressed as Calories per gram molecule, based on Regnault's determination of the specific heat of water, Hydrocarbon Heat of Combustion Formula Name At Const. Vol. At Const. Pres. Normal Hexane 997.8 999.8 Methylcyclopentane 945.7 947.4 Cyclohexane 943.4 945.1 1.3 Dimethylcyclopentane 1099.5 1101.5 Methylcyclohexane 1100.8 1102.8 " ' Cycloheptane 1096.3 1098.3 C 8 H 16 1.1 Dimethylcyclohexane 1252.8 1255.1 1.3 " " 1248.1 1250.4 1.4 " " 1238.9 1241.2 C 9 H 18 1.3.3 Trimethylcyclohexane 1406.0 1408.6 C T Hi2 Methylcyclohexene, a ..' 1047.6 1049.3 /3 1053.2 1054.9 Cycloheptane 1058.7 1060.5 Roth and Auwers 115 have criticised the technique of Stohmann's earlier work and point out that many of the hydrocarbons investi- gated by Stohmann very probably were impure. They have redeter- mined the heats of combustion of a series of benzene and cyclohexane derivatives which were most carefully purified. Their results differ somewhat from Stohmann's and their results show that the increase in the heat of combustion produced by the addition of two hydrogen atoms to the aromatic hydrocarbon is much greater than that due to the addition of hydrogen to the dihydro and tetrahydro compounds; the latter two increases also are not the same. The difference between the heats of combustion of a conjugated cyclohexadiene and the cor- responding aromatic hydrocarbon is about 64 Calories. The difference between that of a conjugated cyclohexadiene and the cyclohexene is about 50 Calories, and the difference between the cyclohexene and the 114 J. Russ. Phys.-Chem. Soc. 33, 708 (1901). 118 Ann. 407, 145 (1915). 574 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS cyclohexane is about 45 Calories. In the case of simple acyclic ole- fines the difference between the saturated and unsaturated substance is about 37 Calories and Roth and Auwers state that hydrogenation appears always to be an exothermic reaction. The following values (Calories) are the molecular heats of combustion at constant pressure and at the specified initial temperature. Substance Calories Temp. benzene 782.3 20.4 cyclohexene 893.7 21.0 cyclohexane 938.5 17.4 toluene 935.2 19.0 1-methylcyclohexene 1049.6 15.7 ra-xylene 1089.5 20.6 1.4-dimethyl-A 1>3 -cyclohexadiene 1153.7 19.5 1-ethylcyclohexene 1205.4 14.9 l-methyl-4-ethyl-A ll3 -cyclohexadiene 1312.5 20.0 l-methyl-4-isopropyl-A 1-3 -cyclohexadiene 1472.2 20.5 naphthalene (solid) 1235.2 (liquid) 1239.7 A'-dihydronaphthalene 1297.8 20.5 A 8 - " (solid) 1299.8 20.7 (liquid) 1302.7 1.2.3.4. tetrahydronaphthalene 1341 .2 20.0 decahydronaphthalene 1503.9 19.2 The heat energy represented by the single bond carbon to carbon \ / C C may be ascertained by reference to the heats of combus- / \ tion of the paraffine series and the intramolecular energy of H 2 , the hydrogen molecule. Richards and Jesse 116 showed, in a series of unusually accurate determinations on very carefully purified octanes, that the mean value for the octanes, C 8 H 18 is 1299.9 Calories. The average value for CH 2 is 156 Calories. The heat of combustion of carbon and of hydrogen in hydrocarbons of the paraffine series may be determined by the simultaneous equations shown below, using the heats of combustion of ethane and propane found by Berthelot and Matignon, 117 X being the heat of combustion of a carbon atom and Y the heat of combustion of a hydrogen atom, C 2 H 6 , 2X + 6Y = 370.9 Calories C 3 H 8 3X + 8Y = 526.7 From these equations X = 96.5 Calories and Y = 29.65 Calories: also if we take X + 2Y = 156 together with the mean octane value 1299.9 Calories, then, in round figures, X = 96 and Y = 30 Calories. 116 J. Am. Chem. Hoc. 1910, 292. 117 Ann. chim. phys. (0) 30, 547 (1893). PHYSICAL PROPERTIES 575 Since this value is an additive one, it follows that the energy of dis- sociation of C-H and C-C bonds are practically equal, or within the limits of experimental error; for example, if there were a noticeable difference between the heat of dissociation, or rupture, of C-H and C-C, then the factor would be very different for C 2 H 6 , having 6 C-H bonds and one C-C bond, than with C 8 H 18 , having 18 C-H bonds and seven C-C bonds. As noted above hydrogen in the parafBne hydrocarbons has a heat of combustion of 30 Calories; by burning hydrogen gas, however, the molecular heat of combustion is not 2x30 or 60 Calories but 67.5 Calories. This difference, 7.5 Calories, can be due only to the greater intramolecular energy of the hydrogen molecule. Langmuir 118 and Isnardi 119 have calculated the heat of dissociation of hydrogen from other observations and both agree on the value 90 Calories for higher temperatures and Nernst 120 has calculated that at absolute zero the value would be about 100 Calories. Using Nernst's value the heat of dissociation of the carbon-hydrogen bond, and also the carbon-carbon single bond, is 100 7.5 = 92.5 Calories. 121 A second carbon-to-carbon bond as in ethylene, H 2 C = CH 2 . has a much smaller heat of dissociation. Thus for ethylene: Calculated, C 2 H 6 , 2 x 96 + 4 x 30 = 312 Calories Found by Berthelot 122 340 " Found by Mixter 123 344.6 " The increase for the double bond is, according to these observa- tions, 28 to 32.6 Calories, and the differences between the observed values and those calculated in the same manner in the case of other olefines, agree very well with the above values, for example, Propylene, calculated 468.0 Cal. Propylene, found (Berthelot) 497.9 Cal. Difference 29.9 Cal. Hexylene, calculated 936.0 Cal. Hexylene, 124 liquid, 959.9 Cal., and adding 7.8 Cal. for heat of vaporization 967.7 Cal. Difference 31.7 Cal. 118 Z. Elektrochem. 23, 242 (1917). 119 Z. Elektrochem. 21, 416 (1915). 320 Die Grundlagen des neuen Warmesatzes, 1918, 153. 121 Weinberg, er. 52, 1501 (1919). 122 Ann. chim. phj/s. (6) 30, 557 (1893). 323 Cliem. Zentr. 1901 (2), 1250. 1=1 Zubow, cf. Landolt & Bornstein, Physikalische Tabellen, Ed. 1912, 909. 576 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS The mean value for the increase in the molecular heat of com-? bustion of simple olefines is therefore about 30 Calories greater thai! the value calculated from the number of carbon and hydrogen atoma^ contained in the hydrocarbon. In the case of conjugated diolefines, however, the difference between the observed and calculated values is not 2 x 30 = Calories, but a very much smaller value, for example, A 2 " 4 -Hexadiene Calculated (6 x 96 + 10 x 30) 876.0 Cal. Observed 125 884.7 + 7.6 892.3 Cal. Difference 16.3 Cal. Other differences of the same order, between the observed and calcu- lated values for the dienes, have been observed. 126 Dielectric Constants: The electrical insulating value of the paraf* fines, and refined mineral oils generally, is taken advantage of in the utilization of oils for transformers, and the use of paraffine in im- pregnating the cotton insulation of wires carrying low voltage cur- rents. The dielectric constant of paraffine wax has been referred to Under this head. This very property which makes these hydrocarbon oils valuable as insulating material is a source of considerable danger in the case of the more volatile mineral oils of low flash point since they easily become electrostatically charged by friction, as by being pumped through a pipe, or agitating woolen goods in a gasoline clean- ing mixture. Electrical spark discharges caused in this way have re- sulted in many disastrous explosions and fires. Holde 12T states that light petroleum oils can easily acquire a charge amounting to several thousand volts by being pumped through a metal pipe. Even when the pipes and containers are grounded it is possible that, in the case of such good insulators, the electrical charge cannot be sufficiently rapidly dis- sipated. Holde gives the specific conductivity of "light petroleum" as 10" 14 to 10' 15 . Decrease of the dielectric constant with rise in tem- perature is very small, the temperature coefficient for cyclohexane being 0.00078. Viscosity. Measurements of viscosity have been of value as evidence of molecular association, the formation of hydrates in aqueous solution, the existence of racemic liquid substances and Dunstan and 125 Roth & Moosbrugger, Ann. ^07, 153 (1915). 28 The effect of conjugation of double bonds on the heat of combination is of par- ticular interest in connection with the constitution of benzene. Cf. Weinberg, Her. o&j loul (1019^. m Ber. JfH, 3239 (1914). PHYSICAL PROPERTIES 577 Thole 128 have noted certain facts which indicate that this property is to a certain extent influenced by the constitution, of organic substances. From the few facts which are known, it appears that the normal paraf- fines pentane, hexane, heptane and octane, have slightly greater viscos- ities than their branched chain isomers. 129 Ortho-xylene has a greater viscosity than meta or para-xylene, but the viscosities of a number of isomeric non-benzenoid hydrocarbons have never been compared. Alicyclic hydrocarbons have greater viscosities than paraffines of the same boiling-point but as to the constitution of mineral lubricating oils practically nothing is known. Viscosity decreases rapidly with rise in temperature and the curves are apparently hyperbolic. 180 The relation between viscosity and lubrication has been reviewed by Mabery and Mathews, 131 who point out that while viscosity is generally accepted as a standard of value in classifying lubricating oils, it is not certain that it is reliable as indicating the durability and wearing qualities of oils differing widely in composition. The vis- cosity of lubricating oils has received considerable attention from engineers and analysts and many forms of apparatus have been pro- posed for its determination, but these instruments all give arbitrary values, which are the resultants of several factors, of which viscosity is one. Most of these instruments measure the rate of flow of the oil through an orifice and the interpretation of the results is based upon the assumption that the flow of oil through an orifice is a cor- rect measure of surface viscosity between bearing surfaces. Mabery and Mathews obtained a set of relative values for the specific viscosity of hydrocarbons obtained by fractional distillation of petroleum. They employed Ostwald's method in which the oil is made to flow through a capillary tube under a definite pressure. The various frac- tions had approximately the composition indicated by the formulae. Spec. Viscosity Hydrocarbon Boiling-Point Sp. Gr. at 20 CTH M 98-100 0.724 0.51 C 8 H M 125 0.735 0.60 doHa 172-173 0.747 0.96 CisH*, 212-214 0.769 149 Ci 5 H 158-159(50mm.) 0.793 2.79 CicH^ 174-175 " 0.799 335 CwH*. 199-200 " 0.813 5.97 128 Cauwood & Turner, J. Chem. Soc. ion, 276 (1915). 128 Thorpe & Roger, Phil. Trans. 185A, 397 (1894) ; 189A, 71 (1897). uo Bingham & Harrison, Z. physik. Chem. 66, 1 (1909). Dunstan and Stevens have determined the viscosities of a number of typical lubricating oils at temperatures within the range 70-200C and plotted the results in curves. J. Soc. Chem. Ind. 30. 1063 (1921). lil J. Am. Chem. Soc. SO, 992 (1908). 578 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS That the paraffines have markedly lower viscosities than the cyclic hydrocarbons of the same boiling-point is shown by the following, determined at 60. Series Boiling-Point Sp. Gr. Spec. Viscosity C n H 2 n+ 2 ......... 274-276 (50mm.) 0.775 8.51 C n H 2n -2 ............. 274-276 " 0.835 15.63 CnHan+2 ............. 294-296 " 0.781 10.88 C m H 2n -2 ............. 294-296 " 0.841 21.23 The marked effect of ring closing and the very slight effect of un- saturation is also shown by the following comparative values found by Thole. 182 Substance v\25 - 2 x 10 1 Hexane CH 2 CH 2 CH a . . 0.00311 4.2 Cyclohexane : 0.00894 12.6 CH 2 CH* CH 3 Methylbutyl ketone CH 2 CHa CO ............ 0.00584 5.8 Cyclohexanone | | 0.0280 29.1 Cxii2 ~ C^xia . . OHj Methylamyl ketone CH 2 CH 2 CH a CO 0.00766 5.9 .CHa Cycloheptanone 0.0259 20.6 CH 2 -< Hexane ............................................. 0.00311 4.2 Diallyl CH 2 = CH . CH 2 CH.CH = CH 2 .............. 0.00269 4.0 Isopentane m (CH 3 ) 2 CH.CH 2 CH 3 ..................... 0.00223 4.3 Trimethylethylene (CHa) 2 C = CH.CHs .............. 0.00212 4.3 Isoprene CH 2 = C(CH 3 ) .CH = CH 2 ................. 0.00214 4.6 Removal of paraffine wax improves the viscosity of lubricating oils, as is shown by the following data of Mabery and Mathews. INFLUENCE OF SOLID PARAFFINE ON VISCOSITY AT 20. Boiling-Point Specific Hydrocarbon (60mm.) Sp. Gr. Viscosity (a) Penn. distillate C n Ha n -a cooled to 10 and filtered ............... 312-314 0.868 88.16 (a) +2.5% solid paraffine of same boil- ing-point ........................ 312-314 0.868 82.30 (b) Penn. distillate CnH 2n _ 2 cooled to 10 and filtered ................ 276-278 0.861 37.57 (b) -f- 2.5% solid paraffine of same boil- ing-point ......................... 276-278 0.860 36.39 It is generally recognized that the real function of oil in lubrication is to maintain a liquid film between the moving metal surfaces. 13 * Under pressure the tendency is for the oil to be squeezed out and the 2 J. CJiem. Soc. 105, 2004 (1914). Thorpe & Rodger, Phil. Trans. 185A, 570 (1894). f. Ubbelohde, Petr. Rev. 27, 293, 325 (1912). PHYSICAL PROPERTIES 579 oil film broken; the cohesion of the oil film itself and its tendency to wet or adhere to the metal surface and its ability to penetrate inter- stices by capillarity are factors of prime importance. While these factors may not be generally recognized, viscosity has come to be con- sidered in a general way as a measure of the resistance to the break- ing down of the oil film. Jerome Alexander. 135 uses the expression "film" to denote a layer of fluid on the solid surface of the order of 10~ 7 centimeters in thickness and states that with a true lubricant the facility of slipping is maximal when a layer of such excessive tenuity separates the solid faces and nothing is gained by increasing the thickness of the layer, a fact experimentally demonstrated with castor oil. According to Alexander, lubrication depends wholly upon the chemical constitution of a fluid, and the fact that the true lubri- cant is able to render slipping easy when a film of only about one molecule deep is present on the solid faces, suggests that the true lubri- cant is always a fluid which is adsorbed by the solid face. Alexander explains the superior lubricating power of graphite in oil by the for- mation of a graphite surface on the metal to which the oil adheres more strongly and greater pressure is therefore required to break down this oil film. Similar views, supported by experimental evidence, have been expressed by Stanton, Archbutt and Southcombe 136 and by D. R. Mountford. 137 The latter believes that the molecules of the liquid enter into a firm "physico-chemical" union with the metallic surfaces (or "adsorption" according to Alexander). Using a friction testing machine of the Thurston type the friction coefficient of a certain nineral oil was reduced from 0.0065 to 0.0042 by the addition of 2 per cent of fatty acids. The experiments of the former authors were carried out at the National Physical Laboratory and they conclude that viscosity is not the only or the most important factor in cases of difficult lubrication. They also attribute "oilness" to adhesion or chemical (?) affinity between the metal and the lubricant. In their experiments one per cent of the fatty acids of rape oil added to a mineral oil lowered the friction coefficient from 0.0047 to 0.0033 and 60 per cent of neutral rape oil was necessary to produce the same effect. Widespread dissatisfaction exists with the present methods of test- ing employed to determine the lubricating value of oils and Dunstan I 138 J. Ind. & Eng. Chem. 12, 436 (1920). Engineering, 108, 758 (1919) ; Chem. Abs. 14, 491 (1920). 13T Proc. Phys. Soc. London 32, II, 1 (1920) ; Chem. A&8. 1$, 1475 (1920) ; Wells & Southcombe, J. Inst. Petr. Techn. 4, 219 (1918). 580 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS and Thole 138 have recently expressed the opinion (shared by the pres- ent writer) , that no method has yet been developed which gives values which express rationally or accurately the lubricating value of an oil. The friction testing machine of Thurston is designed to duplicate closely conditions actually obtaining in practice, but the friction coeffi- cients so obtained are 'strictly a function of the viscosity and, as pointed out by Ubbelohde, are superfluous if the viscosity is determined. A device for testing film stability under variable pressure and move- ment of surfaces would appear to be rational and might do much to clarify understanding of this subject. It is commonly stated that pressure has no effect upon viscosity but under very great pressures the viscosity of mineral oils is greatly altered. Under a pressure of 5 tons per square inch the viscosity of mineral oils increases 16-fold, and Bridgeman 139 noted that ordinary lubricating oils become very viscous at pressures of a few thousand atmospheres, and kerosene at 10 and 8,000 atmospheres changes to about the consistency of vaseline. The viscosity of mineral and fatty oils increases with pressure and at pressures greater than 800 kilo- grams per square centimeter the rate of change is very great. At 1,000 kg. per sq. cm. mineral oils have viscosities ten to twenty-five times the viscosities at ordinary atmospheric pressure. 140 Solubility: Most petroleums and their distillates are completely miscible in benzene, carbon bisulfide, ether and chloroform. Absolute alcohol does not dissolve crude petroleums completely but amyl alco- hol dissolves the hydrocarbons, leaving asphaltic matter undissolved. Petroleums containing relatively large proportions of aromatic hydro- carbons are dissolved by solvents such as alcohol to a larger extent than other petroleums. Oils containing a maximum proportion of paraffine hydrocarbons, such as light Pennsylvania oil, are generally least soluble. The lighter fractions are more soluble than the higher boiling fractions. In the following table the "critical solution tem- perature" was determined by heating the distillate with an equal volume of the solvent, then cooling slowly and noting the temperature at which turbidity appeared. 141 ***J. Inst. Petr. Techn. 4, 191 (1918). U9 Proc. Am. Acad. 1ft, 345 (1911). "Hyde, Proc. Roy. Soc. 91 A, 240 (1920). 141 Chercheffsky, J. Petr. 1910, 210. Grit. Sol. Temperature C acetic ethyl ale. 96 % anhydride 50. 78.5 68.5 91. 87. 104.5 36. 66. 47.5 72. 60. 79.5 31. 60. 53. 75.5 72.5 89.5 miscible at 20 53. 30. 57. 42. 63.5 PHYSICAL PROPERTIES 581 Sp. Gr. Petroleum oj Fraction American 0.780 (Pennsylvania) 0.800 0.820 Russian 0.780 0.800 0.820 Galician 0.780 0.800 0.820 Roumanian 0.780 0.800 0.820 Although paraffine wax has been a common commercial product for a great many years and finds most varied application both in the industries and in scientific work, very little information has been pub- lished regarding its solubility in various solvents or its solvent power for other substances. The following table gives the solubility of a hard paraffine, melting-point 64-65 (about 10 higher melting-point than that of the average commercial paraffine) which had been pre- pared from ozokerite. The softer waxes are probably more soluble tnan the sample here described. 142 g Paraffine Wt. oj Solvent Dissolved by to Dissolve 1 Solvent 100 g. 100 cc. Part Paraffine C& 12.99 .... 7.6 Light petr. to 75 C Sp. Gr. 0.7233.. 11.73 8.48 8.5 Turpentine, 158-160 6.06 5.21 16.1 Xylene, commercial 135-143 3.95 3.43 25.1 Toluene, 108.5-109.5 3.92 3.41 25.5 Chloroform 2.42 3.61 41.3 Benzene 1.99 1.75 50.3 Ethyl ether 1.95 .... 50.8 Acetone 0.262 0.209 378.7 Ethyl acetate 0.238 419. Ethyl alcohol, 99.5% 0519 .... 453 Amyl alcohol, 127-129 0.202 0.164 495. Methyl alcohol 0.071 0.056 1447. Methyl formate 0.060 .... 1648. Glacial acetic acid 0.060 0.063 1668. Acetic anhydride 0.025 3956. Formic Acid (cryst.) 0.013 0.015 7689. A comprehensive discussion of the problem of separating paraffine wax from viscous oils by the use of solvents has recently appeared, the raw material investigated being the oily distillates obtained by the distillation of shales and lignites at low temperatures. In connec- "Tawlewski and Filemonowicz, Ber. 1, 2973 (1888). 582 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS tion with this work the solubility of paraffine wax, melting-point 56, was determined in mixtures of benzene and alcohol at ordinary tem- peratures and in the cold. 143 SOLUBILITY OF PARAFFINE: GRAMS IN lOOcc. SOLVENT. Solvent 23 C 20 Acetone 0.27 0.06 0.02 Benzene-alcohol, 2:8 0.48 0.10 0.01 " 3:7 0.77 0.18 0.04 4:6 1.14 0.23 0.05 Alcohol, 94.5% 0.16 0.01 0.006 Unsaturated hydrocarbons are generally more soluble than satu- rated hydrocarbons which fact is utilized in their separation by liquid sulfur dioxide. The hydrocarbons, including unsaturated hydrocar- bons are very much less soluble in ethyl alcohol than alcohols, alde- hydes, ketones and esters, which fact is made use of in the manufacture of terpene and sesquiterpene-free essential oils. Very few data bear- ing on this have been published but the solubility of turpentine and 95 per cent alcohol may be taken as a typical example of the solu- bility of this type of hydrocarbon. 144 Temperature of Grams of 95% Alcohol Separation C in 100 g. Mixture 20.7 2.4] 42.2 3.4 }- oil rich phase 53.0 7.2J 53.1 10.2^ 44.0 20.3 37.2 30.6 29.6 48.3 23.9 52.8 > alcohol rich 16.3 61.4 phase 15.5 76.6 24. 81.1 63. 87.1 J Hexane is miscible with methyl alcohol at 42.8. 145 The effect of the hydroxyl group in diminishing the solubility of a substance in hydro- carbons explains the slight solubility of castor oil in lubricating oils. The solubility of castor oil in gasolene, which is of some technical im- portance, is very much like the behavior of aniline and the simpler paraffines. Castor oil is usually stated to be insoluble in gasolene but Atkins finds that it is miscible with isohexane at 40.8, with octane at 47.8 and that it is miscible at ordinary tempertaures in certain gaso- 143 Seidenschnur, Brennstoffchem. 2, 49, 73, 81 (1920). 14 *Vezcs & Mouline, Butt. .soc. chim. (3) 31. 1043 (1904). " 5 Rothmund, Z. physik. Chem. 26 t 433 (1898). PHYSICAL PROPERTIES 583 lenes rich in naphthenes such as that from Roumanian and Galician petroleum. 146 Aniline and aliphatic hydrocarbons are iniscible when warmed but separate into two phases when chilled. Thus amylene and aniline 147 are miscible at temperatures above 14.5. Grams Aniline in 100 g. Amylene Aniline Tempo. C Layer Layer 19.5 "81.5 4 ... 20.5 79.5 8 .'.'.' 24.2 75.8 10 28. 73. 12 34. 68. 14 45. 59. 14.5 miscible The normal hydrocarbons, pentane, hexane, heptane and octane, are miscible with aniline 148 at 72, 69, 70 and 72 respectively, cyclo- pentane at 18 and cyclohexane at 31. Crude petroleum oils contain considerable dissolved methane, ethane and propane and on heating or distilling the oil these dissolved gases are not immediately expelled. Markownikow showed that kero- sene and a sample of machine oil (Sp. Gr. 0.906) dissolved about 220 volumes of isobutylene at ordinary temperatures and that the gas was completely expelled only after heating to about 260. Unsaturated gaseous hydrocarbons, ethylene and propylene are dissolved from oil gas by compressing with heavy oil to a greater extent than the satu- rated hydrocarbons which accompany these olefines in oil gas, and on heating or applying diminished pressure to the heavy oil solution the evolved gas is accordingly richer in olefines. Russian kerosene dis- solves 0.144 volume of methane and 0.164 volume of ethylene at 10 and atmospheric pressure. McDaniel 149 has determined the solubil- ities of methane, ethane, and ethylene in ten organic solvents at tem- peratures from 20 to 60. The solvents in the order of increasing solvent power for methane at 25 are methyl, amyl, ethyl, isopropyl alcohols, benzene, toluene, m-xylene, hexane and heptane. With ethane and ethylene the same solvents fall into a similar series in the same order. Contrary to Just 150 McDaniels finds. that the solubil- ities of these solvents for nitrogen, oxygen and carbon dioxide do not follow in the same order as in the case of the hydrocarbons. Ethylene is more soluble in water than in kerosene 151 (water dissolves 0.149 146 J. Inst. Petr. Technologists, 6, 223 (1920). 14T Konowtilow, Ann. physik. (4) 10, 375 (1903). ""Chavanne & Simon, Compt. rend. 168, 1111 (1919). 149 /. Phys. Chem 15, 587 (1911). 160 Z. physik. Chem. 37, 342 (1901). 151 Gniewosz & Walfisz, Z. physik. Chem. 1889, 70. 584 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS volumes of ethylene at 20). According to Charitschkow, 152 care- fully refined kerosene dissolves more ammonia at 22, 0.4982 volumes, than at 0. The solvent power of compressed gases for certain solids and liquids having very low vapor pressures is a fact frequently over- looked. A compressed gas has noticeable solvent power for such a solid or liquid only when the gas compressed to the liquid state is of such a character as to dissolve the solid or liquid to a marked degree, and when one recalls that the physical properties of gas and liquid become identical at the critical point it becomes evident that the solu- bility curves of gas and liquid must merge smoothly into each other at the critical point. Ethyl chloride dissolves in 5 to 6 volumes of methane under 180 atmospheres pressure, at 17, and at 200 atmos- pheres the two Become miscible and the surface separating gas and liquid phases disappears. Iodine, camphor and paraffine wax dissolve in compressed methane to a marked extent and on releasing the pres- sure the paraffine wax is deposited again. Compressed ethylene also has a very marked solvent power for paraffine wax and stearic acid. 153 Cyclohexane has been proposed as a cryoscopic solvent but is un- reliable for this purpose on account of the tendency of substances containing hydroxyl, carboxyl, carbonyl or nitro groups to associate in this solvent. 154 Iodine is less soluble in cyclohexane than in ben- zene. 155 Hexane generally has less solvent power than benzene. One hun- dred grams of the former dissolves 0.37 grams of anthracene at 25, as compared with 1.86 grams in benzene. Ligroin, 100 g., dissolves 0.72 g. benzoic acid at 16 and turpentine dissolves 5.09 grams (at 25). Sulfur is markedly soluble in hexane, as indicated in the fol- lowing table, SOLUBILITY OF SULFUR IN HEXANE. IM Temp. C g.S. in 100 g. Solution 20 0.07 0.16 20 0.25 40 0.55 60 1.0 80 1.7 100 2.8 120 4.4 130 5.2 140 6.0 Bakuer Teclvn. Ges. 1893, 5; Gurwitsch (Wiss. Grundl, d. Erdolbearb, 1913, 100. 1M Villard, Chem. News, 78, 297, 309 (1898). 1M Mascarelli, Atti accad. Lincei (5) 17, 494. 165 Bruni, Oazz. chim. Ital. 42, 12. 1M Etard, Ann. chim. phys. (7) 2, 526; 3, 275 (1894). PHYSICAL PROPERTIES 585 The true solubility of sulfur in organic solvents is frequently dif- ficult to determine on account of the fact that sulfur frequently forms colloidal solutions, as in the now well-known example of colloidal sulfur in [3.|3-dichloroethyl sulfide. 157 The solubility of sulfur in the hydrocarbon caoutchouc has been such a case, further complicated by the fact that, on warming, the sulfur is able to combine chemically with the double bonds of the hydrocarbon. Also, sulfur appears to be more soluble in organic substances containing one or more chemi- cally bound sulfur atoms and Skellon 158 finds that as the per cent of chemically bound sulfur in vulcanized rubber increases, the solu- bility for sulfur as free sulfur increases. Thus ebonite may contain a greater proportion of free sulfur than soft cured rubber and still not bloom. Loewen 159 observed the solution of sulfur in rubber under a microscope and noted that when the time of "vulcanization" is short, droplets of melted sulfur are visible; on continued heating the mix- ture clears up and the droplets disappear but on cooling sulfur glob- ules may reappear. If the time of heating be a little longer than in the last case, there is not sufficient free sulfur in solution to form droplets on cooling but crystalline sulfur may slowly separate on standing. If the time of vulcanization is still further prolonged, no free sulfur will separate after cooling. The viscosities of rubber solutions in chlorinated solvents are approximately double the viscosities of solutions, of the same con- centration, in gasolene or benzene, but, after heating, all kinds of rubber solutions have about the same viscosity. 160 Gaunt finds that the viscosities of fine hard Para rubber in various solvents, in order of decreasing viscosity, are as follows, benzene, CHC1 3 , gasolene, ethyl ether. 161 Gaunt assumes that such rubber solutions contain aggre- gates of rubber micelles and that heating, mechanical working of the rubber or other processes which break up these aggregates, or which cause depolymerization, decrease the viscosity. A ten per cent solu- tion of raw Para rubber in amyl acetate is fluid enough to filter through ordinary filter paper. Acetone is soluble in rubber to the extent of about 17 per cent. Rubber may be precipitated from ben- zene or ether solutions by the addition of alcohol or acetone. Tetra- hydronaphthalene is said to have marked solvent power for rubber. 157 Wilkinson, Neilson & Wylde, J. Am. Chem. Soc. 42, 1377 (1920). 158 India Rubber J. 1,6, 723 (1913). Gummi Ztg. 27, 1301 (1913). 10 "Kirchoff, Caoutsch d gutta-percha 12, 8649 (1915). 181 India Rubber J. 47, 1054, 1093. 586 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Liquid sulfur dioxide readily dissolves aromatic and unsaturated hydrocarbons but saturated non-benzenoid hydrocarbons are only very slightly soluble in this solvent. Edeleanu 162 has developed a refining method based upon these facts. The method has found greater favor in Europe than in America, although kerosene refined in this way has a lower specific gravity and usually better burning qualities than kerosene refined by sulfuric acid. The unsaturated and aromatic hydrocarbons are easily separated by distillation from the low boil- ing sulfur dioxide but the oils thus recovered have not yet proven to be of special industrial value. The fraction boiling at 150-200 has been recommended as a turpentine substitute. With many oils the liquid sulfur dioxide method does not yield water-white oils, and in such cases refining with small proportions of sulfuric acid must be resorted to in order to get this result. The separation of the aro- matic and unsaturated hydrocarbons from the paraffines is much more efficient at low temperatures, a temperature of about 12 C being recommended. The method readily lends itself to analytical sepa- rations and has been checked by Egloff, Moore and Morrell. 163 Ben- zene, toluene, and xylenes and mesitylene are completely miscible with this solvent at 10 and when using 33 and 66 per cent by volume of liquid sulfur dioxide at 18 the pentane, hexane, octane, monane and decane fractions and gasolene from light Pennsylvania petroleum are practically insoluble. At 10, using the same proportions of solvent, these paraffines are soluble to the extent of about 1.8 per cent. Pennsylvania kerosene was found to be 3.6 per cent soluble at 10, using 66 per cent by volume of the solvent. Amylenes are completely miscible at 10 and 18. Cyclohexane is insoluble at 18 and 3 per cent soluble in an equal volume of sulfur dioxide at 4.5, and naphthenes of higher boiling-point, following the general order of solubility noted above, are less soluble than cyclohexane. The Non-benzenoid Hydrocarbons and Colloid Phenomena. A very large number of organic substances are much more spar- ingly soluble in petroleum ether than in ethyl ether or other organic solvents and this fact accounts for the wide employment of mixtures of petroleum ether and ethyl ether in recrystallizing organic substances in laboratory research work; the ethyl ether evaporates more rapidly 162 German Pat. 216,459; Petroleum 9, 862 (1914) ; Engler & ubbelohde, Z. angew. Chem. 1913, 177. 188 Met. & Chem. Eng. 18, 396 (1918). PHYSICAL PROPERTIES 587 and the substance crystallizes from the solvent mixture as it becomes continually richer in petroleum ether. However, when a high degree of supersaturation is quickly brought about, as by pouring a warm one to two per cent solution of stearic acid in gasoline, into a solution of a little sodium ethylate in gasoline, the whole quickly sets to a jelly. Numerous attempts have been made to prepare stable petroleum jellies, or "solidified petroleum." The sodium stearate jellies are not very firm and soon begin to exude oil, according to the well-known phenomenon of syneresis, common to all jellies of this type. Other more or less solid preparations of petroleum oils are really emul- sions. 164 A high-melting wax is sometimes added .to stiffen the jelly and one patentee adds about 15 per cent of turpentine in order to get a larger proportion of alkali stearate into solution when warm. An- ther patentee prepares an emulsion with gelatin which is then hard- ened by formaldehyde. 165 Calcium soaps, when dry, give clear solutions with mineral oils, gelatinizing on cooling. On stirring in water emulsification and stif- fening of the grease results. Commercial greases frequently contain up to 22 per cent of calcium soaps. 166 Colloids containing mostly soap and a little mineral oil are manu- factured and known usually as naphtha soaps. The presence of free fatty acid or unsaponified fatty oil assists in preventing the separa- tion of the petroleum oil. 167 The solubility of soaps in mineral oils increases rapidly with in- creasing molecular weight of the fatty acid, but in light petroleum ether the lead soaps are very sparingly soluble, 100 cc. of the hydro- carbon dissolving 0.0528 g. lead heptoate, 0.221 g. lead myristate and 0.017 g. lead stearate. 168 The subject of emulsions lies somewhat far afield from the subject matter and purpose of the present volume but anyone working with the non-benzenoid hydrocarbons is apt to be concerned with emul- sions of various types and a limited number of examples will there- fore be briefly mentioned. The theory of emulsions has been very 164 Kuess, J. Soc. Chem. Ind. 1906, 1141. Ten parts of stearin are combined with 9 parts caustic soda in 18 parts of water and 100 parts of kerosene stirred in, melted at 105-115 and the alkali soap converted into the more insoluble Al or Mg soaps by adding magnesium or aluminum sulfate. 165 van der Heyden, J. Soc. Chem. Ind. 1906, 236. For a general review see Behrend, Kunstoffe. 1911,, 356. 18 Cf. Holde, Z. Chem. Ind. Roll. 3, 270 (1908). ""Brit. Pat. 2,137 (1911) ; Chem. Abs. 6, 2014 (1912) ; 8, 1026 (1914). 1B8 Neave, Analyst 37, 399 (1912). 588 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS thoroughly reviewed by Bancroft. 169 The emulsifying power of fatty acid soaps, for mineral or other oils, is well known but the sulfonic acid derivatives of petroleum hydrocarbons or their alkali salts, also possess this property to a high degree. The removal of these sul- fonic acids from the treated oil by washing first with water and then with alkali, without undue loss of hydrocarbon oils, is one of the arts of the petroleum refiner. One of the most troublesome diffi- culties encountered by the refiner is the emulsification of water in oil, and this is particularly liable to occur in the manufacture of highly refined water white oils of the so-called liquid paraffine type. Also, when water is added to a heavy lubricating oil containing a lime soap the water becomes- dispersed in the oil and will change it to a grease. As pointed out by Bancroft, an emulsifying agent is a substance which goes into the interface and produces a film; if the adsorption of the emulsifying agent lowers the surface tension on the water side of the interface more than it does on the oil side, the interface will tend to curve so as to be convex on the water side, and we shall have a ten- dency to emulsify oil in water. If the adsorption of the emulsifying agent lowers the surface tension on the oil side of the interface more than it does on the water side, the interface will tend to curve so as to be concave on the water side, and we shall have a tendency to emulsify water in oil. Pickering 17 has described experiments on the emulsification of kerosene in fungicidal and insecticidal sprays, the oil being emulsi- fied with water and lime or basic copper sulfate. The oil globules in such an emulsion are probably prevented from coalescing by being enveloped in a pellicle consisting of particles of the solid much more minute than the globules themselves. "Apparently a precipitate con- sisting of any insoluble substance which is wetted more easily by water than by oil, if in a sufficiently fine state of division, will equally act as an emulsifier." Emulsions made with such an insoluble emul- sifier are in every respect similar to those made with soap and the like. Quite recently emulsions of heavy, nearly non-volatile oils have been prepared in casein solutions: these solutions can then be dried by spraying in a vacuum and the result is a flour, each globule of oil being protected by a film of dried casein. Flours have been made in this way containing as much as 85 per cent of oil. Other emulsions can probably be dried in the same manner. Apparently no 189 J. Phya. Chem. 16, 177, 345, 475, 739 (1912) ; 11, 501 (1913). "oj. Chem. Soc. 91, 2001 (1907). PHYSICAL PROPERTIES 589 industrial applications of this process have as yet been made in the case of mineral oils. Adsorption phenomena are of more than academic interest. The highly adsorptive charcoals, particularly coconut charcoal, activated by superheated steam, which were developed during the war for the manufacture of military gas masks, have proven to be highly effi- cient in selectively adsorbing the vapors of liquid hydrocarbons from natural gas. 171 Good "fifty minute" charcoal will adsorb ten to fif- teen per cent of its own weight of gasoline vapors. The charcoal granules employed are about 8 to 14 mesh and when saturated with adsorbed hydrocarbons the latter are expelled by superheated steam at about 250C. The selective adsorption of coloring matter from vaseline and lubricating oils by fuller's earth has long been known. When unre- fined black vaseline is filtered through warm fuller's earth the first product is a perfectly fluid oil and the successive portions which come through are progressively more and more viscous. This induced Day 172 to subject a crude petroleum to similar treatment and it was found that the first liquid to come through the column of fuller's earth consisted of light low boiling hydrocarbons. Day realized the significance of these facts and stated that in this way petroleum in passing through strata of clay and fine sand could be greatly altered ; asphaltic matter, if originally present, could in this way be removed by adsorption resulting in light petroleum of the Pennsylvania type. His views were communicated to the Petroleum Congress held in Paris in 1900 and, shortly after, Engler 173 confirmed Day's experi- ments. The subject was further investigated by Gilpin and Cram, 174 who liberated the oil in different sections of the fuller's earth columns by the addition of water. They found that the lighter fractions showed less loss, on treating with concentrated sulfuric acid, than the heavier more viscous fractions which were more strongly adsorbed and they interpreted this to mean that unsaturated hydrocarbons were selectively adsorbed by the fuller's earth. This would indicate that the crude oil employed by them contained large proportions of unsaturated hydrocarbons, which is extremely improbable. As to whether or not unsaturated hydrocarbons are selectively adsorbed by fuller's earth has not definitely been shown. 1T1 Burrell, Oberfell & Voress, Chem. d Met. Eng. 24, 156 (1921). 172 Proc. Am. Phil. Soc. 56, 154 (1897). " 3 Z. angew. chem. 1901, 889. "'Am. Chem. J. 40, 495 (1908). 590 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS Continuing these investigations Gilpin and Bransky 175 pointed out that it is not necessary to assume a vegetable origin for petroleums of the light Pennsylvania type. The petroleum found in the Trenton limestone is supposed to have been derived from organic remains buried in this strata, as indicated by the abundance of fossil remains found in this limestone, and this supposition is supported by the rela- tively large proportions of sulfur and nitrogenous compounds in this oil. By filtration of oils such as the Ohio-Trenton, California and Texas petroleums through a bed of fuller's earth, oils very similar to light Pennsylvania petroleum can be obtained. Any sufficiently fine grained and porous material is capable of absorption, to a greater or less degree, just as in the case of fuller's earth. Beaumont, Texas, petroleum containing 1.75 per cent sulfur, on passing through a kaolin filter gave a fraction containing 0.70 per cent sulfur. 178 In a later paper Gilpin and Schneeberger 177 showed that sulfur and nitrogen derivatives were selectively adsorbed from a heavy petroleum from Kern County, California. By two filtrations of this oil they obtained the following: Sp. Gr. % Sulfur Crude oil 0.912 0.541 Fraction A (1) 0.857 0.06 A (2) 0.8604 0.07 A (3) 0.869 . 0.104 B (1) 0.862 0.072 B (2) 0.8771 0.09 B (3) 0.8803 0.141 One filtration gave the following results with respect to nitrogen, Sp. Gr. % Nitrogen Crude oil 0.889 0.761 Fraction (1) ...' 0.8264 0.08 (2) 0.8421 0.116 (3) 0.852 0.289 (4) 0.8614 0.315 (5) 0.8737 0.332 A similar filtration of another sample of California petroleum of Sp. Gr. 0.9118 and boiling over the range 105-340 gave fractions vary- ing from the lightest, Sp. Gr. 0.8325, boiling-point 160-195 to a fraction Sp. Gr. 0.8984 and boiling-point 329-340. 178 Am. Chem. J. 44, 251 (1910). "'Richardson & Wallace, J. Soc. Chem. Ind. 1902. 177 Am. Chem. J. 50, 59 (1913). Chapter XVII. Physiological and Related Properties Odor: The physiology of odor is exceedingly obscure. However some generalizations can be made with reference to the relation be- tween chemical constitution and odor. The odor of a substance ap- pears to be a property of the whole molecule, but is greatly affected by the presence or absence of certain groups and also by the relative positions of different groups. Thus, the odor of the unsaturated ketones isomeric with a-ionone (q.v.) is markedly affected by each change in the position of the double bond in the ring and by the positions of the methyl group, with reference to the ketonic side chain. The marked difference in odor between isomeric normal, secondary and tertiary alcohols is well known. Many of the saturated non-benzenoid hydrocarbons have faint but more or less characteristic, rather agreeable odors. Methane and ethane are, to most persons, entirely odorless but the pentanes and other comparatively volatile hydrocarbons have odors, and many hydrocarbons having the group C(CH 3 ) 3 or CH(CH 3 ) 2 have camphor like odors, as do the aliphatic tertiary alcohols. The light fraction from the petroleum of the Jennings-Louisiana field and the gasolene obtained from the natural gas of the Houma-Louisiana field contains saturated hydrocarbons whose odor closely resembles turpen- tine. Unsaturated hydrocarbons have somewhat more pronounced odors than the corresponding saturated hydrocarbons but still much less intense than alcohols, esters, ketones, aldehydes, etc. The offen- sive odor of unrefined gasolines, particularly when made by cracking processes, has erroneously been attributed to unsaturated hydrocar- bons but the malodorous constituents in such oils are naphthenic acids and derivatives containing sulfur or nitrogen. The older idea is accounted for probably by the notion that refining by sulfuric acid consists merely, or essentially, in removing unsaturated hydrocarbons. Certain conjugated dienes, for example cyclohexadiene, and the light condensate obtained on compressing Pintsch gas containing cyclo- 591 592 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS hexadiene and probably cyclopentadiene, possess sharp pungent odors when strongly inhaled. Diallyl has an odor resembling horse-radish and the effect of the double bond in changing the odor of propyl alco- hol and propionic aldehyde to the very sharp irritating odors of allyl alcohol and acrolein, is well known. These examples also illustrate the fact that odor is a property of the whole molecule, not a blend or composite odor of the constituent groups; thus the double bond in propylene is practically odorless, and propyl alcohol and propionic aldehyde are quite without the sharp irritating properties of acrolein and allyl alcohol. Cyclic unsaturated hydrocarbons such as the terpenes and ses- quiterpenes have sweet agreeable odors. Fresh turpentine or pure pinene is sweet and agreeable in odor and the irritating offensive odor of the stored or oxidized product is due to formic acid and other oxidation products. Ring closing has very little effect upon odor, as, for example, n.hexane and cyclohexane, secondary hexyl alcohol and cyclohexanol ; when, however, the closing of the ring affects the con- stitution, as in the conversion of an aldehyde group to a hydroxyl group the change in odor is pronounced. Intensity of odor can hardly bear any relation to chemical stability as we find stable borneol, having a strong pepper and camphor-like odor, and ce-terpineol, which readily decomposes, has a very faint odor, and many other illustrations of this relation could be given. Also mercaptans are not appreciably different from alcohols in stabil- ity but their odors are beyond comparison. It should be noted, how- ever, that we are probably handicapped, in attempting to make com- parisons and generalizations, by the fact that we can know only what the human nose tells us, and this we can only very imperfectly de- scribe or record. 1 The subject of odor, while seldom mentioned in chemical texts and reviews, is given attention in these pages as it is a sense which is exceedingly useful to organic chemists. In many cases impurities can be detected with reasonable certainty by means of the nose when chemical tests fail or are not sufficiently delicate. Physiological Action: Strictly speaking, the saturated hydrocar- bons cannot be said to have any physiological properties, although inhalation of the vapors of the more volatile ones quickly produce drowsiness, followed by anesthesia, and, in extreme cases, death by Cf. Henning, "Der Geruch," Lelpslg, 1916. PHYSIOLOGICAL AND RELATED PROPERTIES 593 asphyxiation may result. 2 Workers in paraffine wax plants frequently develop skin sores which are supposed to be due to the closing of the skin pores by the wax. The value of soft flexible paraffine as a coat- ing over burns *is merely that of a non-irritant mechanical protection, protecting the tissue from the air, temperature changes, and giving the growing new tissue mechanical support. The saturated hydro- carbons are not attacked by oxidizing enzymes or other active body fluids and accordingly are entirely inert in the digestive tract. Viscous, water-white, tasteless mineral oils are widely sold for pharmaceutical purposes, under a variety of names, i. e., paraffinum liquidum, liquid petrolatum, paraffine oil and many special trade names. As has been pointed out by Marcusson most oils of this class contain no paraffine whatever but consist, like lubricating oils from which they are in fact made by drastic refining, of cyclic hydrocar- bons of the so-called naphthene or polynaphthene class, C n H 2I1 _ 2 , and C n H 2D _ 4 . Such an oil examined by Marcusson 3 had the specific gravity 20 of 0.8827 and showed the following analysis, Found Calculated for C*>H Carbon 86.74% 86.33% Hydrogen 13.43 13.67 The freezing test, by which small proportions of paraffine may separate, if present, may serve to differentiate between oils made from paraffine base crudes, and those made from paraffine-free oils such as Russian Baku, California or Gulf Coast crudes, but its presence is hardly to be condemned since if the paraffine does not separate at room temperatures, it certainly could not do so in the body. Bastedo* reports a clinical investigation of Russian and American oils and states that the choice between different oils of these types is an open one "to be determined by palatability, depending upon the degree to which the refinement has been carried out." Bastedo agrees with the original recommendation of Sir Arbuthnot Lane that oils for inter- nal use should have a specific gravity of not less than 0.885, on account of low viscosity and leakage. Exposure to sunlight in loosely stop- pered bottles will develop taste and odor in from 4 to 10 days and often serves to differentiate between the quality of oils of equal palatability when freshly prepared. 5 Various chemical tests have 2 Fuhner, Biocliem. Z. 115, 235 (1921). 3 Chem. Ztg. 1913, 550. */. Am. Med. Assoc. March 6, 1915, p. 808. Brooks, J. Am. Med. Assoc. 65, 24 (1916). 594 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS been proposed to determine quality, which are of more or less value in eliminating the personal equation involved in testing by taste and odor. Good oils on treating with concentrated sulfuric acid at room temperature will not be colored more than pale straw yellow in 5 minutes. Unsaturated hydrocarbons, ethylene, propylene, butylene and the amylenes, act upon the nerve centers producing first excitement and then narcosis. Amylene was proposed as an anesthetic by Snow in 1856, but it was found to be dangerous, due to sudden failure of car- diac motion. Its use was condemned by the French Academy soon after its introduction. Gwathmy 6 notes two deaths in 238 adminis- trations. The liquid unsaturated hydrocarbons are mildly irritating to the skin and mucous membranes. Taken internally unsaturated hydrocarbons cause severe gastric irritation and may even lead to convulsions and' death. The higher aliphatic alcohols, such as sec- ondary octyl alcohol, geraniol and the terpene alcohols, have dis- tinct bactericidal values though less than the phenols. The ses- quiterpene alcohol santalol is of value in treatment of gonorrhea but the value of borneol or its esters, cineol, menthol and the like in bronchial infections lies more in the stimulating effect of these sub- stances on the mucous membranes than in their slight bactericidal properties. The use of essential oils in medicine is very ancient and though many of the prescriptions of the old herb doctors have given way to carefully prepared and standardized extracts or to new syn- thetic drugs, many essential oils are used in cosmetics, and a few have distinct medicinal values, as American worm-seed or oil of cheno- podium, the active constituent of which is ascaridol (q.v.). In the treatment of persons suffering from hookworm, oil of chenopodium is more efficaceous than thymol; both are about equally effective in removing necators, but oil of chenopodium is much superior to thymol in removing the more resistant species of hookworm. 7 Essential o ; ls containing the ketone thujone, for example, the volatile oils of thuja, tansy, sage and wormwood (Artemisia absinthium) , produce character- istic disturbances of the central nervous system which, in the case of persons addicted to the drinking of the liqueur absinth, results in "rage tanacetique." Considerable difference of opinion seems to exist regarding the physiological properties of d,l. and d.l, or synthetic camphor. In "Anesthesia," 698, New York, 1914. * Report of Uncinariasis Comm. Rockefeller Inst., N. Y., 1920. PHYSIOLOGICAL AND RELATED PROPERTIES 595 England a court found that synthetic camphor possessed properties identical with those of natural camphor (except optical rotation) and therefore ruled that its use in pharmaceutical preparations was per- missible. 8 One observer 9 stated that he could detect no difference between the three kinds of camphor, by peritoneal injection, and later 10 reaffirmed that all three varieties are equally active. Leyden and Welden 11 treated frogs with chloral hydrate and reduced the heart beat to 7, after which the beat was raised to 20 by either natural d.camphor or la?vo-camphor but state that synthetic camphor was without action on the heart. Perkin's epicamphor was found to have an action on the heart slightly less than natural camphor. 12 On the other hand, Tsakalotos 13 states that synthetic di.camphor has the same heart action, also using frogs, as natural camphor, which state- ment is also made by Lutz. 14 Edsall and Means 15 investigated the effect of natural ^.camphor on respiratory metabolism but their results were so irregular that they were unable to draw any conclusion. Camphor vapor in concentrations of one to two parts per million in air is sufficient to effect the heart action markedly. [Heubner, Z. ges. exp. Med. I, 267 (1913).] Natural camphor has, however, a pronounced effect on the muscular respiratory system. 1 . 6 Heffter 17 states that there is no apparent reason for not using synthetic cam- phor in spirits of camphor but suggests that its use internally be not recommended until adequate clinical results are available. Sassen 18 used cats and dogs in studying the physiological proper- ties of natural and artificial camphor and states that with these ani- mals no material difference could be noted in the physiological effects of the two camphors. For both natural and artificial camphor the fatal dose is 2 grams per 1 kilo weight of the animal. Doses of 0.025 to 0.05 gram per kilo weight, of either camphor caused a per- ceptible increase in the heart's activity. The United States and German Pharmacopoeias prescribe natural camphor. Bruni 19 states that ^.camphor is about 13 times as toxic as natural d.camphor and Langgaard and Maass 20 state that the s Pharm. Zentr. 50, 563 (1909). Joachimoglu, Arch exp. Path. Pharm. 80, 1 (1916). Joachiinoglu, ibid., 80, 259, 282 (1917). 1 Arch. exp. Path. Pharm. 80, 24 (1916). 2 Bredt & Perkin, J. Chem. Soc. 103, 2182 (1913). 3 J. pharm. chim. 17, 198 (1918). *Bcrl. klin. Wochenschr. 52, 322 (1915). 8 Arch. Inf. Med. U, 897 (1914). Tsakalotos, J. pharm. chim. (7) 15, 19 (1917). 7 Chem. Abs. 9, 1970 (1915). "ScMmmel & Co. Semi-Ann. Rep. 1910 (2), 170. "Go**, chim. Ital. 38 (2), 1 (1908). *Therap. Monatsch. 20, 573 (1907). 596 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS pharmacological actions of the two forms of camphor are different but in their paper reference is made to trials in the Charlottenburg hospital, of synthetic and natural camphor and no difference was noted between the two when used internally or externally. However, the use of synthetic camphor for internal use was not approved dur- ing the war by the German advisory committee on medical affairs. It is difficult to draw any definite conclusion from these contradictory findings other than that the experimental methods employed must, in many cases, have been exceedingly crude. Borneol produces physio- logical effects very similar to camphor but less pronounced, which would seem to indicate that the oxidation of borneol in the body is not rapid. The action of fenchone is very similar to that of cam- phor. 21 Camphoric acid has the same antiseptic properties as cam- phor, but is much less stimulating. Large doses can be tolerated without danger and it has been recommended for bronchial and lung infections, including pneumonia, but clinical results are not yet avail- able. Monobromocamphor and the two known isomeric monochloro- camphors have properties not differing materially from camphor; these halogen derivatives are unusually stable. In the attempt to get the same physiological action of camphor more promptly by means of a more soluble substance, the physiological properties of oxy- camphor, obtained by reduction of camphorquinone CO CHOH C 8 H 14 <| -> C 8 H 14 <| CO CO was tried, but this substance was found to produce effects dia- metrically opposite from camphor; whereas camphor stimulates the central nervous system, oxycamphor depresses excitation of the re- spiratory center and is accordingly a rapidly acting drug in dyspnoea. 22 Aminocamphor and bornylamine retard the heart action. None of the numerous derivatives of camphor have been found to possess properties on the whole equal to camphor. The anesthetic action of a large number of volatile chlorinated hydrocarbons has been described. 23 Ethyl chloride, butyl chloride and amyl chloride are said to be dangerous. Carbon tetrachloride is very slow in its action and gives a prolonged anesthesia in which the convulsive stage is apt to be of long duration and acute. It is also 21 Arch. eap. Path. Pharm. 50, 199 (1903). 22 Cf. Frankel, Arzneimittel Synthese, Ed. 3, 1919, p. 748. 23 Cf. Gwathmey, Anesthesia, 1914 ; Frankel, Arzneimittel Synthese, E<3. 3, 1919. PHYSIOLOGICAL AND RELATED PROPERTIES 597 more toxic and irritating to the mucous membranes than chloroform. Of the many chlorine derivatives whose anesthetic action has been more or less carefully determined, methyl chloroform CH 3 CC1 3 , and dichloroethylene CHC1 = CHC1, appear to be the most promising, 24 but none have been clearly shown to be as satisfactory as chloroform, although it is generally recognized that it would be desirable to dis- cover a satisfactory anesthetic of the character of chloroform but which would not form the toxic product phosgene, to which many of the fatalities under chloroform anesthesia, have been attributed. It is beyond the purpose of the present monograph to review the whole field of physiological action and chemical structure, and is also foreign to the author's experience in research. These selected notes are therefore included which seem to bear upon the general thesis of the present volume. Diethyl ketone, C 2 H 5 CO.C 2 H 5 , has pronounced soporific properties and has been recommended 25 as an inhalation anesthetic but seems to possess no particular merit. Cyclopentanone, cyclohexanone and cycloheptanone also have pronounced sleep producing power. Cyclohexylamine and n.hexylamine have practically identical effects upon the blood pressure 26 (normal hexylamine has the most pronounced effect of any of the normal primary amines). That the physiological properties of comparatively simple deriva- tives of the aliphatic hydrocarbons are very imperfectly known has been strikingly demonstrated during the recent war when three of the war poisons of greatest value proved to be substances of this class. The physiological properties of only one of these was com- paratively well known, i. e., phosgene. Trichloromethyl chlorofor- mate and mustard gas [|3,(3-dichloroethyl sulfide] were very imper- fectly known before the war. The latter substance is an excellent illustration of the fact that a substance may have a very mild, sweet, rather agreeable odor and yet be exceedingly irritating, this sub- stance killing tissues with which it comes in contact and, within a few hours after exposure, develops most painful blisters. Halogen derivatives adjacent to a carbonyl group are exceedingly irritating to the eyes and mucous membranes, as illustrated by thje lachrymatory "gases" trichloromethyl and monochloromethyl chloroformates, C1.C0 2 CC1 3 and C1.C0 2 CH 2 C1, acetone derivatives CH 3 COCH 2 Br, CH 2 C1.COCH 2 C1 and the like, bromoacetic ester CH 2 Br.C0 2 C 2 H 5 , ** Wittgenstein, Arch. exp. Path. Pharm. 83, 235 (1918). "Ann. chim. farmac. 1892, 124, 225. Cf. Abelons & Bardier, J. phj/s. 1909, 34. 598 CHEMISTRY OF THE NON-BENZENOID HYDROCARBONS bromo acetophenone C 6 H 5 COCH 2 Br and the like. A case of deep destruction of the tissues was described by Bogert 27 as a result of brominating in acetic anhydride. The tissues were killed nearly to the bone, under unbroken skin. This substance would seem to be more destructive to the deeper tissues than mustard gas but we have no data as to effective concentrations, or, in fact, the identity and nature of the toxic substance. "J. Am. Chem. Soc. 29, 239 (1906). SUBJECT INDEX Abietic acid, 417, 418, 419 Absorption spectra, of hydrocarbons, 548, 549 Acetylene, condensations of, 35 formation from ethylene, 35 isopropyl, 212 partial hydrogenation, 161 Acid sludge, nature of, 131 Acrolein. reaction with bisulfite, 144 Adsorption, coloring matter from oils, 589 gasoline vapor in charcoal, 589 selective of oils, 589 Alcohols, catalysts for dehydration, 155, 156 decomposition to olefines, 155 ethyl, from coal gas. 166, 167 preparation from alkyl sulfuric esters, 127, 128 Aldehydes, electrolytic reduction of, 50 Alkyl* halides, conversion to acetates, 74 halides, conversion to hydrocarbons, 4.", halides. dissociation of, 72, 73 halides, preparation of, 63, 65, 66 halides. reaction with alcoholic alkali, 73, 152 halides, reaction with alkali formates, 75 halides. reaction with aluminum chlor- ide, 70 halides, reduction by hydriodic acid, 47 halides. reduction by metallic couples, 45. 48 sulfates, reaction with alcohol alkali, 74 sulfuric esters, 126, 127, 128 sulfuric esters from petroleum refining, 103. 106 sulfuric esters in refined oils, 132 Allene, polymerization of, 43 dimethyl, 212, 224 Allo-ociraene, 182 Allyl alcohol, reaction with bisulfite, 143 b'orneol, from camphor, 490 bromide, 126 halides, conversion to ethers, 74 Allylene, physical properties, 540, 548 Aluminum bromide, action on hydrocar- bons, 45 chloride, action on hydrocarbons, 45 chloride, as a refining agent. 109 p.Amidobenzoic acid, hydrogenation, 283 Amidocamphor, 484 2-Amidocyclohexane carboxylic acid, 283 Amines, alicyclic, reaction with nitrous acid. 525 physiological action of, 597 Aminocyclohexane, 290 Amorphous wax, 22 Amyl acetates, from chloropentanes, 74 Amylene, 172 anesthetic action of, 594 autoxidation of, 134 conversion to amyl methyl ether, 131 effect in synthetic rubber, 217 oxide, 223 polymerization by fullers' earth, 32, 43 Amylene, preparation from amyl alcohol, 173 preparation from chloropentanes, 153 preparation from n.hexane, 35 preparation by pyrolysis of oils, 27, 36, 37 reaction with acetic acid, 125 reaction with aluminum chloride, 45 reaction with halogen acids, 67, 127 reaction with sulfur dioxide. 144 reaction with sulfuric acid, 126. 131 Amylenes, rearrangements, 72, 150, 217 Amylene. solubility in aniline, 583 solubility in liquid SO 2 , 586 Anesthesia, by inhalation of hydrocar- bons, 593. 594 Aniline, hydrogenation of, 283 reaction with ketones, 378 solvent for hydrocarbons, 583 Anthracene, hydrogenation of, 283, 284, Anthranilic acid, hydrogenation of, 283 Apocamphoric acid. 442, 447 Ascaridol, constitution of, 347, 348 medicinal value of, 594 Asphalt, artificial, 52 artificial, effect of sulfur on, 59, 60 decomposition of, 23 formation of. 32 Autoxidation, effect of light on, 134 of unsaturated hydrocarbons, 133 Azelaic acid, conversion to cyclo-octane, 520 Azocamphor, 484 Azulene, 549 Baeyer's stress theory, 111, 112 stress theory as modified by Thorpe and Ingold. 116, 117 Beckmann rearrangement of menthone oxime, 363 Benzene, absorption spectrum of, 549 detection in and isolation from petro- leum oils, 20 distillation with n.hexane, 20 in oil gas, 37 hydrogenation of, 242 preparation from cyclohexane, 43, 44 Benzocycloheptanone, 237 Benzoic acid, hydrogenation of, 282 Benzylidene menthone, 366 Bertram and Wahlbaum reaction, 125 Bicyclohexane, dimethyl, 249 Bicyclononane, 243 Bihydrolaurolactone, 475 Bis-nitrosates, 146 Bis-nitrosites, 146 Bis-nitrosochlorides, 146 Bisabolene, 314 Blau gas, composition of, 40 Bleaching of mineral oils. 52, 54 Boiling point, effect of ring closing on, 547 relation to density, 542 normal and cyclic hydrocarbons, 547 paraffines, 544 Borneo camphor, 494 0-Borneol, 474, 475 599 600 SUBJECT INDEX Borneol, cataljjtlc (dehydrogenation of. 506 conversion to camphene, 532, 534 occurrence, 494 oxidation to camphor, 505, 506 physical properties, 503 physiological action, 596 preparation from bornyl chloride, 502 preparation from turpentine, 501, 502 structure of. 503 Bornyl chloride, 438, 440, 453, 494 chloride, conversion to camphene, 153, 499 chloride, physical properties, 497, 498 chloride, preparation from turpentine, 496, 510 chloride, reaction with alkalies, 462 chloride, reaction with lime, 502 chloride, reaction with magnesium, 502 chloride, relation to fenchyl chloride, 508 iodide, 438, 453 Bornylene, 453 hydrogenation, 463 preparation and physical properties, 455, 456, 504 reaction with diazoacetic ester, 459 Bornylene carboxylic acid, 455 Bornylene glycol, 486 Bornylene-3-hydroxamic acid, 474 Bornyl magnesium chloride, 502 Bromination by n.bromoacetamide, 123 Bromocyclohexane, Grignard reactions of, 287 Buchu camphor, 371, 372 Butadiene caoutchouc, 211 dimeride of, 228 physical properties, 231 preparation from acetaldehyde, 221, 222 preparation from acetone oil, 221 preparation from butyl alcohol, 219 preparation from butyric aldehyde, 221 preparation from cyclohexane, 219 preparation from hexane, 36 preparation from petroleum, 42, 216 dimethyl, 223, 226, 231 dimethyl, dimeride, 227 dimethyl, preparation from pinacone, 156 Butane, chlorination of, 87 from natural gas, 14, 86 synthesis, 48, 86 vapor pressure of liquid, 17, 85, 88 2-bromo-3.3-dimethyl , 72 2.2-diraethyl , nitration of, 61 2-ethyl , oxidation of, 57 2.2.3,3.-tetramethyl , 93 Butanes, chemical properties of, 172 dichloro, 219 physical properties, 98 Butene, ozonide, 141 Butene (1) 2.3-dimethyl , physical prop- erties, 206 3.3-dimethyl , 206 Butene (2), preparation, 153 2.3-dimethyl, 206 Butene (3), 2.2.3.-trimethyl , 206 Buetene (4). 2.2.-dimethyl , 206 Butenes, physical properties of, 171 preparation from butylalcohols, 150, 156 reaction with sulfur dioxide, 144 reaction with sulfuric acid, 126 Butenes, stereoisomeric dibromo , 186 n. Butyl alcohol, by fermentation, 86 alcohol, for rubber synthesis, 219 Butyl alcohol, tertiary, 70 bromides, preparation, 67 Butyl iodide, tertiary, reaction with sil- ver nitrate, 74 glycol, 221 Cadinene, 412, 413 California petroleum, nitrogen in, 29 Caouprene bromides, 211 Camphane, 463, 464 Camphanic acid, 469, 481 Camphene, 453 acetylation of, 510 2-bromo-12-bromo , 465 chlorohydrine, 124, 465 conversion to isobornyl acetate, 462, 463 glycol, 465 hydrate, 444, 459, 502 hydrochloride, 444, 459, 461, 462 hydrogenation of, 464 impurities in, 508 molecular refraction, 561 nitrosite, 464, 465 oxidation products, 457 ozonide, 139 physical properties, 454, 505 preparation from borneol, 534 preparation from bornyl chloride, 153, 510 preparation, industrial, 499, 500 reaction with acetic acid, 461 reaction with bromine, 465 reaction with diazoacetic ester. 458 reaction with formaldehyde, 465 reaction from hydrogen chloride, 462 reaction with hypochlorous acid, 465 Camphenic acid, 457, 458 Camphenilane aldehyde, 466 Camphenilanic acid, 466 Camphenilol, 398, 466 Camphenilone, 139, 456, 466 conversion to santene, 398 reaction with methyl magnesium chloride, 460 Camphenonic acid, 458 Camphocarbonic acid, 487, 488 conversion to bornylene, 504 Campholactone, 475 Campholide, 471, 474 Campholytic acid, 475, 476 rearrangement of, 482, 483 Camphonanic acid, 476 Camphonene series, nomenclature of, 475, 476 Camphonenic acid, 475, 476 constitution, 478, 479 Camphonolic acid, 480 Camphonololactone, 476 Camphononic acid, 469 constitution of, 479, 480 Camphor, acid addition products of, 467 alkylation of, 491 "artificial," 494 benzylidene compounds, 467 Borneo, 494 p.bromophenylhydrazone, 467 buchu, 371, 372 carvone-camphor, 377 constitution, 467 et seq. conversion to camphane, 463 derivatives of, 484 derivatives of, physiological action, 596 glycol, 486 hydrazone, decomposition of, 461 identification of, 467 imide, 490 impurities in synthetic, 507 impurities in natural, 510 natural, 466 natural, production of, 492 oxidation products, 467 et seq. oxime, 467, 484 SUBJECT INDEX 601 Camphor, physical properties, 466 physiological action, 594, 595 preparation from borneols, 505 preparation from camphoric acid, 471 quinone, 485 quinone, eight oximino derivatives of, 486 reaction with mercuric iodide, 467 synthetic, 492 et seq. synthetic, purification of, 509 synthetic, yields of, 506. 510 0-Camphor, see Epicamphor Camphoric acid, anhydride, 491 conversion to camphor, 471 iso. 472 Komppa's synthesis, 469, 470 Perkin's synthesis, 470, 471 physiological action. 596 preparation and properties, 489 preparation from bornylene, 456 stereochemistry of, 472 Camphoric acids, nomenclature of, 477 Camphoronic acid, 468, 469, 482 Camphoroxalic acid. 489 Camphorphorone, 370 conversion to l-methyl-3-isopropylcyclo- pentane, 490 preparation and properties, 489, 490 Camphyl glycol. 488 Cantharene, 310 Carane. from pulegone, 241, 372 physical properties, 553 Carbides, formation from methane, 79 Carbon black, yield from natural gas, Carbon, colloidal, in oils, 105 Carbon tetrachloride, from methane, 79, 80 Carburetted water gas, composition, 40 Carbyl sulfate, 143 Carene, 403 Carone, 384 Caronic acids, synthesis. 248, 386 cis and trans, 117, 385 Carvenene, 340 Carvenone. 339 Carveol. 356 Carvestrene, 384 Carvomenthene, from limonene, 318 Carvomenthol, 356 Carvomenthone. chlorination of, 365 rearrangement of, 527, 528 Carvone, 318 Carvone-anil, 378 Carvone camphor. 377 constitution, 327, 328, 356 conversion to 2-methylcymene, 376 conversion to eucarvohe, 373, 517 conversion to sylvestrene, 384 hydrogenation of, 374 isomerization by sunlight, 377 occurrence. 375 oxidation of, 328 preparation from pinene, 431. 432 preparation from a-terpineol, 329 reaction with aniline, 378 reaction with hydrogen cyanide. 378 reaction with hydrogen sulfide, 375 reactions of. 375 Carvotanacetone, 356, 375 from a-phellandrene, 380 Carvoxime, 317, 327 conversion to amidothymol, 375 hydrogenation, 375 Caryophellene, 409. 410, 411 0-Caryophellene, 411 7/m-Caryophellene, 412 terp-Caryophellene, 412 Castor oil, solubility in hydrocarbon oils, 582 Cedrene, 416 Cedrone, 416 Ceresine, 18, 22, 96 Chlorinated hydrocarbons, physiological action of, 596, 597 Chlorine, substitution, effect on melting point, 545 Chlorocosane, 102 Chlorohydrines. 123, 124 Cholesterol, 205, 522 decomposition products, 31 decomposition products, optical ac- tivity, 565, 566 Cholesterylene, 205 1.4-Cineol, 336 physical properties, 347 1.4-Cineolic acid, 348 1.8-Cineol, 336 addition products, 345 conversion to terpin-diacetate, 345 isolation of, 347 ketone derivative. 347 occurrence and properties, 344, 346 oxidation, 346 1.8-Cineol. meta, 389 1.8-Cineolic acid, 346 Citral. chemical properties, 200 constitution. 185 conversion to decane, 199 conversion to ionones, 200, 201 hydrogenation, 199 hydrolytic decomposition, 184 physical properties, 197 reaction with alkyl magnesium halides, 204 reaction with ethyl acetoacetate, 204 reaction with j8-naphthylamine. 200 reaction with sodium bisulfite, 144, 145, 198, 199 relation to geraniol and nerol, 186 stereochemistry of, 186 Citronellal, constitution of, 189 conversion to isopulegol. 192, 238 reaction with bisulfite, 145 Citronellic acid, 190 Citronellol, constitution of, 189 physical properties, 197 relation to rhodinol, 190, 191 stability of, 194 Coal, carbonization at low temperatures. 23, 34, 41 hydrogenation of, 284 Coal gas. composition of, 40 Cocain, decomposition products. 512 Color, of hydrocarbons, 548, 549 Coloring matter, of petroleum distillates, 104 Conylene. 179 Copaene, 417 Crithmene, 340. 341 Critical pressures, of hydrocarbons. 548 Critical temperatures of hydrocarbons, 548 Crotonic aldehyde, reaction with bisulfite, 144 Cvclic hydrocarbons, viscosities of, 577. 578 Cyclic structures, relative ease of forma- tion, 116 Cyclobutane, angle of strain (Baeyer), carboxylic acid, 235, 251, 257 derivatives, comparison with n.butane derivatives, 251, 247 derivatives, molecular refraction, 555 1.2-dibromo , 252 1.1-dicarboxylic acid, 256 1.3-dicarboxylic acid. 257 1.2-di-isopropyl , 256 1.2-di-isopropylidene , 256 602 SUBJECT INDEX Cyclobutane, l.l-dimethyl-2-methylene-3- isopropyl , 256 ethyl , 115, 255 methene , 253 methyl, 234, 254 nitromethyl, 525 preparation and properties, 251 preparation by polymerization, 212 stability of the ring, 114, 252, 429, 430, 433 1.1.3.3-tetramethyl-2.4-diethyl , 115 1.1.2.2-tetracarboxylic acid, 257 1.1.2-trirnethyl-3-isopropyl , 256 Cyclobutanol, 251 Cyclobutanone, 255 2-isopropylidene ., 256 2-isopropyl, 256 Cyclobutene, bromination of, 255 from cyclobutylamine, 252 synthesis, 156 Cyclobutylamine, reaction with nitrous acid, 525 Cyclobutylcarbinol, conversion to cyclo- pentyl bromide, 524 dimethyl , rearrangement, 524 Cyclobutylemethylamine, conversion to cyclopentanol, 525 Cyclobutylmethyl carbinol, 255 Cyclocamphane, 464 Cyclocamphanol, 464 Cyclocamphanone, 464 Cyclocampholenic acid, 464 A*-Cyclocitral, 202 Cyclofenchene, 439 Cyclogeraniolene, 180 Ai-s-Cycloheptadiene, 511, 512, 513 Cycloheptane, angle of strain (Baeyer), 114 from cyclohexanes, 530 from petroleum, 24 preparation and properties, 511 ' 1.2-dibromo , 512 1.2-dimethyl-1.2-dihydroxy, 239 methene, 516 1.2.4-tricarboxylic acid, 515 Cycloheptanol, 511 1-methyl, 516 acetic acid, 516 Cycloheptanone, 233, 515 bromination, 516 Cycloheptatriene, dimethyl, from p.xylene, 514 Cycloheptatrienes, preparation, 514 Ai-a-s-Cycloheptatriene, 511, 513 A 2 - 5 - T -Cycloheptatriene, 2.5-diinethyl-7-car- boxylic acid, 514 Cycloheptene, 511, 512 A^Cycloheptene methyl, 516 Ai-Cycloheptenone(3),-2 methyl, 516 Cyclohexadienes, absorption spectrum, o49 preparation and properties, 291 reaction with sulfuric acid, 293 Ai-3-Cyclohexadiene, 291, 292, 557 1.3-dimethyl. 238 A^-Cyclohexadiene, 291 1.4-di-isopropenyl, 309 Cyclohexane, 24, 38, 280, 285 Baeyer's synthesis, 234 bromination, 286 chlorination, 288, 289 conversion to benzene, 34, 43, 282 conversion to cyclopentanes, 368 conversion to methylcyclopentane, 524 critical temperature and pressure, 548 cryoscopic solvent, 583 dehydrogenation, 43, 44, 282 derivatives, stereochemistry, 279 detection of, 2S2 effect of ring closing on molecular re- fraction, 556, 557 Cyclohexane, halogenation, 64 identification of alkyl derivatives, 303 isolation from petroleum, 18 metal derivatives, 287 nitration, 289 preparation, 234, 235, 242, 280, 283 pyrolysis of, 36, 285 separation from methyl-cyclopentane, 43 Cyclohexane, ethylidene, 308 hexacarboxylic acid, 281 hexol, 294 iodo, 289 ' methene, 298, 299 methyl, 296, 297 methyl, chlorine derivatives, 297 methyl, nitration of, 289 methyl, 1.1-nitro, 297 methyl, 1.3-nitro, 297 methyl, reaction with sulfur, 59 pentol, 294 phenyl, nitration of, 611 solubility in liquid SO 2 , 586 tetrachloro , 289 trichloro-, 122 Cyclohexanes, alkyl, conversion to ben- zene derivatives, 303 alkyl derivatives, table of physical properties, 301, 303 amino, 283, 290 bromo, conversion to cyclohexanol, 75 carboxylic acid, 235, 282 carboxylic acid, 4-amido, 283 1.1-diacetic acid, 117 diamino, 290 1.2-dicarboxylic acid, 282 1.3-dicarboxylic acid, 236 1.4-dicarboxylic acid, 279 dichloro, 219 1.4-di-isopropyl , 308 dimethyl, 243 1.1-dimethyl , 300 1.2-diol, 294 1.3-diol, 294 1.4-diol, 234 1.4-dione, 233, 280 1.3-diones, 280 Cyclohexanol, decomposition of, 288 preparation and properties, 75, 293 Cyclohexanol ( 1 ) ,-2.2-dimethyl, 533 -1-methyl, 298, 299 -1-methyl, oxidation of, 254 Cyclohexanols, table of physical prop- erties, 295 Cyclohexanone, 233, 234, 296 -4-carboxylic acid, 321 chlorination of, 64 conversion to cyclopentanones, 116, 370, 371 dimethyl, from tetrahydro-eucarvone, 517, 518 preparation from nitrocyclohexane, 289 4-propyl , one (3), 367 reaction with bromoacetic ester, 298 reduction to cyclohexanol, 285 Cyclohexene, absorption spectrum, 549 ' catalysis of autoxidation, 134 conversion to butadiene, 217, 218 ozonide, 142 preparation and properties, 288 preparation from amino-cyclohexanes, 290 preparation from cyclopenthylcarbinol, 526 reaction with acetyl chloride, 121 Cyclohexenes, table of physical properties, 305 A^Cyclohexene aldehyde, 299 A 2 -Cyclohexene carboxylic acid, 282 A^Cyclohexene-l-methyl^-carboxylic acid, 321 SUBJECT INDEX A 3 -Cyclohexene-l-methyl-4-carboxylic acid, 332 A s -Cyelohexene-l.l-dimethyl, 300 Ai-Cyclohexene-1.2-dimethyl, 307, 536 A3-Cyclohexene-1.3-dimethyl, 238 A^Cyclohexene-l^-dimethyM-isopropyl, Ot>3 A 8 -Cyelobexene 1.3-dimethyl-3-ethenyl, 228 A*-Cyclohexene-l-ethenyl, 228 A-Cyclohexene, methyl, 299 Cyclohexenes, methyl , preparation, 297, 298 Cyclohexyl aldehyde, 299 Cyclohexylmenthene, 360 Cyelohexylmethylamine, conversion to cy- clopentanol, 525 Cyclohexylnitromethane, 297 Cyclo-isocamphoronic acid, 464 Cyclo-octane, 519, 520 Cyclo-octadiene, A 1 - 3 , 520 A*-*, 520 A*-", 520 A 1 -*, ozonide of, 140 A-, -3-4-dimethyl, 520 Cyclo-octatetrene, 521 molecular refraction, 560 Cyclononane, 521 Cyclopentadiene, bromination of, 260 dimeride, 260 preparation, 260 reactivity of the CH 2 group, 261 reaction with ketones, 261 reaction with sulfuric acid, 261 4-methyl-2-ethyl, 261, 262 Cyclopentane, 38, 258 angle of strain (Baeyer), 114 dehydrogenation of, 43 relative ease of formation, 116, 240 1.2-diearboxylic acid, 276 dimethyl, 38 dimethyl, from cycloheptyl-iodide, 524 1.1-dimethyl, 270 1.2-dimethyl, 271 1.3-dimethyl, 272 1.2-dimethyl-3-isopropyl, 272 ethylidene, 266 isopropylidene, 267 methene, 265, 266 methyl, 262 methyl from benzene, 258 methyl from cyclohexane, 36 methyl from cyclohexanol, 258 methyl petroleum, 24, 38 l-methyl-2-carboxylic acid, 275 methyl, nitro derivatives, 262, 263 l-methyl-2.3-dicarboxylic acid, 277 l-methyl-3-ethyl, 272 l-methyl-2-isopropyl, 272 l-methyl-3-isopropyl, 490 l-methyl-3-methene, 269 l-methyl-2-nitro, 262 1.2.4-tricarboxylic acid, 277 1.1.3-trimethyl, 408 sulflnic acid, 75 sulfonic acid, 259 Cyclopentanol, 259 from cyclobutyl-methyl amine, 525 (l),-l-methyl, 236 (2),-l-methyl, 262 Cyclopentanone, 233, 267 conversion to amidocapronic acid, 270 Cyclopentanone, enolization of, 264 preparation, 116, 239, 264 preparation from cyclohexanone, 370 preparation from 1.3-dibromo-cyclohexa- none, 536 reaction with aldehydes, 265 reaction with alkyl magnesium halides, 266 Cyclopentanone, reaction with bromoacetic ester, 267 reaction with formic acid, 265 2.5-dimethyl, 264 2-ethyl, 264 isopropylidene, 265 2-methyl, 262, 268 3-methyl, 268, 269 Cyclopentanone (2), 1 methyl-3-isopropyl, 370 sulfonal, 265 Cyclopentene, cyclopentyl, 267 dicyclopentyl, 268 iodohydrine, 276 oxide, 276 ozonide, 141 preparation and properties, 259 A 2 Cyclopentene, 1.1-diethyl, 524 A Cyclopentene, 1.2-diethyl-, 116, 524 A* Cyclopentene, 1.2-dimethyl, 271 A* Cyclopentene, 1.1-dimethyl, 270, 271 AI Cyclopentene,isopropyl, 267, 533, 534 A* Cyclopentene,2-methyl. 269 Cyclopentenes, table of physical proper- ties, 274 Cyclopentyl carbinol, 526 methylamine, 525 Cyclopropane, 234, 247 angle of strain (Baeyer), 114 derivatives, Kishner's synthesis, 241 derivatives, molecular refraction, 553 derivatives, synthesis by diazomethane, 239, 242 preparation and properties, 247 ring as intermediate in rearrangements, 71, 526 ring in tricyclene, 444 ring rupture in sabinane, 526 ring rupture in thujone, 400. 401 ring, stability of, 67, 114, 245, 246 Cvclopropane, dicarboxylic acid, 235 .1-dimethyl, 247, 248 .2-dimethyl, 114, 248 .l-dimethyl-2-carboxylic acid, 115 .l-dimethyl-2-isobutenyl, 115, 241, 250 ethyl, 254 methyl, 247 -methyl-1.2-diethyl, 249 -methyl-2-isobutyl. 249, 409 methylisopropyl, 241. 249 methyl, 1.2.3-tricarboxylic acid, 434 nitro derivatives, 247 phenyl, 241 1.2.3-tricarboxylic acid. 459 1.1.2-trimethyl, 114. 248 1.2.3-trimethyl, 248 Cyclopropyl ethyl ketone, 246 methyl ketone, 236 pp-Cymene, hydrogenation of, 243 Deblooming reagents, 105 Decadienes, 181 Decahydronaphthalene, 242 in petroleum, 24 n.Decane, 99 Decanes, 94, 99 Decatrienes, 181 Decene, from undecylenic acid, 151 Decene(l), 207 reaction with benzoyl peroxide, 135 Decolorizing of petroleum oils, 105 Dehydrocamphoric acid, 476 Dehydrogenation, 43 Demethylated pinone, 233 Density, relation to boiling point, 542 relation to molecular volume, 538, 540 Desulfurizing of petroleum oils, 28 Diallyl, polymerization of, 43 Diaminocyclohexanes, 290 604 SUBJECT INDEX Dibromocyclic ketones, rearrangement, o27, 528 Dicamphyl ethane, 489 Dichlorocyclohexanes, 289 Dichloroethylene, as an anesthetic, 597 &3-Dichloroethyl sulfide, 164 physiological action, 597 Dichloropropyl sulfide, 170 Dicyclobutane, derivatives of 250 Dicyclohexane, 2.6.6-trimethyl 409 Dicyclohexylamine, 283 aa-Dicyclohexylethane, 405 Dicyclohexylpropane, 243 Dielectric constants, 97, 576 Dienes, absorption spectra of, 549 conjugated, heats of combustion, 572 conjugated, odor of, 591 conjugated, polymerization, 212 conjugated, physical properties, 231, 292, 5o6 conjugated, reaction with sulfuric acid, 132 conjugated, refractive index, 292 556 Diethylene oxide, 342 Dihydrocamphorphorone, 371 Dihydrocarveol, 129, 327, 355, 356 Dihydrocarvone, 356 Dihydrocedrene, 416 Dihydrocitral (see Citronellal) Dihydrocuminaldehyde, 381 Dihydroeucarvone, 374, 518 Dihydrolinalool, 196 Dihydrolimonene, 48 Dihydromyrcene, 181, 196 Dihydroperillic alcohol, 379 Dihydropinolone, 350 Dihydrosylverterpineol, 391 Di-isobutene, oxidation of, 135 Di-isobutyl, 92 Di-isoprene, 227 Di-isopropyl, nitration of, 61 critical constants, 548 p-Diketocamphane, 486, 487 2.2-Dimethyl-l-chloropropane, rearrange- ment of, 463 3.3-Dimethyl- (0.1.3 )-dicyclohexane, 408 Dimethyl granatinine, 519 Dimethyl norcampholide, 456 Diosphenol, 372, 373 Dipentene (see also Limonene) dihydrochloride, 507 occurrence, 315 physical properties, 317 pyrolysis of, 216 synthesis of, 321, 323 Diphenylamine, hydrogenation of, 283 Dispersion, molecular, 562 molecular effect of ring closing, 554 n.Docosane, 100 n.Dodecane, 99 Dodecene(l), physical properties, 208 Dotriacontane, 19, 101 Drying oils, polymerization of, 213 Dutch liquid, 165 Edeleanu, method of oil refining, 109 144 586 n.Eicosane, 100 Emulsions, 587, 588 Epiborneol, 474, 475 Epicamphor, 473, 474, 475 Essential oils, paraffines in, 95, 96 Ethane, 82 et scq. chlorination, 84 occurrence in natural gas, 13, 14, 83 oxidation, 57 physical properties, 17, 83, 540, 548 preparation from ethylene, 35 Ethane, preparation from sodium acetate, 50 reaction with ozone, 142 separation from methane, 83 vapor pressure of liquid, 17, 83, 84 Ethanol mercury salts, 168 Ether formation, from alkyl halides, 73 Ethylene, compounds with metallic salts, Io9 compressibility, 159 conversion to acetylene, 35 decomposition by heat, 33, 35, 160 heat of combustion, 571, 575 hydrogenation to ethane, 36, 48 oxidation to formaldehyde, 162 percent in oil gas, 40 physical properties, 84, 158, 540, 548 polymerization, 211 preparation, 125, 155, 160, 161 preparation from ethyl chloride, 73, 153 reaction with benzoyl chloride, 165 reaction with boron trifluoride, 166 reaction with bromine, 165 reaction with chlorine, 165 reaction with iodine, 166 reaction with mercury salts, 168, 169 reaction with nitrosyl chloride, 146 reaction with ozone, 137 reaction with phosgene, 165 reaction with selenium chloride, 164 reaction with sulfur chloride, 164 reaction with sulfuric acid, 166 reaction with sulfur trioxide, 143 separation by mercury salts, 169 solubility, 158, 160, 583 solvent power of compressed, 583 vapor pressures of liquid, 84 Ethylenes, effect of substituents on ad- dition of bromine, 120, 121 substituted, nitration of, 122 substituted, polymerization, 122, 211 Ethylene bond, absorption spectrum, 548, 549 bond, addition of water, 125 bond, angle of strain (Baeyer), 114 bond, chemical properties as modified by substituents, 122 et seq. bond, conjugated, mol. refraction, 556 bond, heats of combustion of hydro- carbons containing, 572 bond, hydration by organic acids, 125 bond, hydration by sulfuric acid, 126 bond, hydrolytic rupture by alkali, 149 bond, molecular compounds with, 120 bond, nature of, 111 bond, reaction in presence of aluminum chloride, 121 bond, reaction with acetoacetic ester, 149 bond, reaction with amines, 147, 148 bond, reaction with bromine, 120, 121, bond, reaction with hydrocyanic acid, bond, reaction with hydrogen sulfide, 148 bond, reaction with hypochlorous acid, 123 bond, reaction with iodine, 123 bond, reaction with metallic sodium, 149 bond, reaction with nitrosyl chloride, 145 bond, reaction with organic acids, 125 bond, reaction with organic peroxides, Io4, 13o bond, reaction with ozone, 137 et scq. bond, reaction with permanganate, 135 bond, reaction with sulfur, 135 SUBJECT INDEX 605 Ethylene bond, reaction with sulfuric acid, 126 bond, reaction with sulfurous acid, 144 bond, reaction with sulfur trioxide, 143 Ethylene bonds, refractivity of conjugated, OOD bonds, refractivity of semi cyclic, 560, 561 bonds, rupture by air oxidation, 134 bonds, thermochemistry of, 575, 576 bonds, type reactions of, 119 Ethylene bromide, reaction with silver' nitrate, 74 Ethylene bromohydrine, 163 chloride, 165 chloride, conversion to glycol, 75 ehlorobromide, 166 chlorphydrine, 163 diamine, 147 oxide, constitution of Thomsen, 569 oxide, preparation, 341, 343 oxide, derivatives, in rearrangements, 531 Ethyl chloride, conversion to ethylene, 73 Ethyl hydrogen sulfate, hydrolysis of, 127 2-Ethyl-p-menthanone, 360 2-Ethyl-menthol, 360 Eucalyptus oils, 346, 415 parafflnes in, 19 Eucarvone, 373, 519 constitution, 517 Eudesmene, 415 Fatty acids, decomposition to hydrocar- bons, 33 oxidation of, 58 from paraffine, 52, 55, 56 from oleh'nes by ozone, 57, 143 reduction to hydrocarbons, 47, 50 salts, electrolysis of, 50 Fatty oils, polymerization of, 213 Fenchane, 452 a-Fenchene, 446, 448, 451 0-Fenchene, 146, 448 7-Fenchene, 449 5-Fenchene, 448 Fenchenes in camphene, 507, 508, 509 Fenchenonic acid, 442 Fenchocamphorone, 447 Fencholic acid, 451 Fenchone, physical properties, 446, 447 synthesis, 449 in synthetic camphor, 507 and isofenchone, relation to fenchyl alcohols and tricyclene, 509 Fenchosantenone, 450 Fenchyl alcohol, 440, 441 alcohol, dehydration of, 451 alcohol, from pinene, 501, 507 chloride, reactions of, 451 chloride, relation to bornyl chloride, 508 Fish liver oils, 205 Fish oil, distillation under pressure, 31 Formaldehyde, from hydrocarbons by partial oxidation, 57 from ethylene, 162 from methane, 78 Formolite reactions, 286, 287 Formylmenthylamines, 364 Fluorenone, reduction of, 530 Fluorescence, of hydrocarbons, 549, 550 of petroleum distillates, 105, 550 Friedel and Craft's reaction, for ring clos- ing, 45, 237 Fullers' earth, filtration of oils through, 589 polymerization by, 32, 43 as a refining agent, 110 Fulvenes, 261 absorption spectra, 549 Fusel oil, 220 Gas, coal, 33 natural, analysis by fractional distilla- tion, 14 natural, composition, 13, 14 natural, fuel value of, 14, 15 natural, origin of, 19 Pintsch, 37 Gases, composition of various industrial, 40 solvent power of compressed, 583 Gasoline, electrostatic charges, 576 hydrocarbons in, 24 oxidation by air, 133 properties of highly "cracked," 42, 43 refining of, 102, 109, 110, 131, 132 removal from natural gas, 14, 16 temperatures for producing by pyrolysis, 36 Geraniol, conversion to dipentene, 187 conversion to linalool, 188 occurrence, 192, 193 odors, 204 oxidation, 188 oxides, 194 physical properties, 187 reaction with bisulfite, 196 relation to citral "a," 186 stability of, 194 Geraniolene, 180 Geranyl acetone, 193 chloride, 193 Glycerine, synthesis from propylene, 86 Glycols, as products of oxidation, 135 preparation from oxides, 342 Greases, calcium soaps in, 587 Grignard reactions, for hydrocarbons, 46, 47, 75 reactions, for ring closing, 234, 236 reactions, on carvone, 376, 377 Haber's methane whistle, 77 Hall refining process, 110 Haller's reactions, on camphor, 491 reactions, on menthone, 360 Heats of combustion, benzene and cyclo- hexenes, 574 CH 2 ;n paraffine series, 574 cyclic hydrocarbons, 573 effect of conjugation of double bonds, 572 in paraffine hydrocarbons, 575 hydrocarbons, table, 571 isomeric substances, 571, 572 unsaturated hydrocarbons, 575, 576 Heats of dissociation of C-C and C-H bonds, 574, 575 n.Heneicosane, 100 n.Hentriacontane, in natural waxes, 19 physical properties, 101 n.Heptacosane, in beeswax, 19 physical properties, 101 Heptadiene A*, 232 A^-3.5-dimethyl, 232 A3'-3-methyl, 232 A^-6-methyl, 232 n.Heptadecane, 100 Heptadecene, from oleic acid, 151 Heptane, bromination, 64 n. Heptane, critical constants, 548 occurrence, 18, 19, 20 physical properties, 99 preparation from cycloheptane, 511 pyrolysis of, 36 sulfonation of, 63 606 SUBJECT INDEX n. Heptane, vapor pressures of, 17 2,6-dimethyl, nitration of, 60 2,6-dione, 215 2-methyl, 91 3-methyl, 91 4-methyl, 92 Heptanes, 89, 91 Heptatriene, A' - 8 - 5 , 232 Heptene(l), 206 -6-methyl, 207 Hepteue(2), 206 -2-methyl, 207 Heptene(3), preparation, 152 -2-methyl, 207 Heptene ( 4 ) ,-4-methyl, 207 3.3.5-trimethyl, 208 Heptene(5)-2-methyl-5-ethyl, 208 n.Heptyl aldehyde, 90 n.Hexacontane, 51 n.Hexacosane, 101 n.Hexadecane. 100 Hexadecene, 208 Hexadiene, A l! *-2.5-dimethyl, ozonide, 140 ozonide of, 140 -2.5-dimethyl, 179 , heat of combustion, 576 , physical properties, 232 -3.5-dimethyl, 232 3-methyl, 232 A3-M-methyl, 232 n.Hexane and benzene, distillation of, 20 critical constants, 548 dehydrogenation, 43, 44 oxidation, 57, 58 in petroleum, 18, 20 physical properties, 99, 548 pyrolysis of, 36, 150 reaction with sulfur, 59 solubility of sulfur in, 583 sulfonation of, 63 synthesis of, 46 vapor pressures, 17 -2-butyl, oxidation of, 57 -2.4-dimethyl, 92 -2.5-dimethyl, 92 -3.4-dimethyl, 93 Iso, in petroleum, 20 iso, physical properties, 99 Hexanes, 89 molecular volume of, 539 n.Hexatriacontane, 101 Hexatriene A , 232 Hexene(l), preparation, 151, 152 physical properties, 206 -3-methyl, 206 Hexene(2), hydrogenation of, 48 Hexene(4), -2-5-dimethyl-4-isobutyl, 208 -4-methyl, 206 Hexenes, in petroleum distillates, 27 reaction with sulfuric acid, 107 Homo-apocamphoric acid, 447 Homocamphor, 491 Homocamphoric acid, 471 Homomenthene, 366, 367 Homoterpenylic acid, 325, 326 Homo-a-terpinol, 353 Hydrocamphorylacetic acid, 491 Hydrocarbons, benzenoid, from oil gas, 37 benzenoid, in petroleum, 26, 38 benzenoid, from petroleum, 36 cyclic, table of physical properties, 25 optically active, in oils, 50 oxidation by nitric acid, 57 oxidation of saturated, 52, 57 oxidation by permanganate, 57 paraffine, table of physical properties, 21, 206 synthesis of optically active, 564 Hydrocarbons, unsaturated, analytical de- termination, 132 unsaturated, conversion to alcohols, 106 unsaturated, determination of constitu- tion, 131, 135, 150 unsaturated, effect of heat and pressure, 42 unsaturated, occurrence in petroleum, 27 unsaturated, odor of, 103 unsaturated, oxidation of, 133, 135 unsaturated, and petroleum refining, 131 unsaturated, polymerization, 210 et seq. unsaturated, preparation of, 150 et aeq. unsaturated, preparation of from amines, 156, 290 unsaturated, preparation by Grignard reaction, 152, 156 unsaturated, preparation from methyl xanthogenates, 154 unsaturated, preparation from palmitic esters, 154 unsaturated, preparation by pyrolysis, 157 unsaturated, reaction with aluminum chloride, 45 unsaturated, reaction with fullers' earth, 32, 43 unsaturated, reaction with nitrosyl chloride, 146 unsaturated, reaction with organic per- oxides, 134 unsaturated, reaction with oxygen, 54 unsaturated, separation from saturated, 109 Hydrogen, selective combustion in pres- ence of methane, 77 Hydrogenation of benzenoid hydrocarbons, 281 by Ipatiev's method, 283, 284 by Skita's method, 285 without a catalyst, 49 Hydrotropilidene, 511, 512 Hypochlorous acid, reaction with defines, 124 Humulene, 412 Ichthyol, 28, 33 Ingold, theory of valence of cyclic hydro- carbons, 117, 118 Inosite, 294 Iodine numbers of unsaturated hydrocar- bons, 123 lonone, ketones related to, 203 preparation, 200, 201 Irone, 202 Isoallofenchene, 448 Iso-amyl-a-dehydrophellandrene, 313 Isoborneol, 460 from camphene, 500, 510 Isobornyl acetate from camphene, 462 Isobornyl chloride, 444 reaction with alkalies, 461, 462 Isobutane, 48, 87 Isobutene, 87 oxide, 342 polymerization, 210 Isobutyl alcohol, decomposition by heat, 70 Isocamphane, 444, 464, 489 Isocampholytic acid, 475, 476, 483 Isocamphorene, 195 Isocamphoric acid, 472 Iso-ethionic acid, 143 Isofenchene, 448, 507 Isogeraniol, 194 Isoheptene(l), preparation, 152 Isohexane, pyrolysis, 36 Isolaurolene, 483 Isolauronolic acid, 475 SUBJECT INDEX 607 Isomenthol, 360 Isonitrosocamphor, 484 iso-octene(l), preparation, 152 Isoparaffines, 22, 23 Isopentane, critical constants, 548 from trimethylethylene, 48 Isopentene, from n.pentane, 217 Isopinene, 441, 448 Isoprene, chemical properties, 177 condensation with limonene, 43 conversion to dipentene, 216 dimerides of, 227, 228 physical properties, 231 polymerization, 215 preparation, 216, 219, 221, 224 preparation from petroleum, 216 reaction with sulfur, 144 separation from hydrocarbon mixtures, Isopropyl alcohol from propylene, 167, 169 Isopulegol, 130, 192, 356 Isopulegone, rearrangement to pulegone, 130 Isothujone, 400, 401 Isozingiberene, 313 Jellies, of mineral oils, 587 Ketene, 125 Ketones, bromination of cyclic, 365 nitration of, 62 physiological action of certain, 597 Kerogen, 23, 96 Kerosene, emulsion of, 588 halogen derivatives, 69 refining of, 102, 131, 132 Kishner's reaction, 366, 372, 409, 452, 463, 464 Krafft's synthesis of unsaturated hydro- carbons, 154 Langmuir, theory of atomic structure and the ethylene bond, 111, 112, 113 Latent heat of vaporization of hydrocar- bons, 567 Laurenone, 481, 482 Laurolanic acid, 477 Laurolecue, synthesis of, 483 series, nomenclature of, 475, 476 Laurolenic acid, 475, 476, 481 Laurololactone, 477 Laurololic acid, 477 Lauronolic acid, 475, 476, 481 Lewis, theory of the ethylene bond, 111, 112 Light, absorption of, by hydrocarbons, 548, 549 Lignite tar oils, action of ozone on, 143 Limonene, condensation with isoprene, 43 constitution, 319 et seq, conversion to isoprene, 216 dihydrochloride, 319 hydrogenation, 48, 318, 319 nitrolanilides 316 nitroso-azide, 318 nitrosochloride, 317 occurrence, 315, 422 oxidation of, 318 ozouide, 142 physical properties, 316, 317 reaction with formaldehyde, 318 rearrangements through addition of wa- ter, 130 relations to a-terpineol and 1.8-terpin, 328 Linalool, chemical properties, 195 constitution of, 188, 189 conversion to terpinene and dipentene, 195 Linalool, occurrence and identification 195 oxidation of, 189 preparation from geraniol, 188 reaction with bisulfite, 196 synthesis of, 189 Linalyl acetate, reaction with sulfur, 196 Liquid petrolatum, 593 Lubrication, by oil films, 579 relation to viscosity, 577 Lubricating oils, chemical character of 522, 523 from coal, 284 from tetrahydronaphthalene, 522 formation by polymerization, 522 oxidation of, 52, 53 refining of, 102, 108 resinification, 53 viscosity of, 578, 579 Magnetic rotation, 562 Marsh gas, see Methane Melting points, 543 effect of structural differences, 545, 546 effect of unsaturation on, 545 of n.parafflnes, 544 o.Menthadiene A J - 8 (), 330 m.Menthadienes, possible. 388 A 2 -( 9 ), 330 *(), 330, 391 p.Menthadiene A 1 * (a-terpinene) 339 A-() synthesis, 329, 330, 331 A- d and I, 564 Menthaue carboxylic acid, 361 2.8-dihydroxy-2-methyl, 376 meta derivatives, 384 ortho derivatives, 392 para, preparation and properties, 319 para, from cymene, 243, 282 Menthanol(3), para, see Menthol Menthatriene A 2 -- 8 ()-2-methyl 376 A 2. e .8 (9) _2-propyl, 292, 293 ' Menthene, para, A s , 364 para, A( 8 ), 364 Menthenols, hydration of, 129 meta, 390 ortho, 392 Menthenol(S), meta A, 387 meta, A', 391 meta, A, 387 para, A 2 , 353 para, A(), 355 para, A(), 356 para, A, 331, 353 Menthanone(5), ortho, 395 Menthenone, A, 355, 367 (2), A, 375 Menthocitronellol, 363 Menthol, alkyl derivatives, 360 catalytic dehydrogenation, 360 crystalline forms of, 357 esters of, 362 occurrence, 357 oxidation, 365 preparation from thymol, 358 preparation from pulegone, 359, 360 stereochemistry of, 360 Menthone, alkylation of, 360, 366 bromination, 365 chemical reactions of, 360 CJ and (ran* forms, 358 conversion to l-methyl-3-isopropyl cy- clopentanone(2), 371 conversion to p-mentnane, 366 conversion to buchu camphor, 371, 372 derivatives, characteristic, 363 electrolytic reduction of, 360 isoxime, 363 nitration of, 62 normal, 367 608 SUBJECT INDEX Merrthone, optical inversion of lavo., 366 rearrangement to cyclo-pentane deriva- tives, 527, 528 stereochemistry of, 361 synthesis, 364 Menthonylamine, 363 Menthyl chloride, 361 Menthyl hydrogen phthalate, resolution of, 359 Menthyl phenyl ether, 361, 362 Menthylamine, 364 Menthylidene hydrazrine, 366 Mesityl oxide, ozonide of, 138 Methane, carbon black from, 78 chlorination of, 79 combustion, mechanism of, 78 compressibility of, 15 conversion to hydrocyanic acid, 81 conversion to formaldehyde, 78 liquefaction of, 76 luminosity of flame, 76 occurrence, 13, 14, 33 oxidation by permanganate, 57 physical properties, 76, 98 physiological effect, 77 preparation from carbon monoxide or water gas, 81, 82 preparation from cellulose, 18, 19, 31 preparation from ethylene, 33, 35 preparation from hexane, 35 pyrolysis of, 35, 78 solubility in oils, 583 solvent power of compressed, 583 warning for , in mine gases, 77 Methyl borneol, 462, 463 caniphene, 462 camphenilol, 460 camphor, 488 carveol, 376 chloride, from methane, 79, 80 chloride, physical properties, 80, 81 chloroform, 597 cyclohexane methyl ketone, 237 cyclohexeues, 276 cyclopentane, isolation of, 18 cyclopentene methyl ketone, 237 (2)-dihydrocarveol, 376 -a-fenchene, 462 fenchyl alcohol, 462, 463 Methyl group, nitration of, 62 Methyl heptenone, condensation of, 238 2-methyl heptene(2)-one(/6), 184 2-Methyl homolimonene, 376 Methyl nopinol, 438 / norcamphor, 450 I pyrollidine, 224 / Mexican petroleum, sulfur in, 28 Molecular dispersion, 562 dispersion, effect of ring closing on , 554 refraction, 550 refraction, effect of ring closing on , 553, 554, 555 volume and density, 538. 539, 540 volume, effect of ring closing on , 541, 542 volume of isomeric hydrocarbons, 539 Montan wax, 23 Myrcene, 182 Myrtenol, 437 Mustard gas, physiological action, 597 see also 0j8-Dichloro-ethyl sulfide Naphthalene, hydrogenation of, 242, 283, 404 in petroleum, 20, 24, 26 Naphthanols, preparation and properties, 405 Naphthanone 405 Naphthenes, 280 in Borneo petroleum, 38 sulfonation of, 63 Naphthenic acid, 56, 273 optically active, 566 removal from petroleum distillates, 104 synthesis of, 275, 276 Naphthols, hydrogenation of, 283 Natural gas, composition, 13, 14 Neomenthol, 359 Nerol, 186, 187 Nickel, effect on pyrolysis of hydrocarbons, Nitration, of non-benzenoid hydrocarbons, 60 to detect benzenoid hydrocarbons, 26 Nitro group, effect on melting point, 546 Nitrocamphene, 465 Nitrocyclohexane, 289 Nitrogen bases, in petroleum, 29, 103 Nitroparaffines, solubility in alkali, 61 Nitrosochlorides, 145 conversion to oximes, 147 method of preparation, 146 Nitrosolimonene, see carvoxime Nitrosopinene, 431 Nomenclature of camphoric acids, 477 of spiro compounds and bridged rings, 405, 406, 407 n.Nonacosane, 101 n.Nonadecane, 100 Nonadienes, 180 Nonanes, 94, 99 Nonene(2), 207 Nonene(4)-4.8-dimethyl, 208 -2.5.8-trimethyl, 208 Nopinic acid, 426, 445 Nopinolacetic acid, 445 Nopinone, 444, 445 conversion to pinene-hydrochloride, 438, 498 Norcamphane, 396 Norcaradienecarboxylic acid, 514 Norcarane, 396 Norpinane, 396 Norpinic acid, 425 Ocimene, 182 Ocimenol, 183 n.Octacosane. 101 Octadiene, 1.6^dimethyl cyclo, 142 A2-*-3.7-dimethyl, 232 A2--3.7-dimethyl, 232 A2-8-2.6-dimethyl, 232 A^-2.7-dimethyl, 232 A-5-4-methyl, 232 As-B-7-methyl, 232 Octahydrindene, 243 Octahydro-anthracene, 284 Octane, 2.6-dimethyl, derivatives of, 183 2.6-dimethyl, oxidation, 57 2.7-dimethyl, nitration of, 61 n. Octane, critical constants, 548 physical properties, 99, 538, 548 sulfonation, 63 vapor pressure, 17 Octanes, heats of combustion of isomeric, 571, 572 synthesis of, 91, 92 molecular volumes of, 539 melting points and constitution, 545 physical properties, 93, 99 pyrolysis of, 36 Octene(l), hydrogenation of, 48 physical properties, 206 -2-methyl, 207 Octene(2), hydrogenation, 48 SUBJECT INDEX 609 Octene(2), physical properties, 207 2.6-dimethyl, 207 2.7-dimethyl, 207 3.7-dimethyl, 207 Octene(4)-4 methyl, 207 4.7-dimethyl, 208 Octenes, from nonylenic acid, 151 reaction with sulfuric acid, 127 Odor, of unrefined petroleum distillates. 102, 103, 591 relation of , to constitution, 203 Oenanthol, 90 Oil gas, composition, 40 butadiene from, 216 chlorination, 165 Dayton process, 40 effect of temperature and pressure on composition, 40, 41 experimental production, 36 isopropyl alcohol from, 167, 169 liquid condensate from, 37 time factor in producing, 36 yields at different temperatures, 38, 39 yields in Hall apparatus, 40 Oiliness. 579 Oklahoma petroleum, benzene, homologues in, 38 Okonite, 96 defines, formation from alkyl halides, 73 in petroleum oils, reaction with sul- furic acid, 106, 107 See also Hydrocarbons, unsaturated Optically active hydrocarbons, 564 Optical activity, 563 Oxidation, of saturated hydrocarbons, 52, 54 Oxides, alkylene, 341, 342, 343 alkylene, formation of, 135 of the terpene series, 341 1.4-Oxidopentane (tetramethylene oxide), 343 1.5-Oxidopentane (pentamethylene oxide), 344 Oxycamphenilanic acid, 465 Oxyfenchenic acid, 447 Oxymethylene camphor, 488 Oxymethylenementhone, 366 Ozokerite, 95, 96 Ozone, reaction with pinene, 502 Ozonides, decomposition of, 137, 138 Paraffine hydrocarbons, occurrence, 13, 16, 19 table of melting and boiling points, 21 Paraffine oil, 104, 105, 593 Paraffine, pyrolysis of, 39 Paraffine wax, 16, 18, 95 chemical properties, 23, 63, 102 composition, 21, 22 crystallization of, 96 dielectric constant, 97 effect of on viscosity of lubricating oils, 578 oxidation, 53, 55, 57 physical properties, 97 reaction with sulfur, 97 solubility in compressed methane, 583 solubility in various solvents, 97, 581 Parafflnes, Ci H22 to CaoH^, tables, 94-97 conversion to benzenoid hydrocarbons, 26 formation, methods of, 33 formation by biological processes, 18 formation, by decomposition of other hy- drocarbons, 42 halogenation, 63, 64 molecular volumes, 539 occurrence in essential oils, 95, 96 Paraffines, physical properties, 98, 99 reaction with sulfur, 58, 59, 97 solubility in liquid SOa, 586 syntheses, 45, 94 viscosity, 577, 578 Paraffinum liquidum, 593 Pennsylvania petroleum distillates spe- cific heats of, 567 Pentacontane, from coal, 19 n.Pentacosane, 101 n.Pentadecane, 100 Pentadiene, A 1 -*, 177 A'-'-S-methyl, 232 Pentamethylene oxide, 344 Pentane, brominatiou, 64 chlorination, 63, 89 from petroleum, 20 phvFical properties, 98 pyrwysis of, 36 separation from isopentane, 18, 87 vapor pressures, 17, 88 Pentane, 1.4-dibromo , conversion to methyl cyclobutane, 234 1.5-dibromo, 234 1.5 dicarhoxylic acid, 280 2.2-dimethyl, nitration of, 61 2-methyl-3-ethyl, 93 Pentane, iso, from petroleum, 20 1*0, physical properties, 98 n.Pentane, conversion to isopentane, 67, conversion to isoprene, 217 critical constants, 548 Pentanes, chloro, conversion to acetates, 74 chloro, decomposition, 153 conversion to rubber, 217, 218 dichioro, 219 dichloro, preparation. 88 Peutatriacontane, 19, 101 Pentene(2), -2.3-dimethyl, 206 -2.4-dimethyl, 206 -3-ethyl, 127, 152, 206 -2-methyl, 206 2-methyl-3-ethyl, 207 3-methyl, 206 Pentene(S), 2-methyl, 206 Pentene(2)-ol (4), 223 Perhydro-anthracene, 284 Perillic acid, 379 alcohol, 379 aldehyde, 379 Peroxides, organic, formation of, 183 Petroleum, filtration through fullers* earth, 30, 589, 590 fluorescence of, 550 optical activity of, 31, 565, 566 origin, 19, 20, 29, 30 pyrolysis of, 34 et seq. distillates, action of ozone, 142 distillates, decolorizing, 105 distillates, hydrogenation, 49 distillates, latent heat of vaporization, 567 distillates, refining of, 102, 103, 107, 108, 126, 131, 586 distillates, refractive indices, 562 distillates, specific heats, 567 distillates, treatment with sulfur diox- ide, 586 distillates, solubility in various solvents, 580, 582 Petroleum jellies, 587 Petroleums, gases dissolved in, 583 Pharmaceutical paraffine oil, 104, 105, 109, 593 Phenanthrene, hydrogenation of, 284 Phellandrenes, 380 et Beq. 610 SUBJECT INDEX a-Phellandrene, conversion to carvotanace- tone, 380 molecular refraction, 559 nitrosite, 381 synthesis of, 381 iS-Phellandrene, 381 Phenols, reduction of, 281 Phenyl ether, hydrogenation of, 284 Phthalic acid, hydrogenation of, 282 Pinacoline alcohol, 460 Pinacoline rearrangement, 528-531 Pinacone, for rubber synthesis, 223, 226 Pinacones, cyclic, 528, 529 intramolecular condensation, 239 Pinane, 429 Pine oil, 421 a-Pinene, constitution, 424 et seq. conversion to carvone, 431, 432 dehydrogenation, 430 dichlorohydrine, 124, 428 hydrochloride, 438, 439, 497, 498 hydrogenation, 429 identification, 424 nitrosoazide, 433 nitrosochloride, 430, 431 occurrence, 423 ozonide, 427, 502 reaction with acetic acid, 501 reaction with benzoyl peroxide, 429 reaction with diazoacetic ester, 434 reaction with hydrogen chloride, 438 reaction with hydrogen peroxide, 429 reaction with hypochlorous acid, 124 reaction with mercuric acetate, 429 reaction with organic acids, 501 reaction with oxalic acid, 126, 439, 501 reaction with ozone, 427, 502 0-Pinene, 444, 445 nitroso , 446 oxidation, 426 synthesis, 154 Pinic acid, 425 Finite, 295 Pinocamphone, 432 Pinocamphylamine, 433 Pinol, 348 et seq. hydrate, 349, 351, 427, 428 oxide, 427, 428 tribromide, conversion to pinolone, 350 0-Pinolene, 439, 441, 442 relation to fenchyl and isofenchyl al- cohols, 509 Pinolone, 350 a-Pinonic acid, 425, 426, 427 Pinononic acid, 435 Pinoyl formic acid, 425 Pintsch gas, 37, 40 Pinylamine, 432, 433 Piperitone, 367, 368 Piperylene, 177, 223, 231 dimeride of, 228 Polymerization, different types of, 229 by surfuric acid, 127 in refining petroleum oils, 106, 137 influence of oxygen, 226 of olefines by aluminum chloride, 45 of olefines by heat and pressure, 42 Polymers, source of, in refined oils, 131 Polynaphthenes, 522 Producer gas, composition of, 40 Propane, 85 in natural gas, 14 physical properties, 85, 98, 540, 548 pyrolysis, 41 synthesis, 48 l-bromo-2.2-dimethyl, 72 2.2-dimethyl, 98 1.2-diphenyl, 62 Propane, 1.2.3-trichloro, 86 Propanol(l)-sulfonic acid (2), 143 Propyl bromides, decomposition, 69 Propylene, 1 and 2 chloro, 86 chemical properties, 169, 170 chlorohydrines, 124, 170 conversion to isopropyl alcohol, 167, 169 hydrogenation, 48 in oil gas, 37, 40 liquid, vapor pressures, 85 physical properties, 540, 548 preparation, 169 reaction with HBr., 68 reaction with hydriodic acid, 66 reaction with nitrosyl chloride, 146 Protoparaffine, 22, 96 Pseudo-ionone, 201 Pseudo-pelletierine, 519 Pulegene, 369, 370 Pulegenone, 370, 372 Pulegoue, conversion to carane, 372 conversion by hydrogenation to men- thols, 359 conversion to A3-p.-menthenol(3), 332 conversion to pulenone, 368 ozonide, 138 physical properties, 368 sulfonic acid derivative, 144 Pulenone, 368, 369 Pyrolysis, effect of pressure, 42 effect of metals, 43, 44 in presence of steam, 44, 45 of petroleum oils, 34, 42, 44 Racemization, of hydrocarbons, 564 Rearrangements, by addition of water, 130 borneol to camphene, 532 of carbocyclic structures, theories of, 526, 528, 529 chloroketones, 527 cyclobutylamine to cyclopropylcarbinol, 525 cyclobutyl-methyl amine to cyclopenta- nol, 525 cyclobutyl-diethyl carbinol to 1.1-diethyl- A 2 -cyclopentene, 524 cyclobutyl-dimethyl carbinol to 1.2-di- methyl-Ai-cyclopentene, 524 cyclobutyl carbinol to bromocyclopen- tene, 524 cycloheptyl iodide to methyl cyclohex- ane, 524 cyclohexane to methyl cyclopentane, 524 cyclohexanes to cycloheptanes, 530 cyclohexanones to cyclopentanones, 536 2.2-dimethyl cyclohexanol(l) to isopro- pyl-Ai-cyclopentene, 534 2-chlorocyclohexanone to cyclopentane- carboxylic acid, 527 cyclohexylmethyl amine to cyclohep- tanol, 525 cyclopentanes to cyclohexanes, 433, 529, 535 cyclopentyl carbinol to cyclohexene, 526 cyclopentylmethylamine to cyclohexanol, 525 cyclopentyl nitrite to nitromethyl cy- clobutane, 525 dibromo cyclic ketones, 527 2.2-dimethyl-l-chloropropane *to 2-me*- thyl-2-chlorobutane, 463 2-iodocyclohexanol to cyclo-pentyl alde- hyde, 526 ethylene oxide derivatives, 531 isoborneol to camphene, 463 menthone to cyclopentanone derivative, 527, 528, 537 pinacoline, 528 SUBJECT INDEX 611 Rearrangements, . pinacoline alcohol to tetra-methyl ethylene, 532, 534, 535 retropinacoline, 463 the Wagner, 528 et seq. Refining of petroleum distillates, 102, 126, 131 effect on specific gravity, 107 by sulfur dioxide, 586 effect of nitric acid, 108 effect of temperature. 108 Reformatsky's reaction, 154, 366 Refractivity, Eisenlohr's revised values, 551 Refractive indices of conjugated dienes, 292 effect of ethylene bond, 551, 552 effect of ring closing, 553-557 exaltation caused by conjugation of CO and ethylene bond, 559 of petroleum distillates, 562 of semi-cyclic double bonds, 560, 561 Retene, 418 Retropinacoline rearrangements, 463 Rhodinol, 190, 191 Ring formation, effect on boiling point, 547 effect on melting point, 545 effect on molecular refraction, 553-557 effect on molecular volume, 541-542 effect on viscosity, 578 methods of, 233 et seq. by polymerization, 212 Rosin oil, 303, 418, 419 Rosin spirit, 419, 421 Rubber, behavior in various solvents, 584 chlorinated, 123 destructive distillation, 216 ozonides, 142. 214 plantation, 214 reaction with sulfur chloride, 136 solutions, viscosity of, 584 synthetic, 214 et seq., 225 Rubbers, synthetic, classification, 215 synthetic, from petroleum, 216 synthetic, vulcanization by sulfur. 136 synthetic, vulcanized, sulfur in, ob4 Sabina ketone, 399 Sabinane, 553 Sabinene, oxidation products, 399 physical properties. 553 sulfonic acid derivative, 144 Santalene, o and /3, 415 from a-phellandrene and isoprene, 416 Santalols, a and j3, 416 Santene, 397 Selinene, 413, 414 Sesquicitronellene, 204 Sesquiterpenes, 310 et seq. Sludge, from petroleum refining, 103 Sobrerol. see Pinol hydrate in suKur dioxide, petroleum fractions in ethyl alcohol and acetic anhydride, 580, 581 solids in compressed gases, 583 Specific heat, 566, 567 Specific viscosity of hydrocarbons, 577 Spectra absorption, of hydrocarbons, 548, 549 Spinacene. 204 Spiro and cyclopropane derivatives, rela- tive stability, 116, 117 Spirocyclene, 252 Squalene, see Spinacene Stearoptene, 19 Styrene, oxidation and polymerization, 134 Suberone, see Cycloheptanone Suberyl alcohol. 511 Sulfonation of hydrocarbons, 63 Sulfonic acids, 144 Sulfur, colloidal, in oils, 105 colloidal, in rubber, 137, 584 in petroleum distillates. 27, 103, 106, 107 derivatives, effect of aluminum chlor- ide on. 45 Sulfur dioxide, liquid, for refining oils, 109, 586 reaction with ethylene bond, 59 reaction with paraffines, 58. 59 solubility in hexane, 583 state of, in vulcanized rubber, 584 Sulfuryl chloride, as a chlorinating re- agent, 165 catalysts for preparation, 366 Sylvestrene. 384 et seq. synthesis, 386, 389, 390 Sylveterpin, 389. 390 Sylveterpineols, 390 Terpenes, absorption spectra of, 549 autoxidation, 134 comparison with aliphatic unsaturated hydrocarbons, 174 et seq. general reactions of, 174 et seq. rearrangements by hydration, 130 Terpenylic acid, 325, 326 Terpin, 1.4, 336, 337 1.8, 319 synthesis, 320, 323 Terpinene, from linalool, 195 o-Terpinene, from thymohydroquinone, 340 molecular refraction, 559 oxidation, 338 transformation to carvenone, 339 7-Terpinene, 339, 340 Terpinenes, 333, 335-337 Terpinenol(l), 355 Terpinenol(4), 338, 354, 355 Terpineols, para, 351 o-Terpineol, chlorohydrinc of, 351 conversion to carvone, 329 conversion to terpin hydrate, 323 constitution, 320, 321 dibromide, 349 identification, 322 occurrence and properties, 321, 322 occurrence in pine oil, 421 preparation, 424 nitrosochloride. 329 oxidation, 324 synthesis, 320 et eq. fl-Terpineol properties and synthesis, 352 y-Terpineol, 334, 335, 353 Terpin hydrate, 323, 324 Terpinolene, 333-335 n.Tetracosane, 100 n.Tetradecane, 100 Tetradecene(l), 208 Tetrahydro-anthracene, 284 Tetrahydrobenzene, see Cyclohexene Tetrahydrocarvone, 356 Tetrahydrocitral, 200 Tetrahydro-eucarvone, 374, 517, 519 conversion to dimethyl cyclohexanone, 374 Tetrahydrogeraniol, 200 Tetrahydronaphthalene, 242, 404 reaction with formaldehyde, 522 Tetrahydrosantalene, 416 Tetrahydroterephthalic acids, 278 Tetrahydro-xylene, meta, 308 Tetramethyl allene, 212 612 SUBJECT INDEX Tetramethylene oxide, 343 Tetramethyl ethylene, 72, 532 physical properties, 206 preparation, 460 reaction with n. bromo acetamide, 123 reaction with sulfuric acid, 126 Tetramethyl methane, 46 Tetratriacontane, 101 Thermochemistry of non-benzenoid hydro- carbons, 117, 568 Thiophanes, 28 Thiozonides, 135 Thujamenthols, 401 Thujane, 402 Thujene, 400, 402 Thujone, 400, 401 physiological action, 594 Thurlow process, for synthetic camphor, 501, 505 Thymol, hydrogenation, 358 Toluene, conversion to methyl-cyclohep- tatriene, 514 from methyl cyclohexane, 43 from petroleum, 37 hydrogenation of, 282 Transformer oils, oxidation of, 52, 53 water in, 53, 54 n.Triacontane, 101 Triazo-ethylene, 163 Trichloro-ethylene, 122 n.Tricosane, 100 Tricyclal, 443 Tricyclene, 439, 443, 444, 460, 461 from fenchyl alcohol, 508 Tricyclenic acid, 443 Tricyclol, 443 n.Tridecene, 99 Trienes, molecular refraction, 557, 558 2.6.6-Trimethyl-0.1.3-dicyclohexane, 409 Trimethylene oxide, 343 Trimethylethylene, 217 hydration of, 125, 131 hydrogenation, 48 oxidation, 135 preparation, 72, 131 reaction with HC1, 67 n.Tritriacontane, 101 Tropidene, exhaustive methylatlon, 512 Tropilidene, 512 Tropine bases, 512 Trouton's rule, 542, 543 Tschugaeff's synthesis of un saturated hy- drocarbons, 154, 455 Tung oil, polymerization, 213 Turpentine, American, 421 autoxidation, 133, 428, 429 conversion to camphor, 495 French, 422 from copals, 422 Greek, 423 optical activity of, 563 physical properties, 422 production, 420 purification, 497 Turpentine, pyrolysis. 216, 430 reaction with hydrogen chloride, 497, reaction with oxalic acid, 501 reaction with sulfur, 135 "recovered," 421 Russian, 422 solubility in alcohol, 582 substitute, 94 Swedish, 387, 422 tests for purity of, 495 wood, 421 n.Undecane, 99 Undecene(l), 208 Undecene(2), 208 -2-methyl, 127 Unsaturation, effect on physical constants, 545, 548 Vanillin, synthesis by ozone, 142 Vaporization, latent heat of, 567 Vaseline, composition, 17 Verbenene, 435, 436 Verbenol, 429 Verbenone, 429 435 Vestrylamine, 385 Vinyl acrylic acid, polymerization, 236 Vinyl bromide, polymerization, 211 reaction with HBr, 69 Vinyl chloride, reaction with ammonia, Vinyl halides, polymerization, 122 Viscosity, 576 effect of pressure on, 580 effect of ring closing on, 578 relation to lubrication, 577 rubber solutions, 584 Vulcanization of rubber, 136, 584 Wagner rearrangement, 528, 532-534 Walden inversion in case of bornylene, 504 carvoxime, 375 Water gas, composition, 40 conversion to methane, 81, 82 Wax, amorphous, 18, 22, 96 bee's, 19 candelilla, 19 ceresine, 22 montan, 23 paraffine, 21, 22, 23, 34, 95 Wood turpentine, 421 Wurtz, synthesis of hydrocarbons, 50 Xylene, para, conversion to dimethyl-cyclo- heptatriene, 514 para, reaction with diazoacetic ester, 514 Zinc alkyls, 46 Zinc chloride, effect on pyrolysis, 45 Zingiberene, 311, 312 Zingiberol, 311, 312 UNIVERSITY OF CALIFORNIA LIBRARY BERKELEY Return to desk from which borrowed. This book is DUE on the last date stamped below. LD 21-100m-ll,'49(B7146sl6)476