HOROWITZ Study of the Action of Ammonia on Thymol QD 341 P5 H25 A Study of the Action of Ammonia on Thymol DISSERTATION SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIRE- MENTS FOR. THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE FACULTY OF PURE SCIENCE OF COLUMBIA UNIVERSITY BY BENJAMIN HOROWITZ, CHEM., M. A. NEW YORK CITY 1913 NEW YORK SCHOEN PRINTING COMPANY 1913 ) A Study of the Action of Ammonia on Thymol DISSERTATION SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIRE- MENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE FACULTY OF PURE SCIENCE OF COLUMBIA UNIVERSITY BY BENJAMIN HOROWITZ, CHEM., M. A. NEW YORK CITY 1913 NEW YORK SCHOEN PRINTING COMPANY 191} To My Mother and Father ACKNOWLEDGMENT. The author wishes to take this opportunity of expressing his indebtedness to Professor William J. Gies, who suggested the in- vestigation, and under whose direction it was carried out. Prof. Gies' encouragement, constant good counsel, and his many acts of kindness, are gratefully appreciated by the author. B. H. Laboratory of Biological Chemistry, College of Physicians and Surgeons, Columbia University, May, 1913. TABLE OF CONTENTS. PAGE Dedicatory 2 Acknowledgment 3 Introductory 5 History of Thymol 7 The Action of Ammonia on Phenols Other Than Thymol 23 The Action of Ammonia on Thymol 28 Is the Blue Color due to an Impurity ? 30 Certain Quantitative Relationships between Thymol, Ammonia and Alcohol, and Thymol and Ammonia Alone 33 Accelerating, Indifferent, and Retarding Agents in the Formation of the Blue Color 38 Are Ammonia and Thymol Alone the Active Factors in the Formation of the Blue Color? 40 Effect of Several Reagents upon the Blue Solution 46 Tests for the Presence of Nitrogen in the Pigment 47 Isolation of the Blue Product 51 The Chemical Constitution of the Pigment 55 The Action of Sodium Hydroxide, Potassium Hydroxide, and Barium Hydroxide on Thymol 57 Addendum 62 Summary of General Conclusions 63 Bibliography 65 Biographical 67 Publication .. ..68 INTRODUCTION. In the course of experiments on enzymes as possible factors in the development of edema Prof. Gies, greatly to his surprise, found that trypsin in ammonium hydroxide solution containing sodium chloride, failed to give the swelling results with elastin which had previously been observed under similar conditions; instead of swelling, the elastin particles gradually became green and then blue (1). It was then recalled that the elastin used had been prepared some years before, and had been preserved with a solution of thymol in alcohol. That the blue color was due to thymol was confirmed by mixing some thymol with ten per cent, ammonium hydroxide solution and obtaining the color on stand- ing. Alcohol appeared to accelerate the transformation. By evaporating the ether extract (which was red), a purplish red oily product, soluble in ether, toluene, and alcohol, was obtained. On another occasion the oily product became crystalline. One of the most interesting observations was that "the red alcoholic solution was turned deeply bluish by a drop of N/10 sodium hydroxide solution, the red being restored by a drop of N/10 hydrochloric acid solution." This color reaction would clearly indicate its use as an indicator. These observations naturally suggested further inquiry. What was the nature of the blue pigment obtained by the action of ammonia upon thymol? How would it behave when substi- tuted for thymol in the important uses to which the latter is put? How would other phenols behave when treated with am- monia, and how would the thymol pigment, and the other phenol derivatives compare chemically and pharmacologically? It was with this broad object in view that the present inves- tigation was undertaken. Here the more purely chemical side of the problem has been dealt with, for within the given time little more could be done; but pharmacological studies have already been begun, and investigation of various other aspects of the subjects are under way. HISTORY OF THYMOL Thymol, a derivative of cymene, and hence belonging to the important class of organic compounds known as the terpenes and camphors, has engaged the attention of the chemist for many years. Its extensive use in medicine the latest application in hookworm disease (2) being worthy of special mention has increased the necessity for further exact study. L,enher's use of it in a colorimetric deter- mination of titanium (3) tends to show that it is also of value to the analytical chemist. EARLY HISTORY Prior to 1853, thymol was known under such names as "thymolic acid," "thyme oil," "stearoptene," "thymol hydrate," and "thymene oil camphor." Just who discov- ered it, or who first extracted it from one of the several vola- tile oils in which it occurs, is not known. There is good reason to believe that the Indians were familiar with it long before it was known in Europe(4, 5). In 1719 Kaspar Neumann, court apothecary in Berlin, observed that crys- tals were deposited from the oil of thyme (6). Brown (6, 7) stated that this substance had long been known in England under the name of Sal volatile thymi, but that it was quite different from camphor. Not till 1846 was anything definite known with regard to it. In that year Arppe(8) published an excellent paper on the volatile oil of Monarda punctata ("Horse mint"), wherein he showed that the oil consisted essentially of two substances, an "elaeoptene" and a "stea- roptene." The latter, which we now know to be thymol, was further purified by steam distillation, and an analysis yielded the formula C 10 H 14 O the one accepted to-day. Arppe found its melting point to be 48, and noticed that when heated beyond this point, the "stearoptene" remained in a fluid condition even after room temperature had been reached. The insertion of a crystal of the substance caused the solidification of the whole mass. Doveri(9), in an investigation of the oil of thyme, found that upon standing it deposited a "stearoptene." 1 The oil, which at first was reddish brown in color, could be converted to a light yellow by repeated distillation. The portion, which boiled at a temperature between 230-235, gave results, upon analysis, that indicated a composition corre- sponding to the formula, C 10 H 15 O. Relationship to Phenol Historically, the most important contribution to our knowl- ,edge of thymol is undoubtedly Lallemand's who, in a brilliant series of researches (10) extending between 1853-1857, clearly showed its relationship to phenol. He first found that the oil of thyme consists of a substance containing oxygen, C 10 H U O (which he called "thymol"), and of a hydrocarbon, C 10 H 16 . The thymol could be readily extracted with alkalis, and reprecipitated by the addition of hydrochloric acid. Its melting point was recorded as 44, and the boiling point 230. Sulphuric acid was found to dissolve it quite readily, and the resulting substance, obtained by the evaporation of the solution, was soluble in water. In his next paper Lallemand deals with the action of oxidizing agents. By mixing thymol with H 2 SO 4 and MnO 2 and distilling, orange-yellow, oily drops were obtained which soon solidified. This substance, known to us as thymo- quinone, and which Lallemand called "thymol," was found to be analogous to quinine. This "thymol" could be re- 1 Many oils which can be distilled either alone or with steam, without undergoing decomposition the so-called volatile or essential oils deposit a solid (stearoptene) on cooling, leaving a liquid portion (elaeoptene) . duced with sulphur dioxide, giving "thymoilol" (thymo- hydroquinone), analogous to hydroquinone. To complete the analogy, "thymoil" and "thymoilol" combine similarly to quinone and hydroquinone. In his further contributions Lallemand brought forward still more evidence to prove his contention that thymol was a homologue of phenol. With that object in view, he pre- pared the di- and tri-nitro derivatives, and also studied the action of chlorine. Identical Products Obtained from Various Volatile Oils Haines(n) investigated the oil of Ptychotis ajowan, ob- tained in Rajpootana, and which the native doctors used. The " stearoptene " obtained from it was found to be iden- tical with the crystals sold in the bazaar under the name of Ajwa Ka Phul (flowers of Ajwa). " I have not," he writes, "been able to discover by what method the natives contrive to obtain the stearoptene from the oil: it is probably so loaded with it as to crystallize out on dropping in a frag- ment ready formed, without it being necessary to redistil the oil." Of great importance is his finding that his com- pound is identical with the one obtained by Lallemand from the oil of thyme, though he adds, curiously enough, "I could not observe, however, the property which Lallemand as- signs to thymol, of combining with caustic alkalis." Stenhouse(i2) came to the same conclusion as Haines, and found in addition that his stearoptene from Ptychotis ajowan agreed in every way with Arppe's from monarda oil and Doveri's from the oil of thyme. Most striking was Kolbe and Lauterman's work (13) in showing how close was the relationship existing between phenol and thymol. These workers, whose researches upon the action of phenols in the presence of sodium and carbon dioxide have become classical, showed that thymol could be converted to thymotic acid in the same way as phenol could be transformed into salicylic acid. Effect of Kekule's Benzene Theory Kekule"'s benzene theory, advanced in 1865, had no small influence upon the views taken with regard to the consti- tution of thymol (7), as indeed it had with regard to the con- stitution of phenols in general. Now for the first time a sharp differentiation came to be made between the aromatic alcohols which, on oxidation, yielded the corresponding alde- hydes and acids, and their isomers, the phenols, which did not yield acids containing the same number of carbon atoms. But this did not differentiate the phenols from the tertiary alcohols; and so, as characteristic of the former, it was stated that they dissolved in caustic alkalis, the hydrogen of the hydroxyl being replaced by sodium a reaction which, in the case of the alcohols, could be accomplished only by the ac- tion of the alkali metals. Relationship to Cymene and to Carvacrol Considerable progress on the constitution of thymol was made by the researches of Engelhard t and Latschinoff(i4). By fusing cymene sulphonic acid with potassium hydroxide they obtained a phenol which, though not thymol, was found to be isomeric with it. Upon examination they found it to be identical with a phenol similarly isolated by Muller(i5), and which he regarded as thymol, and another by Pott(i6), who first obtained his cymene from camphor by the action of phosphorus pentasulphide, and then the phenol by fusing its sulphonic acid with potassium hydroxide. We now know this substance to be carvacrol, isomeric with thymol. The synthetic process having failed, Engelhardt and Latschi- noff substituted analytical methods. Recalling one of Lalle- mand's experiments, wherein he obtained a chlorine deriva- tive of thymol, C 10 H 8 C1 5 O, which, on warming to 200, broke down into propylene and tetrachlorcresol, they treated thymol with phosphorus pentoxide, and obtained propylene and cresol. Just which of the three cresols this was could not be ascertained with exactness, as little was known with II regard to the position of the methyl and hydroxyl groups. Only at a later period was it shown that the cresol under question was identical with the meta variety. Carstanjen(i7) brilliantly followed up the work of Engel- hardt and Latschinoff. He first tried the action of reducing agents upon thymol, such as zinc dust and hydriodic acid, but the results were unsatisfactory. He next prepared the chloride C 10 H 13 C1 by heating with phosphorus pentachloride, but the product obtained was very impure, and distillation failed to purify it, for at 120, the substance blackened; and even under 7 mm. pressure the conditions were no better. However, the product was treated with sodium amalgam in a weak acid solution, and the resulting hydrocar- bon was purified by repeated distillation over metallic sodium. Analysis gave the formula C 10 H 14 . But Carstanjen's further work upon cymene (for that is what his hydrocarbon proved to be) showed perhaps still greater originality. He heated cymene with a mixture of potassium dichromate and sulphuric acid and obtained a white powder which dissolved in ammonia, and was re- precipitated by hydrochloric acid. This proved to be tere- phthalic acid, C 6 H 4 (COOH) 2 , where the carboxyl groups are para to one another. From this, then, it followed that the methyl and propyl groups in cymene (and therefore in thymol, which had been reduced to cymene) must also be para to one another: CH 3 COOH I ' I H C : H H C \C H p H -* I ii H-C. .C-H H-C. ,C-H X(V^ ^C' 'jjf COOH Cymene Terephthalic acid A further important conclusion by Carstanjen was that thymol could form but two isomeric phenols, which at once becomes evident from its relationship to cymene: C 3 H 7 formula i representing thymol and 2 carvacrol. As further confirmation, the latter with phosphorus pentoxide yields o-cresol and (propylene) in contradistinction to m-cresol from thymol. The question as to whether the propyl group was normal or iso was not settled until some years later. In the meantime, Fittica(iS) modified Carstanjen's method of ob- taining cymene from thymol by acting directly upon the lat- ter with phosphorus pentasulphide, and thereby obtaining the hydrocarbon. From hasty observations, based on in- conclusive observations by previous workers, Fittica arrived at the conclusion that the C 3 H 7 in cymene was normal and not iso; hence it followed that the C 3 H 7 in thymol was also normal. This erroneous view prevailed till 1891, when Widman solved the problem. Synthesis The first synthesis of thymol was accomplished by Wid- man (19). This research is a model of its kind. In a former contribution (20) the author showed how w-toluidine could be obtained from benzaldehyde by first nitrating the latter, acting upon the nitro compound thus formed with phos- phorus pentachloride, obtaining the w-nitrobenzaldehyde, and 13 reducing this with zinc and hydrochloric acid to the tolui- dine: /CHO /CH C.H/ C,H/ \H \N0 2 CHO /CHC1, /CH 3 / / By a similar series of reactions thymol was built up. Wid- man's starting point was cuminol, or ^-propylbenzaldehyde: 1 /CHO C - H C.H 3 ^OH + N 2 + H 2 C 3 H 7 \C 3 H 7 Thymol When NO 2 is introduced into the ring containing CHO, the NO 3 enters m- to CHO : 1 The C 3 H 7 in this compound is iso, though this was not known at the time. The other two compounds can therefore be represented thus: NH 2 The fact that thymol was obtained from m-nitrocuminoi is in itself good evidence for the belief that the hydroxyl group in thymol is ra- with respect to the CH 3 . This view receives abundant support in that w-cresol can be obtained from thymol. As the reactions were never allowed to go above 100, there is every reason for believing that the propyl group in thymol stands in the same relation to the CH 3 - group as it does in cuminol; and for this again we have evidence, since thymol can be reduced to cymene. The Nature of the Propyl Group The position of the propyl group, whether normal or iso, still remained unsettled. This problem Widman set him- self to solve, and in 1891 published his results in a paper (21) which was a worthy successor to his former one. The aim in view was to investigate the propyl group in cymene; 15 once this could be established, the nature of the C 3 H 7 in thy- mol would follow. In an excellent review of the work hitherto done in this direction, Widman recalls the labors of such masters as Dumas, Gerhardt, Fittig, Konig, Beilstein, Jacob- sen, Robert Meyer, and Kekule, and finds a total disagree- ment among them. The question at issue was whether the cymene was />-methylpropylbenzene or />-methylisopropylbenzene. Wid- man therefore decided to prepare synthetically both those compounds, and to compare the products so obtained with cymene. He prepared the normal variety by the action of />-bromtoluol on propyl bromide in the presence of sodium (Fittig's synthesis) : CH S CH 2 .CH 2 .CH 3 . The iso modification was obtained by first preparing isopropylbenzene by the action of isopropylbromide and ben- zene in the presence of aluminium chloride (Friedel and Craft's synthesis), CH 3 .CH.CH, CH 3 CH, i6 then brominating the product, and finally treating it with methyl bromide in the presence of sodium: Br CH X\ CH 3 CH S "Na" Na"! Br Br j CH, CH, By comparing these two products with cymene Widman came to the conclusion that the latter was methyliyopropyl- benzene. A tabulation of his results .as he gives them is appended : i 7 WWa 2 W A ? & , ? > a s R w >,$ p oP . >' LW H- p fi . IS Several Syntheses After Widman's observations quite a number of chemists succeeded in synthetically preparing thymol in several in- teresting ways, one or two of these having since been made the basis for commerical production. A few of these methods are here discussed. Beckmann and Eickelberg(22) obtained thymol from menthone by brominating the latter, and heating the result- ing dibrommenthone with quinoline : CH, / CH \ CH 2 CH 2 CH, CO \CH/ 4 Br Meathone x\ CH 2 CH 2 CH 2 CO CBr C,H r Dibrotnmenthoae 2HBr CH, CBr CH 2 CH 2 Quinoline r c \ CH 3 CH 2 CO ( I II CH C.OH u Ketone form of thymol C 3 H 7 Thymol The authors presume that the ketone form of thymol is an intermediate production, though they failed to isolate it. Dinesmann(23) has taken out a patent for his method of preparing thymol one that recalls Engelhardt and Latsch- inoff's unsuccessful attempts. Having in mind the fact that cymene sulfonic acid when fused with KOH gives 19 carvacrol and not thymol, owing to the sulphonic acid rad- icle being o- with respect to the CH 3 group, Dinesmann started with 2-brom-/>-cymol, C 3 H 7 and by dissolving it in fuming H 2 SO 4 (containing 15-20% SO 3 ) he obtained the 2-brom-5-(or 3)-sulphonic acid: CH, (J S0 3 Hl J C 3 H 7 He next proceeded to eliminate the Br by heating the com- pound in an autoclave at 170 with zinc dust and ammonia, and thereby obtained 3-cymolsulphonic acid (zinc salt), which on fusion with KOH gave thymol. Wallach(24) obtained thymol (in very small quantities he states) by brominating menthenone (in glacial acetic acid) and warming the resulting dibrommenthenone : 20 CH, CH, CH, CH CO CHBr CO CH ^ CBr I CH C H 3 T Menthenone Dibrommenthenone Ketoae form of thymo I \y^ \C-OH /-A / c / Thymol a reaction that shows certain similarities to Beckmann and Eickelberg's. Semmler and McKenzie(25) by heating buccocamphor with HC1 for two hours at 150-180 obtained a quantitative yield of thymol: CH.CH< ' CH.CH/ ' /\ \CH 3 /\ \CH, CH 3 CO CH 2 CO II or | | H 8 CH 2 C.OH CH 2 CO -^ ? CH.CH, Buccocamphor Ketone form As an interesting example of intramolecular change Semm- ler's production of thymol from umbellulon(26) is worthy of note: 21 CH, CH H C CH 3 CH L \*,L~Lm Umbelluloa A CH COH II CH Thymol 23 THE ACTION OF AMMONIA ON PHENOLS OTHER THAN THYMOL.* As far back as 1835 Robiquet (27) published a paper on the action of ammonia upon orcinol, in which he showed that am- monia gas in the presence of dry air produced no color with orcinol, part of the ammonia simply being absorbed ; but as soon as moisture was allowed to enter the mixture, a blue color began to develop. The resulting product, which could be isolated by extracting the blue substance with ether and evaporating the lat- ter, could not be obtained in a crystalline form, and this fact prevented Robiquet from pursuing investigations into the nature of the substance. However, he satisfied himself that it was not simply an ammonium salt, since it failed to yield ammonia by heating with an alkali. He also showed that other alkalis could not be substituted for ammonia in this reaction. Potassium hydroxide, for example, produced a brown color one which could be obtained by exposing moist orcinol to the air. Upon ex- traction with ether and evaporation of the solvent, orcinol was again obtained. By a modification of Robiquet's process for obtaining orcein, De Luynes (28) succeeded in isolating the coloring matter of litmus. Starting with orcinol, which could be extracted from lichens, De Luynes mixed this phenol with given quantities of sodium carbonate, water and ammonia, and heated the mixture for 4-5 days at 60-80. On diluting and acidifying with hydro- chloric acid the coloring matter was precipitated. Bearing in mind that a certain analogy existed between resor- cinol and orcinol, Malm (29) tried the action of ammonia upon the former. By allowing a mixture of resorcinol, ammonia, and sodium hydroxide to stand in a warm place for several days, and then acidifying, a reddish-brown precipitate was obtained, which, upon collecting and drying, was found to possess a metallic lustre. The behaviour of this substance with acids and alkalis was found to be similar to that of litmus. Lex (30) found that in the presence of certain oxidizing * Ammonium salts are not considered here. For a general discussion of these see Hantsch: Ber. d. D. Chem. Gesell., 40, 3798 (1907). 2 4 agents, such as bleaching powder, or bromine water, and even when merely exposed to the air for a sufficient length of time, a solution of phenol in ammonia would turn blue. With the object of employing Lex's reaction as a test for carbolic acid, Salkowski (31) made a more careful study of it. He found that by adding ammonia (one part) to phenol (four parts), and then adding a few drops of bleaching powder solu- tion (obtained 1>y dissolving one part of bleaching powder in twenty parts of water, filtering, and using the filtrate), and warming, a blue color would instantly make its appearance if much phenol were present. If the amounts were small the color would develop in a few minutes or in a quarter of an hour, de- pending upon the quantity of phenol. Very dilute solutions gave a green color. Too strong heating, or the presence of an exces- sive quantity of bleaching powder, prevented the reaction. Salkowski maintained that the test was a far more delicate one than with ferric chloride, for whereas a concentration of 1 : 2,000 failed to give a reaction with the latter, 1 : 4,000 gave a strong blue with ammonia. In 1873 Phipson (32) isolated a product obtained by the action of ammonia on phenol, which he called "phenolcyanine." He prepared it by dissolving phenol in alcohol, adding ammonia, and allowing the mixture to remain for some weeks in a partially closed flask. In about fifteen days, when the liquid had become dark green in color, twice its volume of water and y^ vol. am- monia were added, and the mixture again allowed to stand for about six weeks. By this time the liquid had taken a very tine blue tint, was very dark, and a certain quantity of phenolcyanine was found at the bottom of the vessel adhering strongly to the glass. That which remained in solution could be precipitated by saturating the liquid with salt. The product was collected, dis- solved in hot alcohol or benzol, and recovered by evaporating the solvent. Its properties were described as follows : A dark blue sub- stance, soluble in alcohol, yielding a fine blue solution, in ether reddish-blue solution ; in benzol reddish-purple solution. Cone, sulphuric acid gave a bluish-green coloration ; hydrochloric acid had little action; nitric acid produced a nitro compound very 25 different from picric acid. Slightly soluble in water. Deep sky blue by day, red by night. Acids reddened the solution, alkalis bringing back the blue. Nascent hydrogen destroyed the blue color, but upon adding ammonia and exposing to the air the color was reformed. As but a small amount of substance was available the analysis did not prove very successful. In a succeeding paper (33) he attempted to show the rela- tionship between phenylcyanine and indigo, but his method of procedure would have thrown discredit even upon an embryonic chemist. Since, he argues, phenolcyanine has twelve carbon atoms, whereas indigo has sixteen, the problem is to introduce four more carbon atoms into the former to convert it into the latter. Whereupon he records with the greatest gravity how he proceeded, first, to melt phenolcyanine at a moderate tempera- ture with sodium acetate, then to dissolve the product in con- centrated sulphuric acid, and finally throw down the sulpho acid by adding an excess of water. A similar experiment was car- ried out with phenolcyanine and nitro-naphthalene, using equal equivalents of each ; for the former contained 12 carbon atoms, and the latter twenty; and 12 -f- 20 -^ 2 16, which gives us the number of carbon atoms in indigo ( !). "These sulpho- acids," he writes, "mixed and saturated with ammonia or am- monium carbonate, gave a small quantity of a purple black product, insoluble in water and alcohol, but soluble in concen- trated sulphuric acid, producing a dark, emerald-green solution. This product is very similar, if not identical, to the black indigo produced when the leaves are badly fermented." And this is the last we hear of Phipson ! A notable contribution to the orcein question is that by Liebermann (34). In the course of a study of the action of ammonia upon orcinol, it occurred to him that possibly the com- bined action of the ammonia and the oxygen of the air was equivalent to nitrous acid. Accordingly he dissolved orcinol in concentrated sulphuric acid, and added potassium nitrite, where- upon a deep purple coloration was obtained. The addition of water threw down a red flocculent precipitate, which was found 26 to be soluble in alkalis, giving a beautiful red solution. But the substance was not identical with orcein. Further work upon phenols assured Liebermann that phenols in general responded to color tests with nitrous acid. Since his day the Liebermann reagent (6 per cent, potassium nitrite in cone, sulphuric acid) has been extensively used. In a subsequent paper Liebermann (35) describes how he repeated De Luynes's work and found that the action of am- monia upon orcinol gave him a product part of which was read- ily soluble in ammonia. The more insoluble part was dissolved in sodium hydroxide. Both parts were now treated alike ; namely, acid was added to each to precipitate the substance, which was then washed, dissolved in alcohol, and the latter then evaporated. Both substances were found to be amorphous, and outwardly could not be differentiated. The purple color obtained with alkalis showed a decided reddish tint with the first, and a more bluish one with the second. Wurster's paper on "The formation of color by means of hydrogen peroxide" (36) is particularly worthy of close study. He found that in the presence of hydrogen peroxide, ammonia and phenol gave a blue coloration, which gradually changed to green and then to yellow. The solution became quite decolorized when an excessive amount of hydrogen peroxide was added. The addition of acetone, alcohol, or oxalic acid considerably hastened the reaction. Hydroxylamine was found to be still more effective. This substance, together with phenol and hydro- gen peroxide, formed nitroso-phenol, but neither a blue nor a green color was evident. With the addition of ammonia the color rapidly formed. As but a small quantity of hydroxylamine was necessary, Wurster would explain this reaction by saying that the hydroxylamine is oxidized to nitric oxide, which in the solu- tion acts the part of an oxygen carrier. To isolate the product Wurster gives these directions : To an emulsion of phenol in water ammonia is added in such amount a? to leave some of the phenol still undissolved. Some sodium hydroxide and an equal volume of hydrogen peroxide are now added, the whole diluted with ten times its volume of water, and well shaken. The addition of a small crystal of a salt of hydrox- 2 7 ylamine causes a light blue color to appear within a few minutes. This soon changes to deep blue, and in one or two days becomes quite green. Without hydroxylamine the color develops quite slowly, and the maximum intensity is not reached before twenty- four hours, whereas with it a full development of color is notice- able within a quarter of an hour. By extracting the blue color with ether part goes into solu- tion, giving a red coloration. A solution of amyl alcohol and ether extracts more completely. As the ether extract of the acid solution of the dye* is red, and the Liebermann dye (nitrous acid on phenol) under similar conditions is yellow, Wurster is inclined to believe that the two are quite different; and this view receives support in that their spectroscopic behavior is not the same. Dealing next with the constitution of the blue product Wurster states that the whole behavior of the compound points to its identity with phenolquinoneimid, a substance first prepared by Hirsch (37) by the action of quinonechlorimid on phenol, though he could not isolate it. Wurster prepared phenolquinoneimid by adding ammonia to a watery solution of quinone in the presence of an excess of phenol. The yellow quinone solution quickly becomes green on the addition of ammonia, and blue by stirring in contact with air. A still easier method is to start with /-amido phenol. By dis- solving this in sodium hydroxide a red solution is obtained which becomes yellow on contact with air. By the addition of phenol, phenolquinoneimid is at once obtained. By studying other phenols Wurster found that all phenols that have the para position with respect to the hydroxyl group unsubstituted form quinoneimids. Where the para position was substituted the substances oxidized to a yellow product or were not attacked at all. The author concludes significantly by saying that since it has been shown that both ammonia and phenol are decomposi- tion products of proteins, the formation of color in plants bears a certain relationship to the hydrogen peroxide that is present. Zulkowski and Peters (38) carefully repeated Liebermann's * Here the word is rather loosely employed. 28 work on the action of ammonia upon orcinol. By allowing a mixture of 50 gms. orcin in 200 c.c. water and 200 c.c. ammonium hyclroxid solution [strength not given] to stand for two months a thick jelly was obtained, from which the following three colored substances were isolated : 1) A red-colored product, orcein, obtained in microscopi- cal crystals from a mixture of water and alcohol. With alcohol it gives a carmine-red solution, and with ammonia, the fixed alkalis, and the alkaline carbonates, a bluish-violet. Insoluble in water. Yield, 50%. 2) A yellow crystalline substance, soluble in ether and al- cohol, and less so in boiling water, in each case giving a yellow solution. 3) An amorphous, litmus-like substance, with a greenish metallic-luster. Insoluble in alcohol, soluble in alkalis to a dark blue solution, which turns red on the addition of acids. By the addition of hydrogen peroxide, the rapidity of the process could be greatly increased. What required days before could be accomplished in so many hours. Under otherwise identical conditions resorcinol did not yield an orcei'n-like substance ; but the latter could be obtained by allowing orcinol (142 parts), resorcinol (110 parts), 22,% ammonium hydroxide solution (7.7 p.), and 3% iI 2 O 2 (3 MM) p.) to stand for several days. By recrystallizing the product from alcohol a bronze lustrous substance was obtained, the red acetic acid solution becoming blue upon the addition of ammonia or alkali. Maseau (39) in a study of the comparative color reactions of the phenols with ammonia and iodine in the presence of alco- hol, describes the colors with ammonia as follows : Pyrocate- chin, reddish-brown ; hydroquinone, yellow ; pyrogallol, black- ish-brown ; orcin, red to violet. Phenol, resorcinol and naphthol remained colorless. It is quite apparent that the author did not take the time element into consideration. THE ACTION OF AMMONIA ON THYMOL. The brilliant Lallemand, in his exhaustive study of thymol, refers to the action of ammonia upon it in no uncertain terms. 2 9 "Le thymol," he \vrites (10), "n'est pas altere par 1'ammoniaque liquide ; mais il dissout une grande quantite de gaz ammoniac qu'il abandonne lentement en se solidifant." This view became the accepted one. We find Watt, for example, quoting it in his dictionary (40). Just as in many another reaction a neglect of the time element failed to give any visible result. The first one to record any reaction is Lex (30). In de- scribing some new color tests for phenol, the author records that the addition of ammonia to phenol causes a blue color to form under either one of the following conditions : a) Warm- ing the solution with bleaching powder; b) warming with ba- rium peroxide; or c) allowing to stand exposed to the air. He adds, in a casual note, that he finds thymol to behave simi- larly.* In a comparative study of the behavior of thymol and phenol with various reagents, Hirschsohn (41) found that when thymol (1:1000) was heated with bleaching powder and am- monia the solution became cloudy, and after a time showed a greenish tinge. In a concentration of 1 : 2000 and I : 4000 only a cloudiness was obtained. The substitution of chlorine water for bleaching powder produced a bluish-green coloration. Wurster, in the article already quoted at some length (36) showed that thymol, in the presence of hydrogen peroxide and ammonia, forms a quinoneimid analogous to that obtained from phenol. The acid properties of the resulting imid a red oil insoluble in water were found to be so slight that dilute am- monia had no tendency to salt formation. The blue sodium salt could be decomposed with a large quantity of water, but the potassium salt was more stable. * The casual reference to thymol comes in the last three lines of the article. The quotation in full is as follows: "Uebrigens zeigt das einzige fernere Glied der Ph-enolreihe, welches ich zu priifen Gelegenheit hatte. das thymol, insofern ein ganz analoges Verhalten." That this observa- tion attracted no attention is seen by the fact that no mention of it wliat- ever w to be found in She literature. It was only by the merest chance that we came across this article, and then long after the experimental part of this research had been begun. However, even had we stumbled across it at the very beginning of this inquiry Lex's comment would merely have given added impetus to our desire to inaugurate this investigation. 30 From the title of his essay "The role of hydrogen perox- ide in the formation of color" we not only get the object of this research, but also the author's opinion that the oxidizing agent plays an indispensable part in the formation of the color. This is emphasized more than once in the contents. In a recent communication, Gies (42), in following up his previous observations (1), shows that filter paper soaked in the blue, alkaline, alcoholic solution (obtained from the blue of an ammonium hydroxide-thymol mixture by extracting the color with ether, evaporating the latter, dissolving in alcohol, and rendering slightly alkaline), and then dried at room tempera- ture, "assumes a bright red color as the alcohol disappears. Treated with alcohol, such red filter paper, particularly if slightly moist, becomes bright green." Interesting probabilities suggested by these results, and the possible relationship of these color phenomena to the pigments in the Monardas* and other plants, will be investigated. IS THE BLUE COLOR DUE TO AN IMPURITY? The first question that suggested itself was whether the blue color was due to some impurity that was present in the thymol? If so, the probabilities were that the same amounts of different varieties of thymol would show marked differences in intensity of color. Three varieties on the market, Merck's, Eimer and Amend's, and Kahlbaum's, were procured, but no differences could be detected (Table I). In order to eliminate further doubt, it was decided to purify the thymol. After several preliminary experiments, a conven- * Wakeman : Bulletin of the Univ. of Wisconsin, No. 448 ; Science series, 1911, IV, p. 25. Colored pigments in the corollas of Monardas didyma, fistulosa, and punctata, are described. These are regarded as probable oxidation products of the thymol and carvacrol. Hydrothymo- quinone, thymoquinone, and dihydroxythymoquinone, have been isolated. These are known to form colored compounds by combining with monatomic phenols. lent means of doing this was found to be to steam distil the substance, and recrystallize the product from glacial acetic acid.* As tests of purity, the melting points of the different sam- ples were compared. However, even here one is beset with diffi- culties, for in the literature melting points ranging anywhere from 44 to 53 are to be found.f With a view to explaining some of these differences, slight modifications in procedure were adopted at different times (Table II). The last determination (number 8), giving the melting point from 50-50.5 should be taken as the most reliable. From a glance at the table it will be seen that the melting points of the different samples agree remarkably well. This shows that the samples obtained from Merck, Eimer Amend, and Kahlbaum, were equally pure, and that steam distillation * In some of these steam distillations the thymol came over in the form of a colorless oil; by merely transferring into another flask the oil would solidify into a shining white mass. This solidification by mere agitation is a marked characteristic, and has been noticed by many workers. If the condenser is too cold the thymol very readily solidifies in the tube. A little alcohol or acetic acid will easily dissolve this. As this phenol is exceedingly soluble in glacial acetic acid, and as crystallization will not take place from a too dilute solution, quantities of the steam distilled thymol were added to a small quantity of acetic till an almost saturated solution was obtained (by the aid of gentle heat). Then one or two c.c. more of the acid were added. Thymol crystallizes from glacial acetic in large, colorless, hexa- gonal plates, a mosaic of the crystals appearing on the surface. In the course of these crystallizations an interesting means of study- ing the development and formation of crystals was hit upon. This method consisted in very slowly pouring an almost saturated solution of thymol in glacial acetic acid into a large quantity of water. Soon small droplets appear, and in these latter the nucleus of the crystal, in the s'hape of a speck of white solid, springs into being. This enlarges and spreads until the crystal is formed. f Carnelly, in his "Melting and Boiling Point Tables" (1885), I, 224, gives the following: Liquid (Febve) ; 44 (Lallemand, Sten'house, Widman) ; 46 (Kekule and Fleischer); 48 (Arppe) ; 49.2 (Schiff) ; 51 (Andresen) ; 53 (Haines). Fehling ("Handworterbuch der Chemie," 7, 969 (1905) ), supple- ments this by giving Mentschutkin's (50) and Reinsert' s (49.7) figures. 3 2 and recrystallization did not tend to increase the purity. Above all, the purified products reacted with ammonia in precisely the same way, and to precisely the same extent as the non-purified materials. TABLE I. Comparison of intensity of color for different samp-es of thymol. ABC Thymol 0.5 gm. 0.5 gm. 0.5 gm. Ammonia (10%)* 100 c.c. 100 c.c. IGOc.c. Alcohol (95%)* 10 c.c. 10 c.c. 10 c.c. A Merck product. B Eimer & Amend product. C Kahlbaum product. Conclusion. Intensity about same in all. TABLE II. Determination of the melting point of the ordinary and purified thymol. [The numbers refer to C.] Merck's Merck's steam . p . , v , ,, steam distil, and Merck Eimer & Amend Kahlbaum d ; st j lle j recryst. from acetic 1. 48 47-48 47-48 2. 48 47-48 48 3. 49+ 48 49 4. 49+ (0.2-0.3) 49 49 5. 49 +(0.4-0.5) 49 6. 48 49 7. 48.5 48.5 49 8. 50-50.5 50-50.5 50.5 50.5 50.5 Notes. Roth's Melting Point Apparatus was used [see Ber. d. D. Chem. Gesell. 19, 19TO (1886)]. Melting points were taken only after complete fusion. Note on the method of drying thymol: It was noticed that when concentrated sulphuric acid was used as a drying agent in the desiccator the acid gradually changed to a brownish-red color, and fine violet colored deposits were repeatedly obtained. This result was always obtained in the presence of thymol. Calcium chloride was therefore substituted. * Unless otherwise stated the ammonia and alcohol used throughout are respectively 10% and 95%. Alcohol was used on the supposition that it favored the reaction by increasing the solubility of thymol. See page 44. 33 CERTAIN QUANTITATIVE RELATIONSHIPS BETWEEN THYMOL, AM- MONIA AND ALCOHOL, AND THYMOL AND AMMONIA ALONE. Having shown that the blue color is not due to an impurity in the thymol, it became important now to establish definite quantitative relationships in order to determine to what extent the different reagents took part in the reaction. Since we were here dealing with three factors, thymol, ammonia and alcohol, the most logical method that suggested itself was to conduct experiments on the basis of one variable and two constants. Table III shows that the intensity of the color is proportional to the amount of thymol present, but Tables IV and X show that beyond certain limits this does not hold. In the same way it was shown that the intensity of the color varied directly with the amount of ammonia present (Table V), and subsequently, that more concentrated solutions of ammonia inhibited the for- mation of color. The action of alcohol was most peculiar. It was at first supposed, from the original experiments on elastin (1), that alcohol would considerably accelerate the reaction. But this did not prove to be the case (Table VI) ; and indeed, under certain conditions, it acted as a retarding agent (Table VII). Of course, this at once suggested the idea of dispensing with the alcohol altogether, and most satisfactory results were obtained (Table VIII).* * One of the most interesting phenomena in these observations was the behavior of thymol when added to the ammoniacal solution. As soon as the powdered thymol touched the surface of the liquid it tended to form globules. This was particularly marked when the quantities of thymol added were comparatively large. When a globule had reached a certain size it would sink to the bottom. Upon standing the globule would grad- ually assume a reddish or violet tint, and a pear-shaped form. Two op- posing forces now came into play, the upper liquid portion of the globule which tended to force its way upward, and the lower, which restrained it. The upper portion gradually increased in size, and after it had attained a certain volume it broke away from the rest of the globule and came to the surface, forming colored (usually violet) oily layers. A repetition of the above phenomenon would now commence in the globule at the bottom ; and, indeed, this would continue for several days, till finally the entire globule had disintegrated. This might offer an interesting field for physico-chemical study. 34 Under certain conditions water was found to take an inter- esting part in the formation of color (Table XI). With regard to the delicacy of the reaction, it was found that within five days a distinct coloration was obtained in a concentration of 1:25,000 (Table IX). TABLE III. Effect of varying thymol, with ammonia and alcohol constant. 12345 Thymol 0.1 gm. 0.2 gm. 0.3 gm. 0.4 gm. 0.5 gm. Ammonia (10%) 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. Alcohol (95%) 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. Conclusion. Intensity of color varies directly with the amount of thymol present. Notes. Color noticeable one hour after commencing. Color gradually changes from green to blue, the change taking four, and perhaps more days. TABLE IV. What amount of thymol gives the maximum color? 12345678 Thymol ..0.02gm. 0.05 gm. 0.5 gm. 0.6 gm. 0.7 gm. 0.8 gm. 0.9 gm. 1.00 gm. Ammonia. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. Alcohol... 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. Conclusion. Three gives maximum color. Notes. One showed first trace of color in \ l / 2 hours. After 46 hours the color was still a very light green. Intensity of colors in 4 and 5 about same as in 3, but pinkish globules on top. In 6 color somewhat less intense than 5 ; 7 and 8 showed faintest trace of color after 1^ hours, but solution was very cloudy. Pink globules on top. After 46 hours' standing the in- tensity of the blue color showed but very slight sign of increase. TABLE V. Effect of varying ammonia, with thymol and alcohol constant. 12 34 5678 9 Thymol 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 35 TABLE V Continued. Ammonia lOc.c. 20c.c. 30c.c. 40c.c. 50 c.c. 60 c.c. 70 c.c. SOc.c. 90c.c. Water 90 c.c. 80 c.c. 70 c.c. 60 c.c. SOc.c. 40 c.c. 30 c.c. 20 c.c. 10 c.c. Alcohol 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c 10 c.c. Conclusion. Intensity of color varies directly with the amount of ammonia present. Notes. After 26 hours, large crystals partly pinkish, partly colorless separated out in 1. After 41 hours' standing the solution in 2 changed to pink. After 91 hours' 1 and 3 were slightly pink. 1 and 2 were repeated, and beyond getting smaller crystals in 1, identical results were obtained. TABLE VI. Effect of varying alcohol, with thymol and ammonia constant. 123456789 10 Thymol 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gtn. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm . Ammonia 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. Alcohol 1 c.c. 2 c.c. 3 c.c. 4 c.c. 5 c.c. 6 c.c. 7 c.c. 8 c.c. 9 c.c. 10 c.c. Water 10 c.c. 9 c.c. 8 c.c. 7 c.c. 6 c.c. 5 c.c. 4 c.c. 3 c.c. 2 c.c. 1 c.c. Conclusion. Intensity of color about the same in all. Notes. No difference in intensity could be noticed even after 22 hours. TABLE VII. Effect of variation of alcohol upon a comparatively large quantity of thymol in the presence of a constant quantity of ammonia. 123456 Thymol 1 gm. 1 gm. 1 gm. 1 gm. 1 gm. 1 gm. 1 gm. Ammonia (10%) 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. Alcohol (95%). 10 c.c. 20 c.c. 30 c.c. 40 c.c. SOc.c. 60 c.c. 70 c.c. Water 70 c.c. 60 c.c. SOc.c. 40 c.c. 30 c.c. 20 c.c. 10 c.c. Conclusion. Intensity of color decreases with increase of alcohol. 36 Notes. The difference in results between this experiment and the one above is probably due to two causes: 1) Greater quantity of thymol used; 2) greater differences in amount of alcohol. Purplish drops in 1, and to a less degree in 2 and 3; 4 hardly any; 5, 6, 7 none at all. This is due to the fact that the more alcohol present, the more perfect the solution. It has been noticed that whenever there is more thymol than will dissolve, the excess tends to change to a pinkish color sometimes appearing in the form of a pinkish precipitate, more often as pinkish or purplish globules. TABLE VIII. Effect of varying ammonia in the presence of a constant quantity of thymol and in the absence of alcohol. 12 345 678 9 Thymol 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. Ammonia (10%) 10 c.c. 20c.c. 30 c.c. 40c.c. 50 c.c. 60 c.c. 70c.c. SOc.c. 90c.c. Water- 90 c.c. SOc.c. 70 c.c. 60 c.c. SOc.c. 40 c.c. 30 c.c. 20 c.c. 10c.c. Conclusion. Intensity of color varies directly with the amount of ammonia present. Notes. The color is obtained as well without alcohol as with it. 1 began to show a faint trace of color only after 4 hrs. (In the presence of 10 c.c. alcohol the same sample gave evi- dence of color in l l / 2 hrs.) 1, 2, 3, 4, 5 contained crystalline precipitates (probably un- changed thymol), the quantity decreasing from 1-5. 6, 7, 8, 9 did not show this. After 39 hours, solution in 1 became pink ; 2 also, but to a less extent. (1 was repeated and the pink color verified.) 37 TABLE IX. How small a quantity of thymol is necessary to produce the blue color? 1234 Thymol (saturated 10 c.c. 15 c.c. 20 c.c. 25 c.c. solution in water) (=0.008gm.) (=0.012gm.) (=0.016gm.) (=0.02gm.) Ammonia (20%). 100 c.c. 100 c.c. 100 c.c. 100 c.c. Water , 90 c.c. 85 c.c. 80 c,c. 75 c.c. [2 : 50,000] [3 : 50,000] [4 : 50,000] [5 : 50,000] Conclusion. After five days 1 became faint green. Notes. 20% ammonia was used as the added volumes of water and thymol solution reduced the solutions to 10%. TABLE X. Effect of ammonia upon comparatively large quantities of thymol. 1 2 3 Thymol 1 gm. 2 gm. 5 gm. Ammonia 100 c.c. 100 c.c. 100 c.c. Conclusion. Excess of thymol tends to change the blue color of the solution to red, purplish layers appearing at the surface. Notes. Turbidity increases with increase of thymol. Red- dish drops on top increase with increase of thymol. After 1 hour 1 began to show the usual greenish tinge, but neither 2 nor 3 showed any color, though they were both very turbid. In 20 hrs. 1 had become blue, 2 bluish-red and 3 almost wholly red. In all reddish oily drops appeared at the surface more so in 3, less in 1. In 5 days all the samples had become more blue. (It has been noticed that whenever the red is more pronounced at first it has a tendency gradually to change to the blue.) After 2 weeks, the blue was more marked than ever. The solutions were all bluish, the oily drops on top purplish. TABLE XI. Effect of different quantities of water upon a relatively large amount of thymol. 123456 Thymol 1 gm. 1 gm. 1 gm. 1 gm. 1 gm. 1 gm. 1 gm. 38 TABLE XL Continued. Ammonia (10%)... 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. 100 c.c. Alcohol (95%) .... 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10c.c. 10c.c. Water Oc.c. 10 c.c. 20 c.c. 30 c.c. 40 c.c. 50 c.c. 60 c.c. Conclusion. 1 is very turbid and has pinkish tinge; 2, 3, 4, 5, 6, 7, all blue. The 10 c.c. difference in water in 1 and 2 plays an important part. Notes. Above was undertaken because thym. 1 ; am. 100 ; ale. 10; water 0, showed hardly a trace of blue color, but decided turbidity, and pinkish tinge in solution ; whereas thym. 1 ; am. 100 ; ale. 10 ; water 70, showed the usual greenish-blue. ACCELERATIN^, INDIFFERENT AND RETARDING AGENTS IN THE FORMATION OF THE BLUE COLOR. Quantitative relationships having been established, it now be- came advisable to investigate other factors that might influence color formation. It was found that heating on the water bath 5-10 minutes accomplished what standing at room temperature would have taken 24 hrs. or more.* (Table XII.) On the other hand, sur- rounding a mixture of ammonia and thymol with ice prevented the formation of the color altogether. (Table XIII.) Bearing in mind the effect light has upon phenol (43) as well as on quinone and thymoquinone (44) t, it became desirable to ascertain what would happen in the absence of sunlight ; but this did not effect the color formation in the least (Table XIV). The effect of the addition of a small quantity of ether was rather unexpected : not a trace of color could be detected after 20 hours' standing (Table XV). * It may be remarked here that in most of the subsequent work heat was not resorted to. Continued heating would tend to change the blue color into red, but the blue could be restored by adding more ammonia. f At this time the theory had been formed that thymoquinone is an intermediate product in the formation of the blue color. To explain this one or two points may be suggested. In the first place, ether and ammonia are not very miscible, and thymol is exceeding- ly soluble in ether. Then, again, the ether, which consists of the upper portion of liquid, prevents to a certain extent the ingress of air; and air, as will be shown later, is an essential factor. 39 TABLE XII. Effect of heat upon the production of the blue color. Thymol 0.5 gm. Ammonia ( 10% ) 100 c.c. Alcohol (95%) 10 c.c. Conclusion. Color well marked within 5 minutes. With- in 10 minutes the color is so intense, that a sample after stand- ing 24 hrs. cannot compare with it. Note. Heating was accomplished on a boiling water bath, under reflux. TABLE XIII. Effect of cold upon the production of the blue color. Into a glass-stoppered cylinder were placed 0.5 gm. thymol and 150 c.c. ammonium hydroxid solution. This was surrounded by ice. Conclusion. Cold prevents formation of blue color. Notes. No trace of color at the end of 5^ hrs., though a control (containing the same quantities, but exposed to room temperature), had become colored in the usual manner. The crystals of thymol were found quite unaffected. TABLE XIV. Effect of ammonia upon thymol in absence of sunlight. 0.5 gm. Thymol 1 Put into dark bottle, then 100 c.c. Ammonia rinto desk locker, and allowed 10 c.c. Alcohol J to stand. Conclusion. Absence of sunlight does not prevent the for- mation of color. Notes. After 2 hours a part of the solution was poured into a transparent bottle, and it was found to have turned green. It continued to change more towards the blue, in every way con- forming to that found in other experiments. TABLE XV. Effect of ether upon thymol in the presence of ammonia. Thymol (about) 0.1 gm. ' Ammonia 50 c.c. Ether 5 c.c. Conclusion. Ether prevents the formation of the color. 4 o Notes. After 20 hours no change could be noticed. Alcohol was then added. 3 days later a slight pink was noticed in the emulsion which had formed on top. For 3 successive days more alcohol was added each day, but no further change was visible. At the end of 11 days the solution showed a decided green- ish tinge, and the pink color in the emulsion was very clear. ARE AMMONIA AND THYMOL ALONE THE ACTIVE FACTORS IN THE FORMATION OF THE BLUE COLOR? Liebermann, in his important paper upon the action of ni- trous acids on phenols [to which reference has already been made (34) ], tells us what led him to this color test. It had been found that ammonia in the presence of air converts orcin into orcein, and it did not seem improbable to him that the ammonia and the oxygen of the air combined to form nitrous acid and water. If that were so, then nitrous acid could be substituted. When nitrous acid was tried, a colored substance was readily obtained, though this did not prove to be orcein.* Liebermann's illuminating work suggested that possibly in our own case the oxygen of the air plays a role in the reaction. The formation of the blue color should therefore be prevented by preventing ingress of air, or replacing air by another gas ; and should be hastened by passing a rapid stream of air through the solution, or adding a suitable oxidizing agent. Several preliminary qualitative experiments were made. The simple method of placing a beaker containing ammonia and thymol into a desiccator and exhausting the air, has the objec- tion that by so doing most of the ammonia is also driven off. The method, again, of filling to overflowing a bottle with am- monia and thymol and then stoppering it, also has the objection that all air, either dissolved or otherwise, cannot be excluded. To overcome this, thymol and ammonia were treated with hy- drogen gas for ten minutes, placed in a tight-stoppered bottle, put in a desiccator, and the air exhausted. After 23 hours' standing, a pale green color formed; but the lack of intensity * "Dieser Farbstoff ist jedoch kern Orcein, sondern ein neue Substanz." of color clearly suggested that the exclusion of air does have an inhibitory influence, and that the ultimate appearance of color was due to air which had found its way in. At the same time, this experiment demonstrated how very readily even a very small quantity of air would suffice to produce the color. Another method was to pass hydrogen for a few minutes through a 20% solution of ammonia (20% because much of the ammonia is lost by evaporation), pour this directly into a bot- tle containing thymol through which hydrogen had also been passed, and continue passing the gas through the mixture for several more minutes. The bottle was now stoppered and al- lowed to stand beneath an inverted funnel connected with a hydrogen generator. After three hours no trace of color could be noticed. A still more efficient means of preventing the formation of color, and one which later led to fruitful results, was to evolve hydrogen within the thymol-ammonia mixture by adding zinc dust. From theoretical conceptions, and as a result of the inhibi- tory influence of the absence of air upon the formation of color, it was argued that an induced current of air should accelerate color production. Here the valid objection can be raised that by passing air through the solution, much of the ammonia is lost, and as the intensity of the color varies with the amount of am- monia present, it would be almost impossible to arrive at any definite conclusion. However, by causing the current of air to pass through several ammonia solutions first, and then through the solution tested, the difficulty was overcome; and it soon became apparent that under these circumstances the color forms sooner than by merely allowing the ammonia thymol mixture to stand. Comparative effects of passing air and hydrogen, and add- ing hydrogen peroxide, to thymol-ammonia mixtures, were made (Tables XVI and XVII). From these and from foregoing ex- periments we were justified in concluding that the oxygen of the air is a vital factor in the ammonia-thymol reaction. The fact that hydrogen peroxide very considerably increased the rapidity of formation of color was further evidence of the important part oxygen played. 42 TABLE XVI. Comparison of effects induced by passing currents of air and hydrogen through ammonia-thymol mixture. 1 (air) 2 (hydrogen) 3 (control) Thymol 0.5 gm. 0.5 gm. 0.5 gm. Ammonia (20%) 100 c.c. 100 c.c. Ammonia (10%) 100 c.c. (Used 20% ammonia because much of it was removed.) Conclusion. Replacing air by hydrogen prevents formation of color. Notes. Through 1 a current of air was passed, through 2 a current of hydrogen, the rate of flow having been so regulated as to be as nearly as possible the same (16 bubbles in 10 sec- onds). Commenced at 11 : 15 a. m. 2 p. m. Put all 3 solutions into bottles of equal size, and compared colors. 1. Greenish-blue. 2. No color. 3. Greenish-blue. The mere transference from one bottle to another was apparently enough to give 2 the necessary amount of oxygen; for when the process was now continued, and 3^ hours later color compari- sons were once again made, 2 showed a VERY faint green, the other two being quite bluish. TABLE XVII. Comparison of action of air (current), hydrogen and hydro- gen peroxide on ammonia-thymol. 1 234 (control) (hydrogen (air cur- (hydrogen) peroxide) rent) Thymol 0.5 gm. 0.5 0.5 0.5 Ammonia (20%) 100 c.c. 100 c.c. Ammonia (10%) 100 c.c. 100 c.c. Hydrogen peroxide (3%) 10 c.c. Conclusions. Replacement of air by hydrogen prevents for- mation of color. Hydrogen peroxide increases rapidity of formation of color another proof of how important a part oxygen plays in this reaction. 43 Notes. A distinct green was produced in 2 in 10-15 min- utes, a bluish-green in 25 minutes, whereas without hydrogen peroxide a FAINT green tinge is noticeable only after 40-50 min- utes. After 5 hours 4 did not show a trace of color. 1 and 3 again failed to exhibit a marked contrast. Methods of visibly demonstrating the absorption of oxygen were now attempted. One of these was to have a U-tube ar- rangement attached to the bottle containing thymol and am- monia. It was thought that as the oxygen within the bottle would be absorbed, the difference in pressure would show itself by the difference in the level of liquid in the U-tube ; but though the color appeared no difference could be noticed. Substituting a narrower U-tube showed alterations in level, but as these in- dicated expansion at one time and contraction at another, the cause was ascribed to differences in the temperature of the room. Another scheme attempted was to put some thymol and ammonia in a test tube and invert it over water, the mixture but partly filling the tube. Any oxygen absorbed would, of course, be shown by a rise of the solution in the tube. But here again the color appeared without showing any increase in volume. However, four or five days later, when the solution had become much darker (more intensely blue)*, the volume had increased. Since then, several repetitions (in all cases using "controls"), have confirmed this observation. The addition of zinc dust to the solution (thymol 0.5 gm. ; ammonium hydroxid solution 100 c.c.), after the blue color had well formed (after 24 hours' standing) caused, what at first ap- peared to be, a driving of the color to the surface, leaving the major portion of the solution colorless. Upon closer examination the conclusion was reached that reduction was really going on, that the color was being destroyed, but that at the surface there was sufficient oxygen to overcome the reducing influence a suppo- sition which the shaking of the solution tended to confirm, for * The color formation continues for days even weeks, the solution constantly becoming darker. One of the interesting sequels to these observations will be to study the velocity of this color formation. 44 it was noticed that then the blue color spread downward, show- ing that the induced air was overcoming the influence of the hy- drogen evolved. Ether was now added this to act as a me- dium between the air and the rest of the solution. The bottle was well stoppered. The blue that was in the solution was extracted by the ether, forming a red ether layer. The next morn- ing the ether layer was found to be quite colorless. This clearly suggested that the nascent hydrogen had succeeded in reducing the colored substance. From this it followed that by loosening the stopper the air from the outside would once again recolor the ether layer. And, indeed, within five minutes the ether portion be- came pink. The bottle was once again tightly stoppered and al- lowed to stand. Three hours later the pink color had altogether disappeared. By loosening the stopper the color very soon re- appeared. These experiments were repeated many times on the same and different samples, with precisely the same result. We are therefore justified in concluding that nascent hydrogen destroys the color, the oxygen of the air being capable of reproducing it. Attempts were now made to isolate this reduced compound. Having in mind the extreme readiness with which it oxidizes, precautions were taken to prevent this. To a blue solution in a flask zinc dust and ether were added. Besides the hydrogen that was evolved in the flask, hydrogen gas was passed through the entire outfit throughout the whole of the experiment. When the ether layer had become colorless, part of this was poured through a side tube into a perfectly dry flask, which, in turn, was connected with an exhaust. This flask was now surrounded with warm water, the evaporation of the ether being hastened by the exhaust, care being taken to so regulate this that no air bubbles were introduced into the hydrogen generator. But the residue that was left was reddish, showing that it had oxidized in some way. This was repeated several times, the utmost pre- cautions being taken with regard to leagage, but the same re- sults were invariably obtained. In one of these experiments difficulties were experienced in decolorizing the ether layer. It was recalled that the blue solution used contained alcohol, and it was thought that this 45 might have a retarding- influence. With the object of testing this, four samples, all containing the blue solution, zinc dust, and ether, were taken, and to two of them alcohol was added. Even on the addition of the alcohol a blue layer was formed between the pink ether layer and the rest of the solution (a phenomenon which has since been repeatedly noticed. Indeed, one of the ways of hastening the oxidation of the decolorized portion is to add alcohol):* All four samples were allowed to stand overnight. In the morning it was found that the ether layers in the two containing alcohol had not decolorized, where- as the other two had. In another experiment, instead of starting out with the blue solution, thymol, zinc dust, ammonia, and ether were added to the flask. This was allowed to stand for three days, but no color was visible in the ether layer. Part of the latter was now poured into another vessel, and the ether evaporated (in a manner already described). What remained was a yellow liquid, in color identical with that obtained by the action of dry ammonia gas upon dry thymol after it had been allowed to stand for a little while. In several other experiments, in addition to the zinc dust, sodium amalgam was also added, as the evolution of hydrogen was far more rapid. The ether layer was decolorized in about two hours, whereas about 24 hours was required without the use of the amalgam. Here purplish particles and a bluish liquid constituted the residue after the evaporation of the ether. Several other arrangements were tried, but in all cases colored deposits were left. At this point a study was made of the action of dry am- monia upon dry thymol. Dry ammonia gas prepared by heat- ing concentrated ammonia solution and passing it through a soda lime tower was passed into thymol from which all moisture had been removed by keeping it in a drying oven for one to two hours around 50, and leaving it in an exhaust desiccator * This appears contradictory to what was found previously (see page 32). It must be remembered, however, that here we are dealing with a substance which very readily takes up oxygen, and the alcohol may be regarded as an "oxygen carrier." 46 overnight. The thymol readily absorbed the ammonia, a col- orless liquid resulting. This solution, however, gradually turned yellow, then brown, and finally a port wine color all in the course of four or five days. That moisture is also necessary in the formation of color seems to be indicated. EFFECT OF SEVERAL REAGENTS UPON THE BLUE SOLUTION. A closer investigation of the pigment was now decided upon, but before this could be carried out a convenient and rapid method of isolating it from the blue solution was necessary. Gies found that ether extracts the blue color forming a red ether layer (1). Chloroform and toluol act similarly. Al- cohol, beyond tending to emphasize a greenish tinge, seems to have no marked effect. Sulphuric acid converts the blue color into red, purplish drops appearing at the surface, the solution being very cloudy. Hydrochloric and nitric acids act similarly. When the red ether extract is evaporated, an amorphous purplish mass remains* which dissolves in sodium hydroxide to a blue solution, and this in turn is converted into a red solution by sulphuric acid a property which would suggest its use as an indicator (1). One of the convenient methods of extracting the blue color was to dilute the solution with water, neutralize and make slightly acid with sulphuric acid (using phenolphthalein as in- dicator), extract with ether, and evaporate the ether portion.f *Subsequent repetitions gave many beautiful blue crystalline deposits, but further work led to the belief that these crystals were largely com- posed of unaltered thymol that, in fact, they were thymol crystals col- ored by the pigment. t This method required much less ether. The same end could be accomplished by nearly saturating the blue solution with salt and then extracting with ether. For fear of modifying the process or introduc- ing impurities, neither method was much used, direct extraction with ether being preferred. 47 TESTS FOR THE PRESENCE OF NITROGEN IN THE PIGMENT. The opinion had been formed that the pigment was thymoquinonethymolimid. As a ready means of testing this sup- position, it was decided to determine the amount of nitrogen in the substance. A sample of the blue solution was extracted with ether and the ether evaporated. The residue was dried in a desiccator for several days, and two portions were weighed out: A. 0.1165 gm. ; B. 0.1217 gm. Each was treated with 20 c.c. of sul- phuric acid and a trace of copper sulphate, and the Kjeldahl run in the ordinary way. The ammonia was received in 30 c.c. N/5 sulphuric acid. After the operation, A required 29.8 c.c. N/5 alkali and B 30 c.c. This at once raised certain questions, the most prominent of which was whether the substance was really the genuine blue ?* It was quite conceivable that the presence of a preponderating amount of some other substance, nitrogen-free possibly un- changed thymol would account for this result. Before repeating the above under modified conditions a qualitative test for nitrogen was made. From the very nature of the reaction it seemed certain that the blue substance con- tained nitrogen. However, the quantitative determination gave rise to doubt. The qualitative test was made by fusing a sample with sodium and treating a solution of the product with ferric and ferrous salts in the presence of hydrochloric acid. Not a trace of Prussian blue was obtained. This was repeated three times with the same result. (Lassaigne's test). In the possibility that some substance in the pigment pre- vented the reaction, some of the pigment was mixed with an- other substance known to contain nitrogen (amidoacetophenone was used), but here no difficulty was experienced in getting a positive result. These results seemed most perplexing. It appeared in the * Thymoquinonethymolimid should have yielded about 4.5% nitrogen. highest degree improbable that the pigment contained no nitro- gen, and yet the test is considered a reliable one even for com- parative traces of that element.* Another method of attacking the problem was now consid- ered. If, in the course of the reaction ammonia is taken up, then by titrating the amount that remains, the necessary data should be obtained. This was carried out by dissolving 0.1 gm. thymol in 20 c.c. 10% ammonium hydroxid solution, allowing to stand for two days, diluting to 200 c.c., extracting the blue with 50 c.c. ether, and titrating the aqueous portion, 25 c.c. at a time. Exactly the same treatment was accorded a "control," containing, of course, no thymol. (In all cases the ammonia was added to an excess of standard acid, and the excess determined.) Congo red served as the indicator. Methyl orange could be substituted. Blue Control 25 c.c. required 42.4 c.c. acid 42.35 c.c. 43.25 " 4245 c.c. 43 " 42.25 c.c. 43.1 " 42.35 c.c. 200 c.c. (20 c.c. 10% ammonia) 342.5 338.7 c.c. This actually tended to show that there was more ammonia * Lassaigne, the originator of the Prussian blue test for nitro- gen (45), says: "Le precede que nous avons mis en pratique, apres 1'avoir soumis a de nombreux essais, est si sensible, qu'il permet de reconnaitre la presence de 1'azote dans des quantites de matieres azotees aussi petites que celles que les meilleures balances peuvent a peine apprecier." Graebe (46) failed to get the ferrocyanide when using the per- bromide of azonaphthalene. Diazo compounds in general failed to give it, as the nitrogen was evolved at a low temperature. In the cases of compounds containing sulphur, unless a large excess of sodium was used, the test would not always be successful. Jacobsen (47) cites the fact that the sulphur compound produced by the oxidation of parasulphaminetoluic acid failed to give the nitrogen test, though by using an excess of sodium a positive result was ob- tained. Tauber (48) confirms Graebe's observations. See Spiegel, "Der Stickstoff" (1903), p. 835. Dr. Heidelberger, in a private communication to the author, stated that negative results were sometimes obtained by the use of too large a quantity of ferric chloride. The merest trace just sufficient to know that some had been added was all that was required. 49 at the end of the reaction than at the beginning! Several repe- titions gave equally anomalous results. Larger quantities of thymol were now used, and the follow- ing mixtures were prepared : 123456 Thymol 2 gm. 5 gm. 10 gm. 15 gm. 20 gm. Ammonia 20 c.c. 20 c.c. 20 c.c. 20 c.c. 20 c.c. 20 c.c. These samples were allowed to stand for two days. Into 6 (counted as "control"), 5 gms. of thymol were added just before titration. It was thought that possibly the presence of thymol would prove an interfering factor, and thereby shed light upon the results). Each was made up to 100 c.c. with water, 50 c.c. ether were added, shaken well, and 10 c.c. with- drawn from the non-ether layer. This was added to 50 c.c. N/5 acid solution, and the excess titrated, with the following results : 1 required 38.9 c.c. Average for 20 c.c. ammonia 387.75 38.65 c.c. 2 required 40.9 c. Average for 20 c.c. ammonia 407.7 40.65 3 required 43.4 Average for 20 c.c. ammonia 433.5 43.3 4 required 44.35 Average for 20 c.c. ammonia 442.7 44.2 5 required 44.35 Average for 20 c.c. ammonia 442.7 44.2 6 required 39 Average for 20 c.c. ammonia 387.7 38.55 Here a similar tendency to what was noticed before is shown ; namely, that the thymol, instead of decreasing the amount of ammonia, increases it! But these results can be explained satisfactorily. Concentration of solutions were not taken into consideration. Since in 5 much more thymol is present than in 1, and hence takes up more volume, diluting each to 200 c.c., and extracting 10 c.c. from each, will yield a more concentrated solution from 5 than from 1. To obviate these shortcomings it was decided to titrate the whole rather than take an aliquot portion. 1 2 3 Thymol 2 gm. 10 gm. Ammonia (10%) 20 c.c. 20 c.c. 20 c.c. 50 (To 3 five gms. of thymol were added just before the ti- tration.) The solutions were allowed to stand for four days. 20 c.c. ether were then added to each and allowed to stand for another day.* Each was then poured into a separating funnel, the bot- tle rinsed with distilled water, and the washing added to the solu- tion in the sep. funnel. The colorless portion (lower layer) was run into 450 c.c. sulphuric acid, and the residue (the red ether layer), washed three times with distilled water, and the wash- ings added. To test the residue for acid it was again washed with water, and the washings run into a known volume of acid. This was repeated until all the acid had been washed out as determined by titration. 1 required 433.65 c.c. acid. 2 required 433.60 c.c. acid. 3 required 433.70 c.c. acid.| These results were the most trustworthy as every possible precaution was taken. Apparently, these tended to confirm the indication from the Kjeldahl determination, as well as the qualita- tive test, that the pigment did not contain nitrogen. A modified Kjeldahl for nitro compounds (49) was now employed. It was argued that possibly the nitrogen was in such a form as to make the use of sulphuric acid alone unre- liable. Two samples of 0.5 gm. each of the pigment were taken. To each 30 c.c. of sulphuric acid containing 2 gms. of salicylic acid were added. The mixture was allowed to stand for twenty minutes. After the addition of 2 gms. of zinc dust, the mix- ture was first gently warmed, and then heated more vigorously. Mercury (about 1 gm.) and potassium sulphate (about 10 gms.) were added, and the whole boiled. The boiling was continued for about 1^ hours after the solution had become colorless. *It is needless to add that in all these operations perfect-fitting stoppers for the bottles were used. f These figures cannot well be compared with the foregoing ones, as the ammonia used here was from a different sample, and though all the samples were approximately of 10% strength, none of them was exactly so. This was cooled somewhat, potassium sulphide and excess of sod- ium hydroxide were added, and the whole distilled into 75 c.c. of N/o acid solution : 1 required 74.2 c.c. alkali. 2 required 74.5 c.c. alkali. Control required 74.1 c.c. alkali. This was further evidence in favor of what had been found before. At this stage attempts were begun to isolate the pure pig- ment. As is recorded elsewhere, one of these methods was to filter a blue solution that had been standing for many weeks. Naturally, that which we were most eager to do was to test the precipitate on the filter paper a sticky mass for nitrogen. Some of this filter paper was cut up, well soaked in ether, and the ether solution poured into two tubes in which the nitrogen test was to be run. These were warmed to 50 to expel the ether. To one a small piece of sodium was added, to the other potassium. Neither gave an immediate test, but after standing about 24 hours a distinct Prussian blue precipitate was detected in both solutions.* ISOLATION OF THE BLUE PRODUCT. From the foregoing work it became clear that before any further headway could be made, a method of isolating the pig- ment, uncontaminated with thymol or any other substance one that may be formed as an intermediate product would have to be devised. For some time in the past it had been recognized that the blue crystals must consist largely of unaltered thymol the very shape of the crystals suggested this. Moreover, the fact that they gave a negative nitrogen test simply led to the conclusion that the amount of colored substance present was very small in comparison with the amount of thymol, for it was inconceiv- able from the very nature of the reaction that nitrogen should be absent. The task now was to find a complete method of separating and isolating the blue product. *We may anticipate here and state that in subsequent Kjeldahl de- terminations where the pure pigment was used, nitrogen was invariably found to be present. 52 1 gm. of thymol and 150 c.c. of ammonium hydroxid solution were placed in a dialyzer (parchment paper) and suspended in wa- ter. After seven days the bag was opened, and the solution found to be perfectly colorless, though parts of the thymol crystals showed a pink color. The outer solution became turbid upon heating, and readily responded to the sulphuric acetic acid test for thymol. A blue solution that is, a mixture of thymol and ammonia which had been allowed to stand, and had become blue was now substituted in the dialyzer, and suspended in water. After some time, the outer solution became blue. Evi- dently, then, this could not prove a method of separation. Bearing in mind the property possessed by aluminum hy- droxide of carrying down coloring matter if precipitated in the medium, a solution of aluminum sulphate was added to the blue liquid (which, of course, contained ammonium hydroxide). The precipitate was allowed to stand for one hour and fil- tered. A flesh-colored precipitate was obtained, the nitrate show- ing just the slightest bluish tinge, which, however, became more pronounced the longer it stood. When the precipitate was ex- tracted with ether only a part went into solution. By evaporating the ether extract a violet residue was left, but no crystals were evident. This method had several objections, not the least of which was the time element. Attempts were now made to prepare one or two derivatives of thymol in the hope that the blue product would not be affected thereby. First, the formation of thymol iodide was tried. Some of the blue crystals were dis- solved in sodium hydroxide, and to the solution iodine in potas- sium iodide was added till a precipitate was formed. Upon fil- tering the filtrate proved to be colorless, thus showing that not only the thymol but the blue had been acted upon. Next the benzoyl derivative was attempted. To 1 gm. of the crystalline blue 0.5 gm. water, 1 c.c. benzoyl chloride, and sodium hydroxide till the solution was alkaline were added. The mixture was stirred till the smell of benzoyl chloride had practically disappeared. It was now poured into a large ex- cess of water. The expected white precipitate was obtained, 53 but the oily, brownish layer on top showed that the blue had decomposed. It was noticed that in all the blue solutions that had been standing for some months a blue sediment, small in amount, had settled out. A sample of such a blue solution (which we shall label X) which had been standing for about 3 months, was filtered (using three filter papers). The filtrate (Y) was much clearer than the original solution, though it darkened and became opaque on standing. On the filter paper there was a dark purplish-bluish mass, oily rather than solid (Z). From each of X and Y some solution was taken, extracted with ether, and the ether evaporated ; Z was likewise extracted with ether, and the solvent evaporated. The ether solution of X was dark violet in color; of Y rose-red; of Z violet with a reddish tinge. After the ether had been evaporated, X showed dark violet crystals,* Y crystals that were almost colorless, though here and there tinged with a light violet, and Z no crystals at all, but a violet sticky mass which adhered to the sides of the beaker. Apparently in Z no thymol was present. This method of separation seemed to be satisfactory on the face of it. But there were two strong objections to its ex- tensive use: a) The long time necessary to obtain the blue; b) The small quantity obtained. To obviate the first, resort was had to heating. Quite satisfactory results, similar to those recorded, were obtained. Heating on the water bath for ten to twelve hours accomplished what standing at room tempera- ture would have taken weeks to attain. However, too large quantities of thymol could not be used, as the thymol remained suspended, and, of course, contaminated the precipi- tate. But even here the yield of blue was such that any ex- tensive investigation of it was out of the question. Another method was now tried. Though in so far as iso- lating the blue was concerned it proved a failure, yet some in- teresting results were obtained. 10 gms. of thymol were dis- *In most cases where these crystals would form, a creeping ten- dency on their part would be exhibited. Many of the beakers were found to be covered with these crystals not only on the inside but out- side as well. 54 solved in two liters of ammonia (IQ%). The next day 30 gms. of thymol were added, and immediately part of it turned vio- let. Five days later the quantity was increased with 20 more gms. After an hour or so, the whole solution under- went a complete change. From blue with a tinge of violet, it now became violet with red predominating. A deep purple layer formed itself on top as if ether had been added (1). Eleven days later the purple layer (A) was separated from the rest (B). To B twenty more gms. of thymol were added. It was reasoned that any thymol that was t added to B should cause a further production of A, since B already had a large excess of thymol. This proved to be the case, for more of the purplish variety was soon obtained. By washing A several times with water it crystallized out to a violet mass. This crystallization was undoubtedly due to the presence of a large quantity of unaltered thymol. ..&, The odor of A was most peculiar. It recalled thymol some- : .what, though quite disagreeable. Upon covering the precipi- tate and allowing it to stand for several days, a most disagree- able odor, vividly suggesting pyridine, was developed. Various substances were now tried upon thymol and the blue pigment in the hope of finding a medium that would prove a solvent for the one and not for the other : 1. Water Thymol Pigment 2. Saturated sodium Insoluble Insol. chloride Insol. Insol. 3. Alcohol (95%) Very sol. colorless Very sol. Violet (bluish) 4. Ether V. sol. Colorless sol., V. sol. Pink 5. Petrol, ether though not as much Sol. Pink 6. Glacial acetic acid V. sol. Colorless V. sol. Violet (reddish) 7. Benzine V. sol. Colorless V. sol. Pink 8. Carbon bi-sulphide V. sol. Colorless V. sol. Violet (bluish) 9. Acetone V. sol. Colorless V. sol. Violet (bluish) 10. Chloroform V. sol. Colorless V. sol. Violet 11. Ethyl acetate V. sol. Colorless V. sol. Violet 12. Methyl alcohol V. sol. Colorless V. sol. Violet (bluish) 13. Amyl alcohol Sol. Colorless Sol. Violet (bluish) 14. Toluene V. sol. Colorless V. sol. Violet (reddish) 15. Xylene V. sol. Colorless V. sol. Violet (reddish) 16. Lyridine V. sol. Colorless V. sol. Violet (bluish) 17. Phenol (water solu-Insol. (hot and cold) Insol. (hot and cold) tion) Sol. yellow Sol. Violet 18. Nitrobenzine Sol. yellow Sol. Violet (bluish) 19. Aniline 55 The common acids and alkalis proved of no better service. Concentrated sulphuric acid, however, forms a sulphonate which is soluble in water. To some of the pigment, then, some cone, sulphuric acid was added, and the solution allowed to stand for several hours. This was now poured into a large excess of water, and a small quantity of a greenish-colored pre- cipitate was obtained. As thymol is volatile with steam, this was thought of as a possible means of separation, and it proved very good. When the pigment was steam distilled, thymol came over, leaving the pure pigment behind. The steam distillation was continued till the distillate no longer gave a sulphuric acid glacial acetic acid test (a "control" was used, for even with a "control" a slight pink is obtained). The solution in the distilling flask was now extracted with ether (sodium chloride was for some time used to first saturate the solution, but this contaminated the ether extract, and was eventually discarded), and the ether evaporated. An intense deep purple product was obtained, but not a trace of crystalline matter. Though this method does not yield the substance in a crys- talline form, as does the sulphuric acid method, yet in subse- quent work it was preferred because : a) Purity of product was more assured; b) product more easily obtained and in larger quantities ; and c) less time required. THE CHEMICAL CONSTITUTION OF THE PIGMENT. The work done so far does not justify definite conclusions with regard to the exact chemical nature of the compound (or compounds) for it is probably a mixture obtained by the ac- tion of ammonia or thymol. However, clues to the path to be traced are not wanting. Wurster (36), in the article so frequently referred to, states that his blue, obtained by the action of hydrogen peroxide and ammonia on phenol, is phenolquinoneimid. This he prepared in two other ways : 1) By the addition of ammonia to a watery solution of quinone in excess of phenol ; 2) by dissolving />-amido- phenol in sodium hydroxide (which becomes yellow in contact with air) and adding phenol. Hirsch (37) prepared the same 56 phenolquinoneimid by dissolving quinonechlorimid in phenol in the presence of sodium hydroxide. Wurster carefully differen- tiates his pigment from that obtained by Liebermann (50), differ- ences in spectroscopic behavior and in the colors obtained by the addition of ether to acid solutions, justifying his conclusion.* Thymol, he records, behaves similarly. Now, if this be correct, then Decker and Solonina's view that the Liebermann pigment obtained from thymol (prepared by the action of nitrous acid on thymol) is also thymoquinone- thymolimid (51) is erroneous. From our own observations so far. we have been led to believe that the two are different. Some of Liebermann's pigment was prepared and compared with our own, with the results indicated below : Liebermann's Our pigment Water Insol. reddish Insol. violet Alcohol Purple (red) Purple (blue) Ether Red Violet (various shades) Glacial acetic Orange-red Violet Amyl alcohol Purple (red) Violet (blue) Liebermann's pigment, when steam distilled, taken up with ether, and the ether evaporated, gives the same pasty residue as ours, but the color is much redder. Whereas our pigment, when dissolved in cone, sulphuric acid, and the solution poured into water gives a very fine greenish-blue precipitate, which is almost impossible to filter, Liebermann's gives a flocculcnt reddish pre- cipitate, readily filtered. That the blue pigment may prove to be a mixture of more than one substance one of which is probably thymoquinone- thymolimid does not seem at all improbable, for various shades of the colored pigment (the violet varying from the red to blue) have been obtained. *"Da die saure Losung des Farbstoffs in Aether roth ist, wahrend der Liebermannsche Farbstoff aus verdiinnter saurer Losung in Aether mit gelber Farbe ubergeht, so ist die Verschiedenheit der beiden Kor- per schon wahrscheinlich ; dies geht auch aus dem spectroskopischen verhalten des Farbstoffs hervor, welcher zwar eine nicht scharf begrenzte Absorption im Roth zeigt aber nicht den charakteristischen Streifen der Liebermann'schen Farbstoff." 57 THE ACTION OF SODIUM HYDROXIDE, POTASSIUM HYDROXIDE AND BARIUM HYDROXIDE ON THYMOL. Early in these investigations it became of interest to com- pare the action of ammonia with that of other alkalis, with the view to ascertaining whether the blue color formation was due specifically to ammonia or merely because ammonia belonged to the class of alkalis. In this respect very few experiments were needed to show how specific in function was the ammonia. Sodium hydroxide developed a pink color (Table XVIII), the intensity of which decreased with increasing quantities of alkali (Table XIX), though it increased up to a certain limit (Table XX), and increased in the presence of alcohol (Tables XXI and XXII). Heat accelerates the color formation (Table XXIII). The red color could not be extracted with ether, but upon evaporating the ether portion, thymol was recovered (Note A). The action of zinc dust was somewhat peculiar (Note B). Potassium hydroxide acted similarly (Tables XXV and XXVI). Barium hydroxide failed to give any color possibly owing to the weakness of the solution (Table XXVII). TABLE XVIII. Comparative action of sodium hydroxide and ammonium hydroxide upon thymol, alcohol being absent. 1. 2. Thymol 0.5 gm. Thymol 0.5 gm. Sodium hydroxide (10%)* 10 c.c. Ammonia (10%) 100 c.c. Water 100 c.c. Water 10 c.c. Conclusion. Whereas 2 develops the usual green to green- ish-blue color, 1 develops a pink. Notes. It took from 24-36 hrs. to develop a pinkish tinge in 1. Used more of ammonia than sodium hydroxide as latter is a stronger base. *Unless otherwise stated, the sodium hydroxide used throughout was 10%. TABLE XIX. Effect of different quantities of sodium hydroxid upon a constant quantity of thymol. 12345 Thymol 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. Sodium hydroxide 10 c.c. 20 c.c. 30 c.c. 40 c.c. 50 c.c. Water 50 c.c. 40 c.c. 30 c.c. 20 c.c. 10 c.c. Conclusion. Pinkish color develops, most pronounced in 1, least in 5. That is, intensity of color decreases with in- creasing quantities of sodium hydroxide. Notes. Within two hours a pink tint can be noticed in all. The intensity of color seems to reach a maximum in about 48 hours and then does not seem to change (whereas when am- monia is used the color deepens continuously). Two days after the maximum color had been reached, the pink color had disappeared altogether. 10 c.c. of alcohol were then added to each. A pinkish color now gradually developed, being most intense in 1, least in 5. TABLE XX. What amount of sodium hydroxid, keeping thymol and al- cohol constant, would give the maximum color? 12345 Thymol 0.5 gm. 0.5 gm. 05 gm. 0.5 gm. 0.5 gm. Sodium hydroxide 2 c.c. 4 c.c. 6 c.c. 8 c.c. 10 c.c. Water 58 c.c. 56 c.c. 54 c.c. 52 c.c. 50 c.c. Alcohol 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. Conclusion. Maximum coloration obtained with 10 c.c. sodium hydroxide solution (10%). Notes. Color least intense in 1, most in 5; and from pre- vious work it was found that when more than 10 c.e. sodium hydroxide solution is present, the color is less intense. 59 TABLE XXI. Effect of variable quantities of sodium hydroxid upon a con- stant quantity of thymol, in the presence of a constant quantity of alcohol. Thymol 1 05 gm 2 0.5 gm. 3 0.5 gm. 4 S crm 5 O S crm Sodium hydroxide Water Alcohol .. . 10 c.c. . 50 c.c. 10 c.c. 20 c.c. 40 c.c. 10 c.c. 30 c.c. 30 c.c. 10 c.c. u.o gm. 40 c.c. 20 c.c. 10 c.c. u.o grn. 50c.c. 10 c.c. 10 c.c. Conclusion. Intensity of color varies inversely with amount of sodium hydroxide present. Color increased by pres- ence of alcohol. Notes. Though this series was prepared 27 hours later than the one containing no alcohol, the colors were far more intense ; so much so, that the least intense here was found to be more colored than the most intense there. TABLE XXII. Effect of different quantities of alcohol upon a relatively large quantity of thymol in the presence of a constant quantity of sodium hydroxide. 12345 Thymol 1 gm. 1 gm. 1 gm. 1 gm. 1 gm. Sodium hydroxide 10 c.c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. Alcohol 10 c.c. 20 c.c. 30 c.c. 40 c.c. 50 c.c. Water 50 c.c. 40 c.c. 30 c.c. 20 c.c. 10 c.c. Conclusion. Color goes from pink to brown in proceeding from 1-5. TABLE XXIII. Effect of heat upon the development of the pink color by the action of sodium hydroxide upon thymol in the presence of alcohol. Thymol 0.5 gms. Sodium hydroxide 50 c.c. Alcohol 10 c.c. Conclusion. Color developed in 15-20 minutes. 6o NOTE A. Action of ether upon the red solution formed by the action of sodium hydroxide upon thymol in the presence of alcohol. To a small quantity of the red solution (taken from a bot- tle in which the red solution had been kept for days, and which was getting darker and more brownish with time), some water was added, shaken, and then some ether added (the object of the water was to dissolve out the alcohol, and thereby lessen its effect upon the ether). It was hoped that the color (as in the case of the NH 4 OH thymol blue formation) would be ex- tracted by the ether, but no such thing happened. On allow- ing to stand overnight it was found that the color of the solu- tion had changed from a dark brownish-red to a light yellow, the ether layer on top remaining colorless. For 4 days no fur- ther change was visible. By separating the ether layer from the rest of the solution, and evaporating the ether rhombic crys- tals, resembling thymol when a concentrated solution in acetic is poured into water, were obtained. The m. p. of this (48) strengthens this belief. NOTE B. Action of Zn dust upon the red solution formed by the action of sodium hydroxide upon thymol in the presence of alcohol. To a red solution composed of thymol 0.5 gms., sodium hydroxide solution (10%) 10 c.c., water 50 c.c., alcohol 10 c.c., and which had been standing almost a month, some Zn dust was added. In 15-20 minutes the solution had become completely de- colorized. But the pink again appeared this time on top. The ad- dition of ether caused the pink to be taken up by the ether as pink. TABLE XXIV. Effect of various mixtures of sodium hydroxide and am- monia upon a constant quantity of thymol. 123456789 Thymol 0.5 gm. .5 gm. .5 gm. .5 gm. .5 gm. .5 gm. .5 gm. .5 gm. .5 gm. Ammonia 90 c.c. 80 c.c. 70 c.c. 60 c.c. 50 c.c. 40 c.c. 30 c.c. 20 c.c. 10 c.c. Sodium hydroxide 10 c.c. 20 c.c. 30 c.c. 40 c.c. 50 c.c. 60 c.c. 70 c.c. 80 c.c. 90 c.c. 6i TABLE XXIV. Continued. After 45 hours 1 , Greenish-blue ; 2, less intense ; 3, faintest greenish tinge; 4, faint pink; 5, somewhat more so; 6, about same; 7, lighter; 8, still lighter; 9, almost no pink. Two days later. 1, A sort of dark, somewhat dirty green, radically differing from ammonia, thymol alone (which becomes blue) ; 2, same, but much lighter; 3, slight pink with a shade of the green; 4, light pink; 5, 6, slightly more intense; 7, slightly less ; 9, less, just faint pink. TABLE XXV. Effect of variable quantities of potassium hydroxide upon a constant quantity of thymol, in the presence of a constant quantity of alcohol. 12345 Thymol 0.5 gtn. 0.5 gm. 0.5 gm. 0.5 gtn. 0.5 gm. *Potassium hydroxide.. 10 c.c. 20 c.c. 30 c.c. 40 c.c. 50 c.c. Water 50 c.c. 40 c.c. 30 c.c. 20 c.c. 10 c.c. Alcohol (95%) 10 c.c. 10 c.c. 10 c.c. 10 c.c 10 c.c. Conclusion. Pink color, gradually tending to brown ob- tained; most in 1, least in 5. Notes. Differences in intensity of the pink, when first de- veloped, could not be distinguished. Five days after commencing the experiment the pink had almost wholly given way to brown, the intensity remaining as be- fore. TABLE XXVI. Effect of different quantities of potassium hydroxide upon a constant quantity of thymol in the absence of alcohol. 12345 Thymol 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. Potassium hydroxide.... 10 c.c. 20 c.c. 30 c.c. 40 c.c. 50 c.c. Water 50 c.c. 40 c.c. 30 c.c. 20 ac. 10 c.c. Conclusion. Pink color developed, most in 1, least in 5. Notes. After 40 hours' standing, there wae but the faintest trace of pink visible. Within the next 24-30 hours it had in- creased appreciably, though the color still remained very slight. * Unless otherwise stated the strength of the potassium hydroxide used was 10%. 62 After 5 days the color had disappeared, and had given place to a slight brownish color; most in 1, least in 5. On the 7th day, 10 c.c. alcohol added to each. A pink color gradually developed in the next 12 hours. At first the intensity was about the same in all, but 2 days later it was most marked ir: 1, least in 5. TABLE XXVII. Effect of different quantities of barium hydroxide upon a constant quantity of thymol in the presence of a constant amount of alcohol. 12345 Thymol 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gm. 0.5 gin. Barium hydroxide (sat. sol.) 10 c.c. 20 c.c. 30 c.c. 40 c.c. 50 c.c. Water 50 c.c. 40 c.c. 30 c.c. 20 c.c. 10 c.c. Alcohol 10 c,c. 10 c.c. 10 c.c. 10 c.c. 10 c.c. Conclusion. No color. Notes. Though standing for 17 days, no color developed. All cloudy. 1 showed a beautiful hexagonal crystal, slightly pinkish around its borders ; 2 showed several such crystals of smaller size, joined together, with pink more pronounced ; 3 and 4 had circular oily layer on top, of same color as solution (cloudy- white) ; 5 same, but slightly pinkish. The crystal in (1) was taken out. When broken up, it was found to smell strongly of alcohol. It would seem as if the alcohol had concentrated around the point. The crystal, though pinkish, when finely ground, gave a fine white powder, resembling thymol. Its m. p. (49) strengthens the assump- tion. ADDENDUM. In the course of one of the experiments it was noticed that by additions of a fairly large excess of thymol to a 10% am- monium hydroxide solution in a glass-stoppered bottle, the blue color tended gradually to disappear. In time a completely color- less, though slightly turbid, solution was obtained, with a purplish oily layer on top. Not the least remarkable was the behaviour of 63 this solution with temperature changes. Warming around 30 caused the solution to become quite opaque; cooling to about 17C again gave a clear greenish solution. Gradually a pale green to blue crystalline substance separated out, and fell to the bottom. When the bottle was opened and shaken up with a new sup- ply of air, the colorless solution became immediately deeply bluish. This implies that the substance was reduced. SUMMARY OF GENERAL CONCLUSIONS. 1. The melting point of thymol, variously given in the literature from 44 to 53, is found to be 50-50.5. 2. The blue color obtained by the action of ammonia upon thymol is not due to any impurity in the preparations of the latter. 3. Within certain limits the intensity of the blue color varies directly with the amount of thymol present. 4. Within certain limits the intensity of the blue color varies directly with the amount of ammonia present. 5. Within certain limits alcohol does not effect the blue color formation. 6. One part of thymol in 25,000 parts of 10% ammonium hydroxide solution shows a distinct color within five days. 7. Heat and hydrogen peroxide accelerate the production of the pigment; cold and hydrogen have a retarding effect; light has no influence; ether inhibits the color formation. 8. Oxygen is an essential factor in the color formation. 9. Moisture is probably also necessary. 10. Nascent hydrogen destroys the red color obtained by extracting the blue solution with ether, the oxygen of the air being capable of reproducing it. 11. Dry thymol absorbs dry ammonia gas forming a col- orless liquid. 12. The pigment isolated by extracting the blue solution with ether and evaporating the latter fails to show the pres- ence of nitrogen. This negative result is due to the presence of preponderating amounts of unaltered thymol (and possibly also intermediate products?). 64 13. The most convenient method of isolating the blue pig- ment in the pure state is to submit the crude product as ob- tained in 12 to steam distillation till the distillate fails to re- spond to thymol tests. 14. The belief is expressed that the pigment is probably a mixture of two or more substances, one of which may pos- sibly be thymoquinonethymolimid. 15. The pigment possesses properties that point to its value as an indicator. 16. The action of sodium hydroxide upon thymol is to produce a pink color, the intensity of which is increased by al- cohol. Increased quantities of sodium hydroxide do not increase the intensity of the color. The color can not be extracted with ether from the solution. 17. Potassium hydroxide acts similarly to sodium hydrox- ide. 18. Barium hydroxide does not produce any color. 65 BIBLIOGRAPHY. 1. Gies: Biochem, Bulletin, 2, 171 (1912). 2. Stiles : Public Health Bulletin, 32. 3. Lehner: Jour. Amer. Chem. Soc., 35, 141 (1912). 4. Stenhouse: Ann. d. Chem., 93, 269 (1855). 5. Haines: Jour. Chem. Soc., 8, 289 (1856). 6. Roscoe and Schorlemmer, Treatise on Chemistry, Vol. Ill, part V, p. 294 (1889). 7. Semmler: Aetherischen Oele, 4, 51 (1907). 8. Arppe: Ann. d. Chem., 58, 41 (1846). 9. Doveri: Ann. d. Chem., 64, 374 (1848). 10. Lallemand: Compt. rend, de 1'Acad. des Sc., 37, 498 (1853) ; 38, 1022 (1854) ; 43, 375, 459 (1856) ; Annals de Chim. et de Phys. [3] 49, 148. 11. Haines: Jour. Chem. Soc., 8, 289 (1856). 12. Stenhouse: Ann. d. Chem., 98, 107 (1856). 13. Kolbe and Lauterman: Ann. d. Chem., 115, 205 (1860). 14. Engelhardt and Latschinoff: Zeit. f. Chem., 615 (1869). 15. Miiller: Ber. d. D. Chem. Gesell., 2, 130 (1869). 16. Pott: Ber. d. D. Chem. Gesell., 2, 121 (1869). 17. Carstanjen: Zeit. f. prak. Chem., [2], 3, 50 (1871). 18. Fittica: Ann. d. Chem., 172, 305 (1874). 19. Widman: Ber. d. D. Chem. Gesell., 15, 166 (1882). 20. Widman: Ber. d. D. Chem. Gesell., 13, 676 (1880). 21. Widman: Ber. d. D. Chem. Gesell., 24, 439 (1891). 22. Beckmann and Eickelberg: Ber. d. D. Chem Gesell., 29, 420 (1896). 23. Dinesmann: D. R. P. 125,097; Chem. Centralb., 2, 1030 (1901). 24. Wallach: Nachrichten k. Ges. wiss. Gottingen, 223 (1903). 25. Semmler and McKenzie : Ber. d. D. Chem. Gesell., 39, 1163 (1906). 26. Semmler: Ber. d. D. Chem. Gesell., 41, 3993 (1908). 27. Robiquet: Ann. d. Pharm., 15, 289 (1835). 66 28. De Luynes : Compt. rend, de 1'Acad. des Sc., 59, 49 (1864). 29. Malin: Ann. d. Chem. u. Pharm., 138, 80 (1866). 30. Lex: Ber. d. D. Chem. GeselL, 3, 457 (1870). 31. Salkowski: Archiv. d. Physiol., 5, 353 (1871). 32. Phipson: Chem. News, 27, 299 (1873). 33. Phipson: Chem. News, 29, 110 (1874). 34. Liebermann : Ber. d. D. Chem. Gesell., 7, 247 (1874). 35. Liebermann: Ber. d. D. Chem. Gesell., 8, 1649 (1875). 36. Wurster: Ber. d. D. Chem. Gesell., 20, 2934 (1887). 37. Hirsch: Ber. d. D. Chem. Gesell., 13, 1909 (1880). 38. Zulkowski and Peters: Monatsh. f. Chem., 11, 227 (1890). 39. Maseau: Chem. Centralb., II, 60 (1901). 40. Watt: Dictionary of Chemistry, 4, 714 (1894). 41. Hirschsohn: Ber. d. D. Chem. Gesell., 14, 2306 (1881). 42. Gies: Biochem. Bulletin, 2, 293 (1913). 43. Gibbs: Jour. Amer. Chem. Soc., 34, 1190 (1912). 44. Davis: Science Progress, Oct., (1912). 45. Lassaigne : Compt. rend, de 1'Acad. des Sc., 16, 387 (1843). 46. Graebe: Ber. d. D. Chem. Gesell., 17, 1178 (1884). 47. Jacobsen: Ber. d. D. Chem. Gesell., 12, 2316 (1879). 48. Tauber: Ber. d. D. Chem. Gesell., 32, 3150 (1899). 49. Sherman: Organic Analysis, p. 13 (1905). 50. Liebermann: Ber. d. D. Chem. Gesell., 7, 1098 (1874). 51. Decker and Solonina : Ber. d. D. Chem. Gesell., 36, 2892 (1903). BIOGRAPHICAL. Benjamin Horowitz was born on August 25, 1888. He re- ceived his secondary school and college training at Raines Foundation School, London, Eng., and Finsbury College, Lon- don, Eng. In the fall of 1907 he entered the School of Chemis- try, School of Applied Science, Columbia University, and in 1911 he received the degree of Chemist. During 1911-1912 he pursued post-graduate work in Organic Chemistry, and in 1912 he received the degree of Master of Arts. During the past year he has been engaged in research work in the department of Biological Chemistry, Columbia University. In the fall of 1911 he was appointed assistant in Organic Chemistry, Clark University, Worcester, Mass., and in February, 1913, he received a similar appointment in the Department of Biological Chemistry, Columbia University. 68 PUBLICATION. Experiments on pigments produced from thymol by the ac- tion of ammonia. Biochem Bulletin, 2, 293 (1913). THE LIBR/RY