TN UC-NRLF SB 32 LIBRARY OF THE UNIVERSITY OF CALIFORNIA, RECEIVED BY EXCHANGE Class The Diffusion of Crude Petroleum through Fuller's Earth. DISSERTATION. SUBMITTED TO THE BOARD OF UNIVERSITY STUDIES OF THE JOHNS HOPKINS UNIVERSITY IN CONFORMITY WITH THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY. By OSCAR ELLIS BRANSKY, EASTON, PA.: ESCHBNBACH PRINTING COMPANY 1911. The Diffusion of Crude Petroleum through Fuller's Earth* DISSERTATION. SUBMITTED TO THE BOARD OF UNIVERSITY STUDIES OF THE JOHNS HOPKINS UNIVERSITY IN CONFORMITY WITH THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY. By OSCAR ELLIS BRANSKY. EASTON, PA.: ESCHENBACH PRINTING COMPANY 1911. ACKNOWLEDGMENT. To President Remsen, and Professors Morse, Jones, Renouf, and Acree, the author is greatly indebted for valuable instruc- tion in the lecture room and laboratory. The author wishes, in particular, to express his gratitude to Dr. Gilpin, under whose guidance this investigation has been pursued, and to Dr. Day, of the United States Geological Survey. Thanks are also due to Prof. Swartz for important suggestions. 222271 CONTENTS. Acknowledgment : 3 Introduction 5 Object of this Investigation 14 Experimental 14 I. Relative Amounts of Oil Lost in Heated and Unheated Fuller's Earth 14 II. The Diffusion of Benzene in Solution through Fuller's Earth 18 III. The Fractionation of Crude Petroleum 31 IV. Chemical Examination of the Fractionated Oils 49 V. Selective Action of Fuller's Earth 53 Summary 5 6 Biographical 5 8 The Diffusion of Crude Petroleum through Fuller's Earth* INTRODUCTION. 1 It is a well-established fact that the petroleum obtained from the sandstones of the Upper Devonian and Mississippian periods, generally known as the Pennsylvania oil, differs markedly from the natural oil found in the Trenton limestone, usually designated as the Ohio oil, or Trenton limestone oil. Both of these oils, in turn, are distinctly different from the petroleum occurring in the loose sands and soft shales of California. The unconsolidated tertiary clays, sands, and gravels in the southern United States, particularly in Texas, yield another variety of petroleum, characterized by proper- ties more or less different from any of the preceding oils. Not only do these differences exist between oils found in separate regions, but extreme variations in color and specific gravity, as well as in chemical composition, often occur between those of neighboring localities. On the other hand, close re- semblances abundantly occur between petroleums of sections widely removed from each other. Some of the South Amer- ican and many of the European oils, for instance, have been found to possess properties very similar to those of the oils of the southern United States; while the oil from the Cornifer- ous limestone of Canada closely resembles the Ohio petroleum. These variations in the oils of the United States and other countries have been carefully studied by many investigators. Such noted workers as Warren, Storer, Mabery, Pelouze, Cahours, Schorlemmer, Beilstein, Markownikoff, Kngler, and Kurbatoff have devoted their lives to this subject. The ques- tion that naturally arises in connection with these variations is: Are these differences fundamental? Is the Pennsyl- vania petroleum as distinctly different from the Ohio oil as one chemical compound is from another? In answer to these 1 This research was aided by a grant received from the C. M. Warren Committee of the American Academy of Arts and Sciences. b questions, the following extract from a paper read by Mabery 1 in 1903 before the American Philosophical Society is of con- siderable importance: "Now, after years of arduous labor, I have reached the conclusion that petroleum from whatever source is one and the same substance, capable of a simple definition a mixture of variable proportions of a few series of hydrocarbons, the product of any particular field differing from that of any other only in the proportion of the series and the members of the series." The evidence supporting this declaration has been and is accumulating constantly, and, at the present time, this view is generally accepted. If petroleum, then, is one and the same substance, how can the extreme variations between the American oils be ex- plained? Were the causes operating in the formation of the Pennsylvania oil, almost barren of sulphur and nitrogenous bodies, different from those acting in the production of the sulphur-bearing oils of Ohio, or the heavy sulphur and nitrog- enous oils of California? To account for the formation of crude petroleum, two views, as is well known, the organic and inorganic, have been ad- vanced. The Pennsylvania oil, according to these theories, may have been formed either from organic or inorganic sub- stances, or from both. It is as yet impossible, however, to state conclusively from which of these sources the oil was de- rived. It is apparent, therefore, that the differences between the Pennsylvania and the Ohio, Texas, and California oils cannot be explained upon the assumption that the former was formed from organic remains, while the latter were pro- duced from inorganic matter, or vice versa. If, however, crude petroleum is organic in origin, it may have been formed either from vegetable or from animal remains. The follow- ing discussion is based upon the assumption that the above- mentioned oils were derived from an organic source. It has been suggested that the differences between these oils may be accounted for by assigning a vegetable origin to the Pennsylvania oil, and an animal origin to the others. P. Am. Phil. Soc., 1903. Mabery 1 states that "It would seem that the small propor- tion of these bodies 2 in the Pennsylvania oil, as compared with the larger proportions in the limestone oils and California oils should be strong evidence in favor of a different origin, that the Pennsylvania oil came from organic vegetable remains, which should permit of the small amounts of sulphur and nitrogen compounds from this class of oils." Newberry, Peckham, Orton, and other geologists also favor the view that the Pennsylvania oil is of vegetable origin, and is derived from the organic matter of the bituminous shales of the Devo- nian period. The facts which have led to the association of this oil with a vegetable source are, first, that the oil is of a different charac- ter from the limestone oils of Ohio, and those of Texas and California; second, that the Pennsylvania petroleum is found in strata that bear but few fossils; third, the belief that the Chemung and immediately overlying formations are barren of animal organic remains; fourth, the existence of large quan- tities of microscopic fossils, whose origin many believe to be vegetable, in the black shales of the Lower and Middle Devo- nian periods, to which formations many investigators are in- clined to refer the origin of the Pennsylvania oil. It is generally recognized that the Pennsylvania oil differs markedly from the Ohio, Texas, and California oils. Inves- tigation has shown that the former contains a much larger proportion of the paraffin hydrocarbons, and a much smaller percentage of benzene and unsaturated hydrocarbons and sulphur and nitrogenous bodies, than the latter oils. It is further generally admitted that the Pennsylvania oil was not formed in situ. These two facts have aided strongly in as- signing a vegetable origin to this oil. To what strata, then, should the source of the oil be referred? The great coal forma- tions of Pennsylvania, lying above the Chemung, seem, at a first glance, to offer a solution of this problem. It is a notable fact, however, that these formations have not, up to the present time, been connected, either chemically or geologically, with the Pennsylvania oil. The possibility exists, however, 1 P. Am. Phil. Soc.. 1903. 2 Reference is made to the sulphur, nitrogen, and oxygen compounds in petroleum. that it was formed from vegetable remains in the Carbonifer- ous formations above, and that then, by downward diffusion, it reached its present position in the Chemung. This view rests upon the physical fact that a liquid diffuses by the force of -capillarity in all directions, downward as well as upward. Little attention has been given to this possibility, but it seems to deserve a very careful study. Owing, however, to the uni- versal association of water under hydrostatic pressure with natural oil and gas, the migration of the latter is generally upward. This fact is attested by the accumulation of oil in anticlinal folds when water is present, and by the existence of the remarkable gushing oil wells. That the Pennsylvania oil, if not formed in situ, ascended to its present location seems, therefore, more probable. In what strata below the Chemung, then, was the oil origi- nally produced? It has been previously mentioned that a number of investigators refer the source of the oil to the black shales of the Lower and Middle Devonian periods. The or- ganic matter of these shales is composed largely of micro- scopic sporangites, which suggest the existence, according to Orton, of masses of floating vegetation, or Sargasso seas. According to this view, the origin of the Pennsylvania oil is vegetable in character, and its primitive abode was in the shales of the Devonian age lying below the Chemung sand- stone, to which it ascended under the influence of natural agencies. Another origin, animal in character, may be as- signed to this oil. This view is that the oil was formed in the fossil-bearing strata of the Chemung age, and that it diffused to the sandstone reservoirs in which it is now found, and that during such a diffusion its original character was changed. Prof. C. K. Swartz, of the Johns Hopkins Uni- versity, who has made a critical study of the Chemung strata in Maryland, informs us that fossil remains exist in considera- ble abundance in the strata of this age in Maryland and ad- joining areas. In Pennsylvania, the corresponding strata have been found to bear many fossils. It is possible that the oil formed in these strata, and then diffused to the strata in which it now exists and which are barren of fossil remains. The evidence accumulated in this investigation seems to show that it is not necessary to assign a vegetable origin to the Pennsylvania oil to explain the differences between it and the oils of Ohio and California. It is clear from the re- sults of this and other investigations that, when such oils as those of Ohio and California and Texas, which seem to be of animal origin, are allowed to diffuse through such porous media as fuller's earth, they yield oils very similar to those of Pennsylvania. By assuming, therefore, that the Pennsylva- nia oil migrated from some primitive source, in which it may have been formed from animal remains, through shales, lime- stones, and sandstones, its peculiar character can be under- stood. Whatever the original home of the oil, it seems probable that it migrated to its present location from some place be- low. It is with the changes occurring in crude petroleum as a result of such a migration through porous strata that the present investigation is primarily concerned. In 1897, Dr. David T. Day, 1 from his own observations, and those of Dr. John N. MacGonigle, proposed the view that the Pennsylvania oil, at some past time, possessed properties very similar to those of the Ohio oil, but that in its migration to its present abode from regions below, its character was changed to its present condition. Guided by this view, he conducted, in the laboratories of the United States Geological Survey, an investigation into the changes occurring in crude petroleum when allowed to diffuse through porous media, such as fuller's earth. He demonstrated clearly that an oil resembling the light Pennsyl- vania oil could be readily produced in the laboratory from the heavier crude Ohio oil. Glass tubes were packed firmly with the dry earth, through which the crude oil diffused by its own force of capillarity. From the earth of the upper sections of the tubes, very light, in some cases colorless, oils were liberated by treatment with water; from the earth of the lower sections of the tubes, much darker and heavier oils were obtained. i P. Am. Phil. Soc., 1897. 10 It will be observed that the fractionation is effected entirely by capillarity; oils with different surface tensions rise with different velocities through the capillary openings, such as the fine interstices and minute pores of the fuller's earth. A separation of the various constituents making up the com- plex of any one oil is thus produced. The view once held that this phenomenon is chemical was clearly disproved by Kngler and Albrecht 1 in 1901, and later by other investiga- tors. Any medium, therefore, sufficiently fine-grained and porous to afford capillary spaces causes a separation of the con- stituents of any mixture, provided they possess different surface tensions. The compact sandstones, shales, and lime- stones that recur in many cycles throughout the earth's crust present an excellent medium for the separation of the con- stituents of such a complex mixture as petroleum. The force of capillarity, assisted by the hydrostatic pressure of the water occurring in the interior of the earth, acting over vast periods of time, is, it seems safe to state, sufficiently powerful to transport the oil from the lower regions to those above. That the condition, therefore, to cause such a migra- tion, with the consequent fractionation of the original oil, are abundantly present, appears extremely probable. Let us examine, now, the conduct of the constituents of petroleum subjected to such a fractionation. The members composing the natural oil may be grouped under the following general heads: paraffin, aromatic, unsaturated hydrocarbons, sulphur, nitrogen, and oxygen compounds. The behavior of the paraffin and unsaturated hydrocarbons will be considered first. Dr. David T. Day early observed that the unsaturated hydrocarbons are less diffusible than the paraffin hydrocarbons. Later, Gilpin and Cram clearly demonstrated that when petro- leum is allowed to diffuse through tubes packed with fuller's earth, the unsaturated hydrocarbons collect in the earth of lower sections of the tubes, while the paraffins tend to accumu- late in the lightest fraction at the top of the tube. In the 1 Z. angew. Chem., 1901, 889. II present investigation, these results have been fully confirmed. On pages 50 to 52 are given the bromine absorption values, and the percentages by volume absorbed by concentrated sulphuric acid of the various oils obtained from definite sec- tions of a tube. These figures indicate conclusively that the amount of unsaturated hydrocarbons in the oils from the lower sections of the tube is much greater than the amount of these hydrocarbons in the lightest fractions at the top of the tube. Furthermore, the bromine absorption values for the oils of similar fractions of the first, second, and third frac- tionation, given on page 51, show that in the progress of the fractionation more and more of the unsaturated hydrocar- bons are removed. Herr, 1 in Russia, has likewise observed that these hydrocarbons are less diffusible than the paraffins. An interesting confirmation in nature of these experiments has been recently presented by Clifford Richardson and K. G. MacKenzie. 2 They found that a colorless natural naphtha from the Province of Santa Clara, Cuba, contained practically no unsaturated hydrocarbons, but was almost entirely a mix- ture of naphthenes and paraffins. Concentrated sulphuric acid absorbed but 0.76 per cent, by volume, while fuming sulphuric acid absorbed only 1.8 per cent. With the naphtha were obtained water and an emulsion of water, oil and clay. These investigators are of the opinion that the naphtha was "undoubtedly formed by the upward filtration of heavy petroleum through the clay stratum, similar to the fuller's earth filtrations of Gilpin and Cram, and the light naphtha in the upper part of the stratum was afterwards partly libera- ted by saline waters, the oil remaining in the clay forming with water the emulsion." 2 A comparison of the proportions of unsaturated hydrocar- bons in Ohio and Pennsylvania oils shows that ttfe latter con- tains a much smaller percentage of these hydrocarbons. By assuming that the Pennsylvania oil diffused upward through such porous media as shales and limestones to its present location in the sandstones, it is possible to account for the 1 Petroleum August, 1909. 2 Am. J. Sci., May, 1910. 12 smaller amounts of the olefins in it on the basis of the ex- perimental work described above. In its passage through the capillary interstices of the clays, limestone and sandstones, a fractionation, resulting in the removal of the unsaturated hydrocarbons, probably occurred. It is reasonable to con- clude, therefore, that the variation in the content of unsatura- ted hydrocarbons between the Ohio, Texas, and California oils, on the one hand, and the Pennsylvania oil on the other, can be probably accounted for by assuming that the latter was subjected to capillary diffusion at some time in its career. That the light-colored naphthas occurring in various parts of the world were originally darker and heavier oils, and that their primitive character was changed by diffusion through media possessing the power of fractionation, seems very probable. The behavior of the aromatic hydrocarbons, in particular benzene, in passing through fuller's earth, constitutes one of the subjects of this investigation. The results of this study, given in detail on pages 18 to 30, indicate clearly that ben- zene, like the olefins, tends to collect in the lower sections of a tube of fuller's earth through which the benzene, in solu- tion, is allowed to diffuse. That the aromatic hydrocarbons in the natural oil behave in a similar manner has not yet been decided. The proportion of these hydrocarbons in the Illi- nois oil investigated was too small to enable us to determine accurately their amounts in the various fractions obtained by the capillary diffusion of the crude oil. The ordinary methods, such as nitration with a mixture of nitric acid and sul- phuric acids, and sulphonation, employed for the quantitative determination of the aromatic hydrocarbons, could not be used in this work, owing to the fact that these reagents readily affect the unsaturated hydrocarbons as well. A study of the conduct of the aromatic hydrocarbons in the natural oil con- taining large amounts of them will be undertaken in the near future. It is probable, however, that the benzene and homol- ogous compounds in crude petroleum behave like the unsatura- ted hydrocarbons. The presence of larger amounts of aromatic hydrocarbons First frac- tionation. Third frac- tionation. O.04 0.003 0.05 0.09 o. 16 13 in the Ohio than in the Pennsylvania petroleum, and still larger amounts in the California and Texas oils, seems to afford further evidence in favor of the view that the Pennsyl- vania oil has undergone much greater diffusion, and conse- quently greater fractionation, than any of the other oils. The conduct of the sulphur compounds in petroleum in the process of diffusion is similar to that of the unsaturated hy- drocarbons. On page 52, the percentages of sulphur present in the oils from different parts of the tube and different stages of fractionation are tabulated. One series of figures will be given to show the behavior of the sulphur compounds: Per cent, of sulphur. Firsi Lot 6. Fraction A B D E It is clear from these figures that the sulphur compounds, like the unsaturated hydrocarbons, tend to collect in the lower sections of a layer of fuller's earth through which petro- leum is allowed to diffuse. In 1902, Clifford Richardson and E. C. Wallace, 1 in an in- vestigation on the occurrence of free sulphur in Beaumont petroleum, passed this oil upward through the kaolin filter described by Dr. D. T. Day at the Petroleum Congress in Paris, in 1900, and obtained a distinct fractionation. The percentages of sulphur in the crude oil, and the oils obtained by this fractionation were determined. The results are given in the following table: Per cent. Sp. gr. 25/25. sulphur. Crude oil o . 9140 i . 75 i st fraction 0.8775 0.70 2nd fraction 0.8986 0.91 3rd fraction o . 9038 i . 04 It seems reasonable to assume from these results that the variations in the sulphur content between the Pennsylvania 1 J. Soc. Chem. Ind., March, 1902. and Ohio oils may be satisfactorily explained by the view that the former oil, as previously stated, diffused from other regions to its present location, and in its migration a large part of its original content of sulphur was removed. Further work upon this point will be undertaken in this laboratory. No careful study of the behavior of the nitrogen and oxy- gen compounds in petroleum diffusing through a porous medium has as yet been undertaken. A careful investigation of this matter will be pursued in this laboratory later on. It is probable that such an investigation will show that the nitrogen compounds conduct themselves like the sulphur and unsaturated compounds. The Object of this Investigation. The present investigation was undertaken for the imme- diate purpose of studying the changes occurring in the crude Illinois oil when allowed to diffuse through fuller's earth. The more distant, but more fundamental, object was to gain further insight into the causes of the variations between the various oils of this country. EXPERIMENTAL. Preliminary Experiments. The Relative Amounts of Oil Lost in Heated and Unheated Fuller's Earth. Before the actual investigation of the Illi- nois oil was undertaken, experiments were made to deter- mine the relative amounts of oil lost in heated and unheated fuller's earth. 1 In the work of Gilpin and Cram, the earth was always heated until geysers ceased to form, and then al- lowed to cool for several hours. The purpose of heating the earth was to obtain larger yields of oil, but towards the close of their investigation it became apparent that the amount of oil lost in unheated fuller's earth was not as large as they had supposed it to be. Since much time and labor is consumed in the process of heating and then cooling the earth, it seemed advisable to settle this point at the outset. Apparatus. The apparatus employed for this investiga- 1 The fuller's earth employed in these investigations was generously supplied by the Atlantic Refining Company of Philadelphia. V tion was essentially the same as that used by Gilpin and Cram. A, A, A, A (Fig. I) are tin reservoirs made to hold somewhat more than a liter. The tin tubes B, B, B, B, 5.5 feet in length, and 1.25 inches in diameter, rest upon narrow tin supports placed upon the bottom of the reservoirs, and are connected with the branched glass tube F by suction tub- ing fitted with pinchcocks at E, E, E, E. The tube F is con- nected with the large tank C, which serves to maintain fairly constant pressures; C is in turn joined by the glass tube D to a manometer, and the latter connected with the Chapman pump. Any number of these tubes may be set up in series under the same diminished pressure. F After the tubes are closed at their lower ends with grooved corks covered with muslin to prevent the earth from sifting out, they are packed to the desired firmness with the fuller's earth. Bach tube is then placed in its own reservoir, con- taining the oil to be fractionated. When they are connected to the branched tube F, the pressure in the system of tubes is reduced by the suction pump. The oil rises at first rapidly, then its diffusion gradually diminishes in power. When the reservoirs are almost exhausted, the tubes are disconnected i6 and clamped, with the bottom ends up, above shorter tubes of the same diameter, into which the oil-laden earth is allowed to slide. These shorter tubes are made of two curved pieces, joined at the bottom by a cap, and held together at the top by a ring. The cylinders are opened by slipping off the ring and cap and removing one of the curved pieces, and the earth divided into the desired sections. When water is added in portions to the earth and the two mixed thoroughly, the oil is displaced and is drawn off in separate portions. Six tubes packed with heated fuller's earth were set up al- ternately with six tubes filled with the unheated earth. Each tube was placed in its own reservoir containing 950 cc. of crude oil. The oil was allowed to diffuse upward through the tubes under diminished pressure. Sixteen hours elapsed before the oil in the reservoirs was exhausted. Since the tubes did not rest directly upon the bottom of the reservoirs, a small amount of oil remained, the volume of which was subtracted from the volume originally supplied. The earth from each tube was shaken into a bucket, and the oil recov- ered by displacement with water, as described above. The results of these experiments are given in the following table : Table I. Heated Fuller's Earth. Weight of fuller's earth. Oil absorbed by earth. Oil recovered. Oil lost. Per cent, oil lost Tubes. Grains. cc. cc. cc. cc. I 1005 850 450 390 4 6 3 1000 792 460 332 41 5 1035 850 500 350 41 7 1070 86 5 450 4*5 48 9 1035 813 430 383 47 ii 1045 88 5 530 355 4i Total, 5055 2830 2225 44 Unheated Fuller's Earth. 2 1075 917 585 332 36 4 1095 853 562 291 34 6 1065 840 500 340 42 8 1045 814 435 379 46 10 1035 873 5io 363 4i 12 1055 850 485 365 4i Total, 5147 3077 2070 40 17 The petroleum employed in the above experiments was a dark, green oil from Venango County, Pennsylvania, possess- ing a specific gravity of 0.810. Since the Illinois oil, which was used in the fractionation proper, described later, differs materially from the Pennsyl- vania petroleum, further experiments were undertaken to determine the relative amounts of this oil retained by heated and unheated earth. Ten tubes, five of which were packed as uniformly as possible with fuller's earth that had been heated until geysers ceased to form, and the other five with unheated earth, were placed in reservoirs, each containing 950 cc. of Illinois oil, specific gravity 0.8375. When the oil was entirely absorbed, the tubes were taken down and the oil-laden earth shaken into two breakable cylinders and divided into the following sec- tions: A constitutes the section, 10 cm. in length, measured downward from the level to which the oil had ascended; B, the next 15 cm. ; C, 20 cm. ; D, 30 cm. ; E, 35 cm. ; the remainder of the earth to the bottom of the tube, designated as F, was entirely discarded. The earth was then treated with separate portions of water. The oils displaced by the successive additions of water were collected separately and are designated in the table below as Aj, A 2 , B lt B 2 , and so on; A^ is the oil first displaced, A 2 the oil next expelled by further additions of water. The volumes and specific gravities of the recovered oils were determined. The results are given in the following table: Table II. Heated fuller's earth. Unheated fuller's earth. Frac. Spec. grav. Vol., cc. Spec. grav. Vol., cc. A, 0.8287 100 0.8320 72 A, .... 0.8352 22 B, 0.8390 157 o . 8405 184 B 2 0.8485 35 0.8451 124 Ci 0.8441 280 0.8443 270 c, 0.8507 67 o - 8495 H7 *>, o . 8450 393 0.8483 368 D, o . 8490 132 0.8517 210 E t 0-8537 339 0.8500 360 E, 0.8564 174 0.8569 185 1701 1942 i8 These results indicate that unheated fuller's earth retains no more oil than the heated earth. Although, in these ex- periments, the percentage of oil lost in the unheated is smaller than that lost in the heated earth, Gilpin and Cram, employ- ing heated earth, recovered, in one test, 5,951 cc. from 9,070 cc., and, in another, 5,415 cc. from 8,915 cc., the amount of oil lost in the earth in the first test corresponding to 34 per cent., in the second to 39 per cent. It is clear, therefore, that there is no sufficient, if any, compensation for the time and labor spent in heating the earth. In the investigations that followed, therefore, the unheated fuller's earth was always used. The Diffusion of Benzene in Solution through Fuller's Earth. In order to deal more intelligently with the fractionation of the crude Illinois petroleum, it seemed advisable to study the behavior of the individual aromatic hydrocarbons, espe- cially benzene, both alone and mixed with paraffin hydrocar- bons, when allowed to diffuse upward through fuller's earth. Gilpin and Cram established the fact that the paraffin hydro- carbons tend to collect in the lightest fractions at the top of the tube. Their method consisted in distilling by heat six samples of oils of different specific gravities, each 300 cc. in volume, and collecting ten fractions between definite inter- vals. Five of these samples consisted of oil partly fractiona- ted by fuller's earth, and the other of the crude oil. The specific gravity and \iscosity of each fraction were deter- mined; then to 30 cc., or to all there was where the amount was less than 30 cc., an equal volume of concentrated sul- phuric acid (specific gravity 1.84) was added, and the two shaken in a machine for half an hour or longer. The volume of the oil unaffected by the acid was measured, and, by sub- traction, the volume of oil absorbed was calculated. This latter volume represents only approximately the percentage of unsaturated hydrocarbons present in the oil, because sul- phuric acid of this strength readily dissolves benzene when the two are thoroughly shaken. In this investigation various solutions of benzene and a refined paraffin oil, boiling between 160 and 240, and only 19 slightly attacked by sulphuric acid, were made up and allowed to rise in tubes packed with unheated fuller's earth. The pressure in the system was reduced very little, because the liquid, under a greatly diminished pressure, rose too rapidly. About 24 hours elapsed before the oil in the reservoirs was exhausted. The earth in each tube was shaken out and divided into six sections. Beginning at the uppermost point to which the oil had ascended, grade A consisted of the first 8 cm. ; grade B of the next 8 cm. ; grade C of 18 cm. ; grade D of 30 cm. ; grade E of 35 cm. ; and, finally, grade F of the remainder of the earth, depending on the height to which the oil had ascended. This division is the same as that used by Gilpin and Cram. The oil in the earth was displaced by water and drawn off. The specific gravity of each fraction was determined by means of the Mohr-Westphal balance at exactly 20. The fourth decimal is not to be considered as strictly accurate, but gives a closer approximation to the truth than if it were entirely discarded. The viscosity was determined by means of the viscosom- eter described by Ostwald and Luther and modified by Jones and Veazey. 1 The time taken for measured volumes of the oils to drain from the small bulb, whose capacity was 4.5 cc., was compared with the time required for a similar amount of water to run through. These values were substituted in the equation TS " y T S ' where y = coefficient of viscosity of water. For this, 0.01002, the value obtained by Thorpe and Rodger, 2 was used. t = time of flow of liquid under examination. S = specific gravity, measured at 20, of liquid under ex- amination. T Q = time of flow of water. S = specific gravity of water. Since the balance was calibrated for water, at 20, the value for S is unity. y = coefficient of viscosity of oil under examination. 1 Z. physik. Chem., 61, 351. 2 Phil. Trans., A, 185, 397 (1894). 20 The amount of benzene present in each fraction was deter- mined by shaking the oil with an excess of ordinary concen- trated sulphuric acid (specific gravity 1.84) for periods of time varying from 30 to 60 minutes, until there was no further diminution in the volume of the oil. The following experiments demonstrate the power of this acid to dissolve benzene, forming benzenesulphonic acid: (1) Twenty-five cc. of benzene were shaken vigorously in a machine with 25 cc. of concentrated sulphuric acid (specific gravity 1.84) for 30 minutes. Amount of benzene dissolved, 7 cc., or 28 per cent. (2) Twenty-five cc. were shaken for 30 minutes with 50 cc. of acid. Amount of benzene dissolved, 18 cc., or 72 per cent. (3) Twenty-five cc. were shaken for 30 minutes with 75 cc. of acid. Amount of benzene dissolved, 25 cc., or 100 per cent. The reagents usually employed for removing benzene are a mixture of fuming nitric and concentrated sulphuric acids. The work of Worstall, 1 Francis and Young, 2 and others, shows that such a mixture readily attacks the paraffin hydrocarbons, especially at higher temperatures, forming nitro derivatives, and also oxidizing them to a considerable extent. Further- more, in working with this mixture the oil must be kept at a low temperature to prevent a violent reaction which results usually in the decomposition of the oil. In this work, there- fore, in order to avoid the danger of attacking the paraffin hydrocarbons, and or the sake of convenience, concentrated sulphuric acid was used. It seems advisable, at this point, to call attention to the fact that the power of ordinary concentrated sulphuric acid to remove benzene and homologous hydrocarbons has been generally overlooked. In order to determine the percent- ages of these hydrocarbons, it is customary to shake the oils to be analyzed with concentrated sulphuric acid, and then to nitrate the unaffected oil. It is assumed that the acid re- moves such substances as the unsaturated hydrocarbons, and does not attack the aromatic hydrocarbons. Thus, P. 1 Amer. Chem. J., 20, 202; 21, 210. 2 J. Chem. Soc.. 1898, 928. 21 Poni, 1 in determining the presence and percentage of aromatic hydrocarbons in Roumanian petroleum, collected fractions between 35 and 70, distilled under diminished pressure. These were purified by shaking with sulphuric acid, and each nitrated with a mixture of i part of nitric acid (specific grav- ity 1.52) and 2 parts sulphuric acid (specific gravity 1.8). The recovered oils were assumed to be paraffins and naphthenes, while the proportion of benzene and unsaturated hydrocar- bons was calculated from the nitro products obtained. It is obvious from the results obtained in the present work that some of the benzene was removed in the process of purifying the fractions. The amount dissolved would depend upon the vigor of the shaking and its duration, as well as on the strength of the sulphuric acid. It is highly probable, there- fore, that his percentage of benzene is too low. In the study of the mixture of benzene and paraffin hydro- carbons, twenty-five cc. of each fraction, or the whole frac- tion when it was less than 25 cc., were shaken vigorously with three times their volume of concentrated sulphuric acid for 30 minutes. The amount unabsorbed was measured over the acid in a burette, after sufficient time was allowed for most of the oil that was mechanically held in suspension to rise. The oil was then reshaken with a little more acid 15 minutes longer, and the volume again read. In cases where the benzene was present only in small quantities one shaking was sufficient, in other cases it was repeated a second time. The paraffin oil employed, specific gravity 0.797, was shaken several times with fresh portions of concentrated sulphuric acid until the acid was no longer colored, and only a slight diminution in volume occurred when a small sample of the oil was thoroughly shaken in a machine for some time with the acid. The oil was then washed with water and sodium hydroxide and dried over calcium chloride. The specific gravity decreased to 0.792. When this oil was mixed with benzene in various propor- tions, and allowed to diffuse upward through fuller's earth, the following results, arranged in series, were obtained: 1 Ann. Sci. Univ. Jassy, 1907, 192-202 (abstracted in J. Chem. Soc., 92, II, 883 (1907)). Volume of oil, cc. Specific gravity. Viscosity. II 0.789 .... I? 60 0.792 0.7912 0.0154 100 0.7915 0.0140 150 0.7913 0.0134 139 0.7915 0:0134 22 Table III. Series i. Oil alone. Specific gravity 0.792. Level of oil, 28 cm. Per cent. Grade. oil, cc. gravity. Viscosity. benzene. 2 A B C D E F 477 1 Orig. vol., 778 Series 2. 90 per cent, oil (o.792)-io per cent, benzene (0.8775). Specific gravity, 0.7983. Level of oil, 22 cm. Volume of Specific Per cent. Grade. oil, cc. gravity. Viscosity. benzene. A ii 0.787 .... 10. o B 16 0.7923 13.3 C 56 0.7935 0.0131 ii. 6 D 109 0.7943 0.0123 14.8 E 145 0.7957 0.0120 14.4 F 245 0.7955 0.0116 14.8 582 Orig. vol., 872 The results that are tabulated in the various series are ex- pressed diagrammatically in the following curves. The ordinates represent the different grades of oil, and the ab- scissas, the percentages of benzene and the specific gravities. The final curve represents in toto the results of the experi- mental work upon the diffusion of benzene in solution through fuller's earth. The ordinates of this curve represent the per- centages of benzene, and the abscissas, the various mixtures of benzene and oil that were allowed to diffuse through the earth. 1 The original volumes of solution vary with each series, owing to the fact that more or less always remained behind in the reservoir below the level of the tin sup- port. In Series 1, 2, 3, and 4, 950 cc. were supplied to each reservoir; in the rest of the series, each reservoir contained originally 1 ,000 cc. 2 In this series the percentages of benzene are not given, because the paraffin oil alone was used. IO 20 Per cent. Benzene. OO O.JQOO O.800O Specific Gravity. Fig. II. Series 2. Series 3. 80 per cent, oil (o.792)-2o per cent, benzene (0.8775) Specific gravity, 0.806. I^evel of oil, 25 cm. Grade. A B C D E F Orig. vol., Volume of Specific Per cent. oil, cc. gravity. Viscosity. benzene. 25 0.7948 0.0147 15-3 35 0.7981 0.0130 16.0 78 0.8017 O.OII7 22.4 126 0.80O5 0.0105 21.6 166 0.801 0.0107 22.4 146 0.798 O.OIIO 20.8 576 , 892 Series 4. 75 per cent, oil (o.792)-25 per cent, benzene (0.8775), Specific gravity, 0.810. Level of oil, 33 cm. Grade. Volume of oil, cc. Specific gravity. Viscosity. 1 Per cent, benzene. A 16 0.800 .... 22.0 B 35 0.803 0.0129 23-3 C 74 0.8077 O.OI26 24.0 D 128 0.805 O.OII4 24.0 E 152 0.8068 O.OIO2 26.O F 120 0.8065 0.0105 28.0 525 Orig. vol., 655 i. O IO 2O 30 O.JQOO 0.8OOO O.SlOO Per cent. Benzene. Specific Gravity. Fig. III. Series a. lo 20 30 0.7900 0.8000 0.8100 Per cent. Benzene. Specific Gravity. Fig. IV. Series 4. 1 The viscosities of Grades A and B in a few of the tables are not given, because, in these series, which were the first to be made, the decision to determine the viscosi- ties was reached only after the fractions had been treated with acid. Since A and B were small, all the oil was used up in this treatment. Series 5. 75 per cent, oil (0.794) *-2 5 P er cent - benzene (0.8775). Specific gravity, 0.8115. Level of oil, 24 cm. Volume of Specific Per cent. Grade. oil, cc. gravity. Viscosity. benzene. A 25 0.7942 0.0123 14.0 B 28 0.8048 0.0104 21 .2 C 70 O.8lO5 O.OO94 31.2 D 140 O.SlOO 0.0094 2 7 .6 E 172 O.SlOO 0.0094 32.0 F 144 0.8093 0.0095 2 7 .6 579 Orig. vol., 875 Series 6. 75 per cent, oil (o.792)-25 per cent, benzene (0.8775). Specific gravity, o 8083. Level of oil, 27 cm. Volume of Specific Per cent. Grade. oil, cc. gravity. Viscosity. benzene. A 22 0-7995 0.0106 17-5 B 32 0.8055 0.0099 24.4 C 82 0.8052 o.oioo 24.0 D 155 0.8085 0.0093 28.8 E 190 0.8085 0.0093 31.2 F 93 o . 8063 o . 0096 28.8 574 Orig. vo 1-, 923 \ | ] } 10 20 30 0.7QOO 0.8200 O O.8000 O.SlOO Per cent. Benzene. Specific Gravity. Fig. V. Series 5. IO 2O 30 O.JQOO 0.820O 0.80OO O.SlOO Per cent. Benzene. Specific Gravity. Fig. VI. Series 6. 1 In Series 5, 8, 9 and 10 the specific gravity of the refined oil is 0.794. Since the quantity of oil of specific gravity 0.792 was not sufficient for all the series, a sec- ond quantity was prepared which had the specific gravity 0.794. This oil was used in the above-mentioned series. 26 Series 7. 59.5 per cent, oil (o.792)~4o.5 per cent, benzene (0.8775). Specific gravity, 0.8223. Level of oil, 9 cm. Grade. Volume of Specific oil, cc. gravity. Viscosity. Per cent, benzerte. A 9 1 .... .... B 15 o . 8069 14.0 C 48 0.816 0.0103 22.4 D 96 0.8182 o . 0086 31.2 E 160 0.820 0.0082 31-6 F 255 0.8185 o . 0083 29.6 583 Orig. vol., 922 Series 8. 50 per cent, oil (o.794)-5o per cent, benzene (0.8775). Specific gravity, 0.8295. Level of oil, 17 cm. Volume of Specific Per cent. Grade. oil, cc. gravity. Viscosity. benzene. A 22 0.8l22 .... 24-5 B 32 0.819 28.4 C 78 0.8287 O.0077 44.8 D III 0.8275 0.0077 47 -6 E 155 0.827 0.0077 39-2 F I 9 2 0.8256 0.0079 36.4 590 Orig. vol., 960 A V \ B \ \ x^ \ C \ ^ { \ ) D | \ ) E 1 F 30 0.7900 0.8100 o 0.8000 0.8200 30 40 0.7000 0.8100 0.8300 0.8000. 0.8200 &"" Per cent. Benzene. Specific Gravity. Per cent. Benzene. Specific Gravity. Fig. VII. Series 7. Fig. VIII. Series 8. 1 In Series 7 the volume of Grade A recovered was so small that no measure- ments could be made. Series 9. 50 per cent, oil (o. 794^50 per cent, benzene (0.8775), Specific gravity, 0.8315. Level of oil, 18 cm. Grade. Volume of oil. cc. Specific gravity. Viscosity. Per cent, benzene. A 18 0.816 O.OO9I 26.0 B 24 0.8210 o . 0085 34-5 C 76 0.8275 0.0078 47.6 D 136 0.8283 0.0077 50.0 E 174 0.8293 O.OO76 49-2 F 144 0.8277 0.0078 40.0 572 Orig. vol., 923 Series 10. 50 per cent. oil (o.794)-5o per cent, benzene (0.8775). Specific gravity, 0.8295. Level of oil, 1 6 cm. Volume of Specific Per cent. Grade. oil, cc. gravity. Viscosity. benzene. A 31 0.8135 0.0097 31.6 B 45 0.8251 o . 008 i 43-6 C 85 0.8290 0.0076 46.4 D 140 0.8280 0.0077 47.6 E 175 0.8285 0.0076 49.6 F 137 0.8272 O.O076 50.0 613 Orig. ^ / rol., 972 \ } \ 1) 10 20 30 40 0.7900 0.8100 0.8300 o 10 20 30 40 0.7900 0.8100 0.8300 0.8000 0.8200 0.8000 0.8200 Per cent. Benzene. Specific Gravity. Per cent. Benzene. Specific Gravity. Fig. IX. Series 9. Fig. X. Series 10. 28 Series u. 75 per cent, crude oil (o.8io)-25 per cent, benzene (0.8775). Specific gravity, 0.8312. Level of oil, 18 cm. Grade. A B C D E F Volume of Specific oil, cc. gravity. Viscosity. 12 0.8255 0.0445 22 0.8268 0.0423 52 0.8280 o . 0300 7 6 0.8290 0.0298 140 o . 8300 0.0263 186 O.8320 0.0276 Per cent. benzene. 1 Orig. vol., 488 890 Series 12. Benzene alone (0.8775). Level of oil, 33 cm. Grade. A B C D E F Volume of Specific oil, cc gravity. 16 0.8765 15 0.877 68 0.878 128 0.8778 157 0.8775 89 0.8771 Viscosity. O . OO66 O . 0066 O . OO66 O . 0066 Per cent, benzene. 473 Orig. vol., An examination of these figures shows conclusively that benzene tends to collect in the lower portions of the tube. The specific gravities and viscosities confirm the results ob- tained by determining the percentages of benzene present by removing the benzene with concentrated sulphuric acid. The specific gravities of Grades F to C run very close together, and are all much greater than those of Grades A and B. Since benzene possesses a high specific gravity (in this work the specimen had a specific gravity of 0.8775), tne larger value for the lower grades indicates the presence of larger amounts of benzene. The specific gravity of the paraffin oil was only 0.792, showing that the higher specific gravities were due to larger percentages of benzene. Further, since the viscosity 1 The percentages of benzene in Series 11, in which crude oil was employed, are not recorded, because, owing to the formation of heavy black emulsions, the loss in volume could not be determined with any degree of accuracy. 3 10% Benzene QO% Oil 20% Benzene 80% Oil Mixtures Fractionated. Fig. XI. 25% Benzene 75% Oil 50% Benzene 50% Oil 30 of the benzene used was 0.0066, and that of the paraffin oil about 0.0150, the viscosities of those fractions containing higher percentages of benzene, we should expect, ought to be much smaller than those containing less benzene. The re- sults show that the viscosities of the Grades F to C are much smaller than those of A and B. It will be observed that the maximum in specific gravity is reached, not at F, as might be expected in the f ractionation of the crude oil, but between C and D. Between B and C there is a marked decrease. This sudden break is found also in the viscosities, and in the percentages of benzene. While the sharp breaks in the curves represent the marked change in the proportion of benzene and the height to which it rises in the tube, no satisfactory explanation has yet been ob- tained as to why it should occur at these points. This action will be studied more carefully later. In order to determine the degree of exactness of the per- centages of benzene obtained, known amounts of benzene were added to the oil until the specific gravity corresponded closely to that obtained by f ractionation. The amount of benzene thus added and the amount actually removed by the acid agree very closely, as the following results show: Benzene found in the Benzene in 25 cc. grades of Series 8. of mixture. Specific gravity. cc. Specific gravity. 7 3 0. 8143 Grade A 7 9 0. 8135 9 4 0. 8213 " ' B 10 9 O 8251 ii .1 O. 8274 M 77* 12 5 ,8272 ii 3 0. 8287 " E 12 4 .8287 ii 9 o. 8293 " C II .6 O .8290 The variations in the specific gravities of the mixtures and those of the grade A-F are due to the fact that in the latter series some f ractionation had taken place, and therefore the paraffin oils mixed with the benzene were not identical with those mixed with the benzene in the series of prepared mix- tures, as the paraffin oil used was not an individual substance but a mixture. The Fractionation of Crude Petroleum. The petroleum employed for the fractionation was an oil obtained from the E. E. Newlin farm, 2.5 miles west of Robin- son, Crawford County, Illinois. The specific gravity of the oil was 0.8375 at 20; its color was dark brown. The fractionation of the oil was effected by upward diffusion through tubes packed with fuller's earth. In order to shorten the time required for the oil to diffuse by capillarity to the upper parts of the tube, the fine interstices and pores of the earth were evacuated by applying diminished pressure at the top of the tube. By this aid, the time required for the oil to reach the top of a tube was reduced from several weeks to one or two days. The apparatus employed is the same as that described on page 14. The tin tubes were packed as uniformly as possible by in- troducing definite amounts of earth, and ramming solidly with rods tipped with rubber stoppers. The degree of com- pactness depended upon the kind of oil to be used. For the crude oil, about one and one-half feet of the tube was filled at a time, and the earth packed as firmly as possible; for the lighter oils, one foot of the tube was filled at a time; for the oils heavier than the crude, between two and three feet of the tube were filled at one time. The tubes were then placed individually in reservoirs con- taining 950 cc. of the crude oil, after which diminished pres- sure was applied at the top of the tubes. The oil rose rapidly at first, then diffused more and more slowly as the tops of the tubes were approached. When the oil in the reservoirs was completely exhausted, the tubes were disconnected from the blanched glass tube F (see Fig. I) and the oil-laden earth shaken into two breakable cylinders. For the various frac- tions, the following divisions of the earth were made: Frac- tion A constituted the first 10 cm., measured downward from the level to which the oil had ascended; fraction B, the next 15 cm.; C, 20 cm.; D, 30 cm.; E, 35 cm., and F, the remainder to the bottom of the tube. In the first fractionation up to 32 Lot 28, fraction F was dicarded; from Lot 28 to the end of the first fractionation, E and F were collected together. After thus dividing the earth, the various portions were placed in separate receptacles and treated with water. After each addition of water the two were thoroughly mixed. The earth, when the oil first appears, is granular; as more water is added, liberating more oil, the earth becomes muddy, and when as much oil as possible has been expelled by the water, the earth has the consistency of glue. The portions of oil liberated by successive additions of water were collected separately. As Gilpin and Cram 1 pointed out, the oil that is first expelled, if not very small in volume as compared with the oils succeeding, possesses a lower specific gravity than the oil liberated by further additions of water; the latter, in turn, is lighter than the next succeeding oil. The oil that is liberated last, therefore, possesses a higher specific gravity than any of the oils preceding it. Sometimes, how- ever, the specific gravity remains constant after the second or third extraction. This fractionation, by means of water, was combined with the fractionation effected by the fuller's earth. In the tables that follow, A l is the oil first liberated, A 2 the oil next liberated; in the lower fractions, i. e., C, D, E, three and sometimes four extractions were made before all the oil that could possibly be liberated by water was recov- ered. The specific gravity of the oils was determined by means of the Mohr-Westphal balance. As mentioned before, the fourth decimal is not to be considered as rigidly accurate, but it gives a closer approximation to the truth than if it were entirely discarded. The temperature at which the specific gravity was measured was exactly 20. 1 Amer. Chem. Jour., 40, 495 (1908). 33 Lot. No. of tubes. Table IV. The First Fractionation. a 3 Hours 1 1 8 req. 23 Spec. Frac. grav. A l 0.8250 A 2 0.8287 BI 0.8367 B 2 0.8392 C\ 0.8413 C 2 0.8460 C 3 0.8488 A 0.8470 D 2 0.8495 D 3 0.8514 A 0.8555 E t 0.8527 E 2 0.8540 E 3 0.8570 Lot. No. of tubes. Hours req. Spec. Frac. grav. A l 0.8295 A 2 0.8315 #1 0.8375 B 2 0.8413 C\ 0.8418 C 2 0.8442 C 3 0.8495 A 0.8449 A 0.8455 A 0.8490 E! 0.8500 0.8510 14 tubes i tube Vol.,2 cc. 312 90 485 250 828 228 126 1014 375 200 172 720 430 4OO 16 2 E 3 0.8567 Vol., cc. 170 100 327 250 505 223 74 495 328 260 545 295 170 17 8 tubes 45 3 tubes 1 1 o o Spec, grav. .8285 .8310 8370 8408 Vol., cc. 73 59 218 78 s g 0. 0. 0. o pec. ;rav. 8223 8270 8372 84.OO Vol., cc. 138 54 258 200 Spec, grav. 0.8233 o . 8405 Vol., cc. 50 130 o o 8440 8442 8430 .8464 .8500 272 136 313 150 112 0. 0. 0. 0. o. 0. 8442 8455 8480 8488 8500 8540 290 235 148 538 295 H5 0.8505 0-8535 0.8546 0.8619 120 65 235 30 8475 8509 .8540 285 135 118 o. 0. 0. 8537 8550 8580 380 245 170 0.8615 172 6 103 177 tubes 24 i tube 17 i tube 4 40 3 tubes Spec, grav. O. O. O. 0-8453 0.8419 o . 84^9 0.8465 0-8454 0.8500 0.8495 0.8513 O. Vol., cc. 130 358 92 425 138 I 3 640 I6 7 195 575 185 130 96- Spec. grav. 0.8320 0-8352 0.8405 0.8451 0.8443 0.8495 0.8483 0.8517 o . 8500 0.8569 tube Vol., cc. 22 184 124 270 147 368 2IO 360 185 17 3 tubes 40 i tube 150 i tube Spec. Vol., grav. cc. 0.8287 85 o . 8490 0.8485 0.8441 0.8507 o . 8450 o . 8490 0.8537 0.8564 !34 35 218 67 302 132 215 174 1 Chapman pump was run day and night. Manometer indicated pressures rang- ing from 30 to 80 mm. 2 In lots 1 to 5, 1000 cc. of crude oil were supplied to each tube. 3 Beginning with lot 6, 950 cc. of crude oil were supplied to each tube. 4 The pressure in the tubes was diminished intermittently. See page 17. 34 Lot. No. of tubes. 7 9 8 10 9 10 Hours 20 7 tubes 20 i tube 19 8 tubes 24 2 tubes req. 24 i ti ube 22 2 ti .ibes 408 tubes Spec. Vol., Spec. Vol., Spec. Vol., Spec. Vol., Frac. grav. cc. grav. cc. grav. CC. grav. cc. A l o .8325 66 .8175 45 .8364 88 o .8215 145 A 2 o .8356 30 8365 64 .8234 90 B l o .8395 164 .8333 no .8400 215 8330 397 B 2 o .8418 140 .8420 240 o .8350 155 g o . 8400 87 C l o .8408 475 O .8417 132 O 8445 368 o .8415 ** / 350 C 2 o .8468 123 .8500 22 .8467 225 o .6436 255 , . . . 8495 82 .8480 160* Z^ 1 o .8449 500 .8468 no .8465 460 .8485 507 > 2 o .8487 270 .8498 1 06 O .8478 260 8495 280 o .8500 260 8545 247 j O .8500 483 8533 228 .8490 450 .8548 313 2 .8524 3i8 o 8495 354 o .8550 275 E, .8521 233 .8580 375 Lot. 10 11 12 13 No. of tubes. 8 10 9 10 Hours req. 14 17 42 24 8 tubes 40 2 tubes Spec. Vol., Spec. Vol., Spec. Vol., Spec. Vol., Frac. grav. cc. grav. cc. grav. cc. grav. cc. A! O .8273 130 .8258 215 .8325 125 8323 122 A 2 o .8288 75 .8318 70 8345 87 -8352 9 6 t 8395 220 .8370 340 o .8430 235 0.8438 245 #2 .8418 1 60 .8480 1 80 .8467 120 .8470 180 C l o .8423 240 .8422 488 o .8470 2 7 8 o .8464 317 C 2 o .8440 195 8450 205 o .8487 288 .8505 235 C 8 o .8500 150 . . . . . . . . . . . D! o .8460 410 O .8465 565 o 8495 452 .8500 312 D 2 o .8475 2IO .8490 310 .8522 305 .8492 375 D 3 o .8500 348 0.8530 187 o .8518 150 E l o .8532 320 o .8510 297 .8505 475 .8505 450 E 2 o -8535 282 o .8520 405 8533 490 .8489 395 E 3 o .8550 215 o 8533 155 .8518 1 80 1 Several cubic centimeters of this fraction were mixed, accidently, with frac- tion E 3 . 35 Lot. No. of tubes. Hours req. Frac. 14 15 Spec. grav. Vol., cc. 26 3 tubes Spec. Vol., grav. cc. 26 3 tubes Spec. Vol., grav. cc. A 0.8355 i32 0.8381 6o 0.8305 73 B l 0.8470 236 0.8487 94 0.8452 143 C l 0.8565 0.8560 v-1 D 1 0.8523 D 2 0.8550 D 3 t 0.8540 E 2 0.8532 98 150 170 205 150 325 0.8430 i 10 0.8465 138 0.8480 57 0.8509 88 0.8475 212 0.8505 158 0.8517 104 0.8522 178 0.8467 184 0.8561 192 0.8502 152 0.8585 140 16 15 40 11 64 4 Spec, grav. 0.8370 0.8357 o . 8449 0.8445 0.8475 0.8509 0.8562 0.8540 0.8530 0.8575 0.8538 0.8562 0.8595 tubes tubes Vol., 200 108 490 226 635 235 90 825 495 150 775 620 205 Lot. No. of tubes. Hours req. 17 9 40 Spec. Vol., 18 8 24 5 482 64 1 Spec. tubes tubes tube Vol., 19 10 40 8 tubes 64 2 tubes Spec. Vol., 20 10 20 6 tubes 30 4 tubes Spec. Vol., Frac. grav. cc. grav. cc. grav. cc. grav. cc. A 0. 8258 225 o .8322 112 0.832O 146 .8281 236 B 0. 8432 452 8435 335 0.8438 385 .8413 518 c t o. 8480 450 8495 250 0.8480 300 .8450 350 0. 8488 168 o .8500 250 0.8472 315 8495 300 D\ 0. 8530 520 o 8530 320 0.8509 4 22 O .8508 325 D 2 0. 8550 350 .8540 350 0.8535 355 -8538 460 E l o. 8585 385 o 8547 90 2 0.8492 580 .8513 445 E 2 0. 8598 460 o 8526 640 0.8560 415 .8540 550 1 TI ,.-,:! J j.1 .*, .4 i short time, the reservoirs were nearly two-thirds exhausted. The pump was stopped, and the remainder of the oil allowed to diffuse during the night under normal pressure. 2 This irregularity, *. ., the liberation of oil with a specific gravity higher than those of the oils immediately following, is observed when an amount of water is added sufficient to replace a very small amount of oil for the first fraction. Lot. 21 No. of tubes 10 Hours 1 24 6, tubes req. 40 2 tubes 64 2 tubes Spec. Vol., Frac. grav. cc. A 0.8275 245 B 0.8410 615 C l 0.8452 520 C 2 0.8488 226 D 1 0.8512 533 A 0-8535 4*5 EI 0.8557 375 E 2 0.8625 282 Lot. 25 No. of tubes 9 Hours 2 48 . 8 tubes req. 72 1 tube Spec. Frac. grav. A B 0.8270 0.8425 0.8495 o . 8492 0.8509 2 0.8510 E l 0.8556 E 2 0.8570 D. Vol., cc. 225 410 75 2 250 320 480 335 395 29 Lot. No. of tubes 10 Hours 4 18 5 tubes req. 40 5 tubes Frac. A B , D, EF, EF, Vol., cc. 3327 22 10 40 6 tubes 4 tubes Vol. cc. Spec, grav. 0.8281 o . 8405 0.8459 0.8472 0.8505 0-8523 0.8615 0.8585 210 508 265 410 435 450 385 365 26 10 17 2 tubes 24 4 tubes 41 4 tubes Spec. Vol., grav. cc. 0.8284 315 0.8422 550 0.8473 520 0.8508 178 0.8515 600 0.8540 230 0.8559 490 0.8586 135 Spec, grav. 0.8262 3OO 0-8395 505 o . 8463 390 0.8488 270 0.8520 510 0.8543 290 0.8550 417 0.8559 645 30 15 207 41 6 632 Spec. grav. 0.8348 0.8468 o . 8490 0.8505 0.8485 0.8502 0.8520 0.8528 23 10 tubes 52 5 tubes 24 10 40 4 tubes 64 6 tubes Spec, grav. 0.8241 0-8395 o . 8448 o . 8470 0-8533 0.8541 0.8650 0.8624 27 Vol., cc. 330 615 420 305 400 465 305 350 Spec, grav. 0.8250 o . 8408 o . 8463 0.8505 0.8540 0.8540 0.8623 o . 8645 28 Vol., cc. 287 535 475 1 86 525 360 393 335 10 17 4 tubes 29 6 tubes 10 24 7 tubes 28 3 tubes Spec, grav. 0.8312 o . 8440 o . 8460 0.8478 o . 8482 o . 8500 O.8520 0.8565 31 Vol., cc. 230 470 400 232 435 420 465 335 Spec, grav. 0-8333 o . 8440 0.8458 o . 8500 0.8470 0.8498 0.8492 0.8505 Vol., cc. 240 410 415 177 387 400 6 9 o 3 600 32 tubes tubes tubes Vol., cc. 335 630 560 277 750 540 1125 880 5097 10 44 4 tubes 15 40 7 tubes 89 6 tubes 89 4 tubes 103 4 tubes Spec. Vol., Spec. Vol., grav. cc. grav. cc. 0.8292 245 0.8270 445 0.8439 576 0.8423 726 0.8495 465 0.8500 730 0.8523 205 0.8500 2 2O 0.8517 670 0-8545 750 0.8552 210 0.8543 540 0.8555 805 0.8580 870 0.8610 360 0.8598 910 3536 5191 1 Pressure in the tubes was diminished intermittently. 2 Some oil of this fraction was lost. 3 Beginning with lot 28, fractions E and F were collected together. 4 Pressure in tubes was diminished intermittently. 37 Lot. No. of 33 34 35 tubes. 10 10 9 Hours' 414 tubes 44 6 tubes 48 6 tubes req. 65 4 tubes 68 4 tubes 72 3 tubes 89 2 tubes Spec. Vol., Spec. Vol., Spec. Vol., Frac. grav. cc. grav. cc. grav. cc. A 0.8330 290 0-8355 320 0.8380 235 B t o . 8440 365 0.8475 525 0.8460 452 B, 0.8462 165 c t 0.8502 500 0.8508 470 0.8508 345 C 2 0.8540 160 0.8543 190 0.8525 245 Dl 0.8555 655 0-8575 530 0.8549 580 D, 0.8562 250 0.8585 325 0-8573 335 EF t 0.8575 735 0-8535 895 0-8557 645 EF, 0.8585 480 0.8555 405 0.8570 492 3600 3660 3329 Observations on the First Fractionation. Specific Gravity. The range of the specific gravity ex- tended from 0.8175, the value for Fraction A l of Lot 7, to 0.8650, the value for Fraction E of Lot 13. The value for the crude oil itself was 0.8375. The limits of the specific gravities of the individual lots averaged from 0.820 to 0.860. The specific gravity decreases gradually from E to B, but be- tween B and A , the decrease, in most of the lots, is much greater than between any two consecutive lower fractions. This marked change was also observed in the study of the diffusion of benzene in solution. A detailed investigation into the cause of this sudden divergence will be undertaken in the near future. Color. The colors of the fractions obtained extended from green to black. The lighter oils possessed a beautiful green fluorescent color, which shaded gradually to brown, and then to the deep black of the heavier oils. Odor. The unpleasant odor of the crude petroleum disap- peared almost entirely in the oils of Fraction A and B ; but the other fractions still possessed to a greater or less extent the odor of the natural oil. The Volume of Oil Retained by the Fuller's Earth. The amount of oil retained by the earth averaged about 55 per 38 cent, of the amount supplied. In the first fractionation of the crude Pennsylvania oil, specific gravity 0.810, Gilpin and Cram found that approximately 40 per cent, of the oil was retained by the earth. It is evident, therefore, that the amount of oil remaining in the earth depends chiefly upon the character of the oil. The Pennsylvania petroleum contains a much smaller percentage of unsaturated hydrocarbons, sulphur, and asphaltic substances than the Illinois oil em- ployed in this investigation. Since the fuller's earth, as will be shown later, readily removes these substances in the process of fractionation, the large percentage of Illinois oil retained by the earth is thus clearly explained. It is safe to conclude that if the heavy Texas or California oil were allowed to diffuse through fuller's earth, the amount of oil retained would ex- ceed the amounts of either of the above-mentioned oils lost in the earth. The Second Fractionation. The products obtained from the first fractionation were united according to the following arrangement : Specific gravity of the Specific gravity of Lot. oils united. mixture. 36 0.8250-0.8350 0.8293 37 0.8350-0.8400 0.8390 38 o . 8400-0 . 8450 o . 8433 39 0.8400-0.8450 0.8433 40 o . 8450-0 . 8500 o . 8490 41 42 43 44 0.8500-0.8600 0.8543 45 U It (I (I (I U 48 49 50 46 47 a . u (I (I U u (i u The oils thus combined were subjected to chilling and filtra- tion for the purpose of removing as much dissolved paraffin 39 as possible. The procedure was as follows : The oils were first chilled at temperatures ranging from o to 10 and then filtered through plaited filter papers. When the oil ceased to drip from the funnel, the residue upon the filter paper was placed in a larger filter press, and the remaining oil separated by pressure from the paraffin. The filter press was simple in construction. A piston, fitted closely in an iron cylinder, was gradually forced down upon the oil-laden paraffin, which rested upon a membrane of cotton duck fastened between perforated tin supports. The retained oil was forced through the membrane and was collected from the outlet below. The lighter oils deposited very little paraffin; from the heavier ones somewhat more paraffin was separated. Owing to the high viscosity of the heavier oils, the filtration proceeded very slowly. Since too much time was consumed in this process, the paraffin of some of the oils of Fraction E was not removed. A slight change in specific gravity occurred in the oils from which the paraffin was removed. The final specific gravities of the united oils were as fol- lows : Lot. Specific gravity. 36 0.8305 Paraffin removed. 37 0.8415 38 o . 8433 Paraffin not removed. 39 0.8455 Paraffin removed. 40 0.8515 41 0.8515 42 0.8515 43 0.8540 44 0.8543 Paraffin not removed. 45- 0.8543 46 0.8543 47 0.8543 48 0.8543 49 0.8557 Paraffin removed. 50 0.8557 When these oils were again allowed to diffuse upward through fuller's earth, the following fractionation was ob- tained : Table V. The Second Fractionation. Lot. 36 No. of tubes. 5 Hours 1 44 3 tubes req. 48 2 tubes 37 4 51 38 8 48 7 tubes 64 1 tube 39 8 29 4 tubes 45 3 tubes 64 1 tube Spec. Vol., Spec. Vol., ' Spec. Vol., Spec. Vol., Frac. grav. cc. grav. cc. grav. cc. grav. cc. A 0.8272 1 60 .8292 135 O .8331 180 0.8290 255 6, 0.8315 216 .8421 215 .8447 175 0.8432 355 B 2 0.8331 58 O 8455 210 0-8458 no Q 0-8334 350 .8467 295 .8490 305 o . 8492 455 c % 0-8355 85 .8505 175 0.8513 1 80 D, 0.8330 360 o .8468 340 .8492 4OO 0.8505 740 D 2 0-8339 320 8485- 152 .8509 295 0.8527 275 EF, 0-8347 72O .8480 535 o .8508 710 0.8546 1166 EF 2 0-8356 320 .8489 215 .8518 355 0.8560 350 2589 1887 3886 2805 Lot. No. of tubes. Hours req. 40 9 48 5 tubes 72 4 tubes Spec. Vol., 41 5 40 Spec. Vol., 42 5 69 Spec. Vol., 43 4 10 days 2 17 days 2 Spec. tubes tubes Vol., Frac. grav. cc. grav. cc. grav. cc. grav. cc. A 0.8305 380 O ,8 3 l6 235 o 8325 210 .8435 65 Bl 0.8438 515 .8460 290 O .8487 265 .8546 115 B 2 0-8453 155 o .8480 65 o .8515 54 . . . . Q 0.8518 600 8523 375 'o ,8540 335 .8575 20O Q 0-8539 170 o .8540 100 O .8567 56 A 0.8550 685 .8558 470 .8572 420 O .8605 220 ft 0.8560 330 o .8571 no o .8582 175 .8640 50 EF l 0.8605 780 o .8620 580 .8640 675 .8650 225 EF 2 0.8620 600 ,8622 320 o 8650 200 .8615 78 2420 953 4215 2545 In this series, as well as those following, the pressure in the tubes was dimin- ished intermittently. Lot. No. of tubes. Hours 44 3 48 2 tubes 45 5 66 46 5 93 47 5 13 days 1 req. 96 1 tube Spec. Vol., Spec. Vol., Spec. Vol., Spec. Vol., Frac. grav. cc. grav. cc. grav. cc. grav. cc. A 8330 85 0.8362 170 o .8332 210 0.8340 145 B 1 o .8505 175 0.8510 210 .8480 260 o . 8500 275 , O.8522 80 o .8505 5O C\ o 8582 I 55 0.8562 265 o 8554 300 0.8553 320 C 2 o .8605 6 5 0.8585 50 o .8567 95 0.8576 50 P t .8605 195 0.8567 425 o .8600 370 0-8595 430 A o .8620 120 0.8580 100 .8613 1 20 0.8618 70 F, o .8672 240 0.8659 615 o .8666 610 0.8665 330 F 2 o .8680 175 0.8670 150 o .8680 130 0.8670 215 1210 2065 2145 1835 Lot. 48 49 50 No. of tubes. 5 7 5 Hours 14 days 2 48 72 4 tubes req. 89 1 tu be Spec. Vol., Spec. Vol., Spec. Vol., Frac. grav. cc. grav. cc. grav. cc. A 0.8385 125 0.8341 255 O .8320 170 B, 0.8530 275 0.8505 395 .8485 230 S, 0.8520 95 .8500 70 c, 0.8568 320 0.8560 380 O 8565 300 C 2 0.8586 90 0.8572 230 8577 IOO 0, 0.8610 325 0.8620 500 .8609 480 A 0.8623 115 0.8625 290 o .8626 125 EF, 0.8695 330 0.8705 500 .8685 640 EF 2 o . 8700 80 0.8705 580 .8700 235 1660 3225 2350 Observations on the Second Fractionation. Specific Gravity. The range of the specific gravities grows smaller as the oils to be fractionated become lighter, and less complex. Thus, in Lot 36, the range of specific gravity ex- tends from 0.8272, the value for Fraction A, to 0.8356, the 1 Owing to the weakness of the water pressure, the pressure in the tubes was only slightly diminished. The tubes were taken down before the reservoirs were com- pletely exhausted. The distances to which the oil had risen were 35, 25, 30, 20, 10 cm. from the tops of the tubes. 2 Owing to the weakness of the water pressure, the pressure in the tubes was di- minished but slightly during this time. The tubes were taken down before the reservoirs were completely exhausted. The distances to which the oil had risen were 50, 35, 30, 60, 55 cm. from the tops of the tubes. value for EF 2 , the difference between them being 0.0084. In Lot 38, the mother oil, of specific gravity 0.8433, yielded fractions whose specific gravities ranged from 0.8331 to 0.8518, amounting to a difference of 0.0187. This fact ap- pears to be general throughout the various lots, and points to the gradual formations of mixtures which will pass through the earth unaltered, just as the fractionation by distillation tends to yield substances with definite boiling points. Color. The color of the oils in this fractionation shaded from a very light yellow to greenish black. Odor. The odor of the crude petroleum vanished com- pletely from the oils of this fractionation. Volume of Oil Retained by the Earth. The oil retained by the earth in this fractionation amounted to approximately 50 per cent., a smaller percentage, as is naturally to be ex- pected, than in the fractionation of the crude petroleum. The Third Fractionation. The following oils obtained from the second fractionation were united for the third fractionation: Lot 51. Specific Gravity 0.8316. Specific Volume, Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 36 A 0.8272 1 60 42 A O. 8325 210 39 A 0.8290 255 44 A 8330 85 37 A O.8292 135 36 BZ O 8331 58 40 A 0-8305 380 38 A 8331 1 80 36 B* 0.8315 216 46 A 8332 210 4 1 A 0.8316 235 36 C l 8334 350 50 A 0.8320 170 49 A 8341 255 2899 Lot 52. Specific Gravity 0.8343. Lot. Fraction. Specific Volume, gravity. cc. Lot. Fraction. 36 36 47 D 0.8330 360 0.8339 320 0.8340 145 36 36 36 EF l EF 2 C 2 Specific Volume, gravity. cc. 0.8347 720 0.8356 320 0.8355 8 5 1950 43 Lot 53. Specific Gravity 0.8433. Specific Volume, Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 45 A 0.8362 170 38 B, 0.8447 175 48 A 0.8385 125 40 B, 0-8453 155 37 B l 0.8421 215 38 B, 0-8455 2IO 39 BI 0.8432 355 39 B, 0.8458 50 40 B 0.8438 515 1970 Lot 54. Specific Gravity 0.8473. Specific Volume, Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 39 B 2 0.8458 60 50 B, 0.8485 230 B. o . 8460 290 42 0.8487 265 37 0.8467 295 39 c l o . 8492 455 B\ 0.8480 65 38 Ci o . 8490 305 1965 Lot 55. Specific Gravity 0.8485. Specific Volume, Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 37 Pi 0.8468 340 37 EF, 0.8489 215 37 D, 0.8485 152 38 D t o . 8492 400 37 EF l 0.8480 535 47 B, o . 8500 275 1917 Lot 56. Specific Gravity 0.8508. Specific Volume. Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 50 B 2 o . 8500 70 45 B, 0.8510 210 49 B, 0.8505 395 39 c* 0.8513 1 80 44 0.8505 175 4 2 B, 0.8515 54 46 B 2 0.8505 50 40 0.8518 600 38 C 2 0.8505 175 1909 Lot 57. Specific Gravity 0.8509. Specific Volume, Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 38 A 0.8505 740 38 A 0.8509 295 39 EF 1 0.8508 710 38 EF 2 0.8518 355 2100 44 Lot 58. Specific Specific Volume, Gravity 0.8558. Specific Volume* Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 49 B 2 0.8520 95 49 c, 0.8560 380 45 B 2 0.8522 80 45 c, 0.8562 265 0.8523 375 50 c, 0.8565 300 48 &t 0.8530 275 42 c, 0.8567 56 4 C 2 0-8539 170 46 c, 0.8567 95 42 c, o . 8540 335 48 C, 0.8568 320 0.8540 100 49 c, 0.8572 230 47 c, 0-8553 320 43 c, 0.8575 200 46 c, 0-8554 300 3896 Lot 59. Specific Gravity 0.8563. Specific Volume, Specific Volume.. Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 39 EF, o . 8546 166 41 ft .8571 1 10 40 D, 0.8550 685 4 2 D! .8572 420 4 1 D t 0.8558 470 45 D, O .8580 100 39 EF, 0.8560 350 42 D, .8582 175 40 D, 0.8560 330 48 c, O .8586 90 45 D, 0.8567 425 47 D, 8595 430 4750 Lot 60. Specific Gravity 0.8615. Specific Volume, Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 46 > t . 8600 370 41 EF, O .8620 580 49 EFi o . 8605 780 44 ft .8620 120 43 n 0.8605 220 49 ft .8620 500 44 D l o . 8605 195 EF, .8622 320 50 n o . 8609 480 48 D, .8623 115 48 D, 0.8610 325 49 D, O .8625 290 46 0.8613 120 50 D, .8626 125 47 D 2 0.8618 70 42 E, o .8640 675 4 EF 2 0.8620 600 5880 Lot 61. Specific Gravity 0.8680. Specific Volume, Specific Volume Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 42 EF 2 0.8650 200 4 6 EF, .8680 130 43 EF, 0.8650 225 44 EF, O .8680 175 45 EF, 0.8659 615 5 EF, .8685 640 47 0.8665 330 48 EF, .8695 330 46 EF\ 0.8666 610 50 EF, .8700 235 47 EF 2 0.8670 215 49 EF, .8705 500 45 EF, 0.8670 150 49 EF, o .8705 580 44 EF, 0.8672 240 4975 45 The oils thus united were fractionated by fuller's earth again, with the results given in Table VI. Table VI. The Third Fractionation. Lot. No. of tubes. Hours req. 51 31 60 Spec. Vol., 52 2 60 Spec. Vol., 53 2 48 Spec. Vol., 54 2 48 Spec. Vol., Frac. grav. cc. grav. cc. grav. cc. grav. cc. A .8213 92 .8219 65 O .8266 73 0.8303 66 B .8303 185 .8333 143 .8431 0.8488 JI 5 C l O 8337 165 O 8375 190 O .8464 175 0.8518 175 C 2 o .8345 90 D l -8353 210 '8388 188 .8468 145 0-8523 1 60 D 2 o .8356 170 O 8393 90 o .8474 115 0.8528 105 E 1 o .8366 385 8403 175 .8473 202 0-8530 245 E 2 .8411 92 .8488 73 0.8548 60 F i o 8373 190 O 8431 88 o .8496 170 0.8548 H5 1487 1031 1068 1091 Lot. 55 56 57 58 No. of tubes. 2 2 2 4 Hours 2 48 1 tube 96 96 723 tubes req. 72 1 tube 92 1 tube Spec. Vol., Spec. Vol., Spec. Vol.. Spec. Vol., Frac. grav. cc. grav. cc. grav. cc. grav. cc. A O 8283 58 O 8313 75 .8336 55 0.8318 170 B 8457 100 .8488 135 .8491 130 0-8531 260 C l 8515 155 8546 170 o 8528 1 80 0.8578 205 C 2 0.8592 105 D, 8521 220 8553 150 o .8551 '185 0.8588 205 D 2 8543 50 8560 92 o 8573 45 0-8593 340 E l O 8540 270 o 8553 145 o .8568 170 o . 8603 325 E 2 8563 90 .8588 70 0.8613 170 F 8566 1 80 8575 130 .8611 170 0.8628 275 1033 987 1005 2055 1 The tin tubes used in these lots were 1.5 inches in diameter. 2 The pressure in the tubes was diminished intermittently. 4 6 Lot. 59 60 61 No. of tubes. 5 65 Hours req. Frac. A B 72 Spec, grav. Vol., cc. 72 Spec, grav. Vol., cc. 5 days. 1 Spec. Vol. grav. cc. 0. 8328 195 .8343 195 0.8413 .. 0. 8508 340 .8540 330 O .8601 .. O. 8578 325 .8601 290 O .8683 .. 0. 8588 112 .86l8 130 O. 8608 490 O .8628 440 O .8709 .. 0. 8623 135 .8638 85 . . . . 0. 8628 475 O .8664 425 8688 .. 0. 8633 155 .8683 140 0. 8673 330 -8703 310 O 8691 .. A A Observations on the Third Fractionation. Specific Gravity. The decrease in the range of specific gravity as the oils supplied become lighter is observed in this fractionation as in the preceding ones. Color. The lightest oils were almost colorless; the heavier oils were dark brown to green. Odor. Most of the oils possessed an agreeable odor. Prolonged Diffusion. In Lot 61, the time required for the oils to reach the tops of the tubes was five days. No frac- tionation, as is evident from an examination of the specific gravities, occurred in the lower parts of the tubes. The heavier oils of fractions D, E, and F were exceedingly viscous. Volume of Oil Retained by the Earth. -The volume of oil retained by the earth in this fractionation amounted to ap- proximately 45 per cent. The increase in the yield of oil indi- cates, therefore, a process of purification, in which, as will be shown later, such compounds as the unsaturated hydrocar- bons are removed. The. Fourth Fractionation. The following fractions obtained from the third fractiona- tion were united for the fourth fractionation : 1 See below, this page. 47 Lot 62. Specific Gravity 0.8298. Specific Volume, Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction gravity. cc. 5 1 A 0.8213 92 51 B o 8303 185 52 A 0.8219 65 56 A 0-8313 75 53 A 0.8266 73 58 A 0.8318 170 55 A 0.8283 66 59 A 0.8328 195 54 A 0.8303 58 979 Lot 63. Specific Gravity 0.834.3. Specific Volume, Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction gravity. cc. 52 B 0.8333 143 51 c 2 0-8345 90 57 A 0.8336 55 3i A 0.8353 210 Ci 0.8337 185 51 A 0.8356 170 60 A 0.8343 195 1048 Lot 64. Specific Gravity 0.8368. Specific Volume, Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 51 , 0.8366 388 52 c, 0-8375 190 51 F 0.8372 190 52 D, 0.8388 188 956 Lot 65. Specific Gravity o. 84.30. Specific Volume, Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 52 D, 0-8393 90 52 F 0-8431 88 52 E, 0.8403 175 55 B l 0-8457 100 52 0.8411 92 53 Q o . 8464 175 53 B! 0.8431 115 53 A 0.8468 980 Lot 66. Specific Gravity o, ,8483. Specific Volume, Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction gravity. cc. 53 fSj 0.8473 202 56 B l 0.8488 135 53 A 0.8474 115 53 E 2 o . 8488 73 54 B l 0.8488 115 59 B,. 0.8508 330 970 Lot 67. Specific Specific Volume, Gravity o '85I3- Specific Volume, Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 57 B l 0.8491 130 54 Q 0.8518 175 59 BI 0.8508 10 55 A 0.8521 220 55 C l 0-8515 155 58 B, 0.8531 26O 950 4 8 Lot 68. Specific Gravity 0.8533. Lot. Fraction. Specific gravity. Volume, cc. Lot. Fraction. Specific gravity. Volume cc. 54 A 0.8523 1 80 54 ^i 0.8530 245 54 D 2 0.8528 105 60 B 0-8540 330 57 c, 0.8528 1 80 1040 Lot 60. Specific Gravity 0.8556. Specific Volume, Specific Volume > Lot. Fraction. *. gravity. cc. Lot. Fraction. gravity. cc. 55 EI o . 8540 270 56 E, 0.8553 H5 55 A 0-8543 50 56 D 2 0.8560 92 56 c, 0.8546 170 56 E 2 0.8563 90 54 E 2 0.8548 60 55 F 0.8566 1 80 54 F 0.8548 145 57 E l 0.8568 170 57 n 0.8551 185 57 A 0-8573 45 56 D i 0.8553 150 56 F 0-8575 130 1882 Lot 70. -Specific Gravity 0.8596. Specific Volume, Specific Volume Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 58 c t 0.8578 205 60 Cj 0.8601 290 59 c t 0-8578 325 58 E, o . 8603 325 58 D 0.8588 205 59 A o . 8608 490 59 C 2 0.8588 112 57 F 0.86II 170 57 E, ' 0.8588 70 58 E 2 0.8613 170 58 0.8592 105 60 C 2 0.8618 130 58 D 0-8593 340 2937 Lot 71. Specific Gravity 0.8638. Specific Volume Specific Volun Lot. Fraction. gravity. cc. Lot. Fraction. gravity. cc. 59 D 2 0.8623 135 60 D 2 0.8638 85 60 A 0.8628 440 59 E 2 0.8633 155 59 0.8628 475 60 E l o . 8664 425 -Q zr ^ QA/-.Q *%**! 1990 49 Table VII. The Fourth Fractionaiion. Lot. 62 63 64 65 No. of tubes. 1111 Hours req. 72 Spec. Vol., 72 Spec. Vol., 90 Spec. Vol., 48 Spec. Vol., Frac. grav. cc. grav. cc. grav. cc. grav. cc. A O 8243 32 0.8273 45 0. 8297 4 1 O .8308 42 B 0.8298 71 0-8357 75 O. 8378 57 .8428 70 C 0. 8323 90 0.8378 95 0. 8401 81 .8463 92 D O 8330 115 0-8383 130 0. 8408 H5 O 8473 130 E 0, 8333 130 0.8388 98 O. 8413 135 .8471 130 F 0. 8341 75 0-8393 95 0. 8418 70 .8485 80 513 538 499 544 Observations on the Fourth Fractionation. Specific Gravity. As in the preceding fractionations, the decrease in the range of specific gravity as the mother oils become lighter is again observed in this fractionation. It is evident, moreover, that there is a constant forward ac- cumulation towards definite and constant mixtures. The lighter oils of one lot are found to possess specific gravities closely approaching those of the heavier oils of the preceding lot. Color. The oils of Fraction A were almost colorless; the color of the heavier oils ranges from green to light brown. Odor. All the oils of this fractionation possessed agreeable odors. Volume of Oil Retained. The volume of oil retained by the earth amounted to approximately 40 per cent. Deposition of Paraffin. In Fractions A and B of several of the lots, a fine, crystalline deposit separated out, and col- lected upon the bottom of the bottles containing the oils. When the oils were warmed, this deposit dissolved completely, showing it to be paraffin. Chemical Examination of the Fractionated Oils. U maturated Hydrocarbons. Action of Concentrated Sulphuric Acid. The percentage of volume of oil absorbed by concentrated sulphuric acid (specific 50 gravity 1.84) was determined according to the following pro- cedure: Ten cc. of the oil to be examined were measured into a glass-stoppered bottle, and thirty cc. of concentrated sul- phuric acid were added. The mixture was thoroughly shaken in a machine for thirty minutes and then poured into a burette. After allowing sufficient time for any oil that might be mechan- ically absorbed in the acid to rise to the top, the volume of un- absorbed oil was read directly over the acid. Owing to the formation of heavy emulsions, no attempt was made to neu- tralize and wash the oil. The results of the analyses are given in the table below : Per cent. Per cent. by volume by volume Lot. Fraction. absorbed. Lot. Fraction. absorbed. 5i A 2.3 51 D 1 11.5 51 B 6.1 51 D 2 12.0 51 C l 9-1 5i E 12.5 51 C 2 10.2 51 F 14.5 Action of Bromine. The method employed for determining the amount of bromine absorbed by the oils was as follows: Between 0.5 and 0.9 of a gram of the oil to be examined was dissolved in 10 to 15 cc. of carbon tetrachloride. Five cc. of a standard solution of bromine in carbon tetrachloride were then introduced, and the solution allowed to remain, with oc- casional shaking, in a dark place for 30 minutes. Ten cc. of a 10 per cent, solution of potassium iodide were then added, and the amount of iodine liberated determined immediately by titrating with a standard solution of sodium thiosulphate. A few drops of a starch solution were introduced to mark ac- curately the end of the titration. The amounts of bromine absorbed by addition and substitution were not estimated separately. The amount of bromine absorbed, calculated upon the basis of loo grams of oil, are given in the following table: Table VIII. First Fractionation. Lot. Fraction. 32 A 32 B 32 C Per cent, of bromine absorbed. Lot. Fraction. Per cent, of bromine absorbed. 5.02 6.96 7.40 32 D 32 E Crude oil 7-87 8.00 7.64 Second Fractionation. Lot. 36 36 36 36 36 Lot. 51 51 51 Lot. 62 Fraction. A d Fraction. A B C Fraction. A Per cent. Per cent. of bromine of bromine absorbed. Lot. Fraction. absorbed. 4-74 36 D, 6.81 54 36 A 6.28 5-66 36 EF l 6-49 5-56 36 HF 2 7.18 6.18 Third Fractionation. Per cent. of bromine absorbed. 3-27 4-47 Fourth Fractionation. Per cent, of bromine absorbed. Lot. Fraction. 2.86 62 E Per cent. of bromine Lot. Fraction. absorbed. 51 D 4.92 51 E 4.71 51 F 5-36 Per cent, of bromine absorbed. 3-73 These results demonstrate conclusively that the unsatura- ted hydrocarbons tend to collect in the lower sections of a layer of fuller's earth through which the oil is allowed to dif- fuse. These figures confirm the results obtained by Gilpin and Cram in their work on the Pennsylvania petroleum. In their investigation, distillation by heat was employed in order to obtain fractions that could be readily studied. In this work the relative amounts of the unsaturated hydrocarbons in the various oils were determined directly as they came from the earth. The percentages by volume of oil absorbed by concentra- ted sulphuric acid represent only approximately the per- centages of unsaturated hydrocarbons, since, as was shown previously, any benzene which may have been present in the oils was also removed by the concentrated acid. This fact rendered impossible a quantitative separation of the aro- matic from the unsaturated hydrocarbons. Since no other methods, besides nitration and sulphonation, neither of which could be here employed, were available, no results as to the 52 relative amounts of the aromatic hydrocarbons in the various fractions could be obtained. It is evident from the results of the bromine determina- tions that, as the fractionation proceeds, the amounts of un- saturated hydrocarbons become smaller and smaller. A comparison of the amounts of bromine absorbed by Fraction A of the first, second, third and fourth fractionations is given below for the purpose of bringing out this point more clearly : Per Cent, of Bromine Absorbed by Fraction A. First Second Third Fourth fractionation. fractionation. fractionation. fractionation. 5.02 4.74 3.27 2.86 Sulphur Compounds. The sulphur in the various oils was determined by the usual method of combustion. For these determinations, the oils obtained from one tube of Lot 6 were employed. The results are given in the following table: Lot 6 Fraction. A B C The percentage of sulphur in the Fractions A, C and E of Lot 51 was also determined. The results were as follows: Table X.Lot 51. Per cent. Per cent. Fraction. of sulphur. Fraction. of sulphur. A 0.003 ^ 0.006 C o . 040 These results show that the sulphur tends to collect in the oils in lower sections of the tube. As the fractionation pro- ceeds, the proportion of sulphur becomes smaller. The fig- ures below indicate that as the oil is subjected to repeated fractionations, the sulphur is gradually removed: Per Cent, of Sulphur. First fractionation. Second fractionation. Third fractionation. Fraction A o . 04 .... o . 003 Fraction E 0.16 .... 0.006 Fraction C o . 08 o . 040 Specific Per cent. Lot 6. Specific Per cent. gravity. sulphur. Fraction. gravity. sulphur. 0.8195 O.O4 D 0.8510 O.O9 0.8362 0.05 E o . 8600 o. 16 o . 8440 Lost 53 Selective Action of Fuller's Earth. When the earth, from which as much oil as possible has been extracted by prolonged treatment of water, is dried, and then digested with ether, oils of surprisingly high specific gravity and viscosity are obtained. In the experiments undertaken to study the selective ac- tion of fuller's earth, the procedure was as follows: The earth under examination was thoroughly treated with water until no more oil appeared. This muddy earth of the consis- tency of thin liquid paste was spread upon porous plates and allowed to dry at room temperature. Several weeks usually elapsed before the earth became completely dry. It was then finely pulverized, and after being thoroughly soaked and shaken with ether the mixture was allowed to remain undis- turbed for 24 hours or more. The mixture was then filtered and the dissolved oil recovered by distilling off the ether from the filtrate. The residual earth was then digested with ether for some time by means of an electric stove that com- pletely surrounded the flask. The oil thus extracted was added to the oil first obtained. In several cases the residual earth was treated further with ether in the Soxhlet extractor. The results of these extractions are given in the following table: Table XL Speci crrax Lot. Fraction. 7 A 8 A 18 A l 18 A 2 19 A t 19 A 2 19 A z 25 ^i 25 A 2 The specific gravity of none of the oils of the first and sec- ond fractionation, extracted with ether, except those of Lot 19, could be determined at 20 C. All were extremely vis- Specific gravity Specific gravity at 50. Lot. Fraction. at 50. o . 8470 25 A 3 0.8391 0.8502 25 B 0.8489 0.8419 Specific gravity at 20. o . 8400 51 A 0.8368 0.8495 51 B 0.8473 0.8495 51 C 0.8491 O.86OO 71 D 0.8568 0-8363 51 E 0.8518 0.8381 51 F 0-8553 54 cous; those of Lot 25 were so viscous at this temperature that they would not flow when the bottles containing them were inclined. The color of the oils ranges from brown to black. The ethereal solutions, however, of many of the oils were very light in color. It is interesting to compare the specific gravities of the oils extracted with ether with those of the corresponding oils extracted with water. For this purpose, the oils extracted by water and by ether from the earth of Lot 51 are chosen. In the table below, the specific gravities of these oils at the same temperature, i. e., 20, are given: Table XII. Lot 51. Specific Gravity at 20. Frac. Ether. Water. Frac. Ether. Water. A 0.8368 0.8213 D 0.8568 0.8353 B 0.8473 0.8303 E 0.8518 0.8366 C 0.8491 0.8337 F 0.8553 0.8373 As the figures above indicate, the specific gravities of oils extracted with ether are much higher than those of the corre- sponding oils extracted with water. The presence of such heavy and viscous oils in the upper sections of the tube can be explained only by assuming that they were carried to these heights in solution with the lighter oils, and were then re- moved by the earth. Since such viscous oils are totally un- able to diffuse by capillarity to any appreciable extent, it is not probable that their transportation to the upper parts of the tube was effected by capillary diffusion. Chemical Examination of the Oils Extracted by Ether Un- saturated Hydrocarbons. Action, of Concentrated Sulphuric Acid. The percentage by volume of oil absorbed by concentrated sulphuric acid (specific gravity 1.84) was determined according to the following pro- cedure: Ten cc. of the oil to be examined were measured into a glass-stoppered bottle, and thirty cc. of concentrated sul- phuric acid were added. The mixture was thoroughly shaken in a machine for 30 minutes and then poured into a burette. 55 After allowing sufficient time for any oil that might be mechan- ically absorbed in the acid to rise to the top, the volume of unabsorbable oil was read directly over the acid. Owing to the formation of heavy emulsions, no attempt was made to neutralize and wash the oil. The oils selected for examination were those extracted by ether from the earth of Lots 36 and 51. The results of the analyses are given in the following table : Table XIII. Oils extracted by ether. Per cent. Oils extracted by water. Per cent. Oils extracted by ether. Per cent. Oils extracted by water. Per cent. by voli time by vol urne by voh ime by vo lume Lot. Fraction. absorbed. absorbed. Lot. Fraction. absorbed. absorbed. 36 A 24 3 .0 51 c 17 .0 9 . I 36 B 37 O 10 4 51 D 16 4 ii 5 51 A 7 2 3 51 E 16 5 12 5 51 B ii 5 6 . i 51 F 18 .0 H 5 Action of Bromine. The method employed for determining the amount of bromine absorbed by the oils has already been described (p. 50). The amounts of bromine absorbed, calculated upon the basis of loo grams of oil, are given in the table below. The values for the corresponding oils extracted with water are also given for comparison. Table XIV. Oils Oils extracted extracted by ether. by water. Per cent. Per cent. of bromine of bromine Oils Oils ex- extracted tracted by ether, by water. Per cent. Per cent, of bromine of bromine Lot. Fraction. absorbed. absorbed. Lot. Fraction. absorbed. absorbe< 32 A 5- 30 5 02 51 B 4 45 4 36 32 B 7- 39 6. 9 6 51 C 6 .27 5 03 36 A 5 72 4 74 51 D 6 .09 4 92 36 B 6. 10 5 40 51 E 5 .98 4 7i 36 C 6. 72 5 56 51 F 5 .20 5 36 51 A 3 .27 3 .27 / As these results clearly demonstrate, fuller's earth retains the unsaturated hydrocarbons, thus exercising a selective action. 56 Sulphur Compounds. The sulphur in the oils obtained by extraction with ether was determined by the usual method of combustion. The results are given in the table below: Table XV. Oils extrac- Oils extrac- Oils extrac- Oils extrac- ted by ted by ted by ted by ether. water. ether. water. Per cent. Per cent. Per cent. Per cent. Lot. 51. of sulphur. of sulphur. Lot 51. of sulphur. of sulphur. A 0.004 0.003 D 0.060 B O.OII E 0.080 C 0.050 0.040 F O.OSO O.OO6 The selective action of the earth, in regard to the sulphur compounds, is indicated by these results. This fact was also pointed out by Richardson and Wallace. It is very prob- able that the earth also retains largely the nitrogen compounds in the oil, and may also remove to a greater or less extent the benzene hydrocarbons. These results seem to furnish evidence in favor of the view that the Pennsylvania oil diffused, at some time in the history, through porous media, which exercised a selective action upon it, removing a large part of the unsaturated and sulphur com- pounds, and probable the benzene and nitrogen compound. SUMMARY. 1. When a solution of benzene and a paraffin oil is allowed to diffuse upward through a tube packed with fuller's earth, the benzene tends to collect in the lower sections and the paraffin oil in the upper sections of the tube. 2. When crude petroleum diffuses upward through a tube packed with fuller's earth, a fractionation of the oil occurs. The oil that is displaced by water from the earth from the top of the tube possesses a lower specific gravity than the oil obtained from the earth at the bottom of the tube. 3. As the fractionation proceeds, the range of specific grav- ity covered in succeeding fractionations becomes smaller, in- dicating a movement towards the production of mixtures which will finally pass through the earth, unaltered. 57 4. In the fractionation of petroleum by capillary diffusion through fuller's earth, the amounts of unsaturated hydrocar- bons and sulphur compounds in the resulting fractions in- crease gradually from the lightest oils at the top to the heavier oils at the bottom of the tube. 5. Fuller's earth tends to retain the unsaturated hydrocar- bons and sulphur compounds in petroleum, thus exercising a selective action upon the oil. BIOGRAPHICAL. The author of this dissertation, Oscar Ellis Bransky, was born in Baltimore, January 29, 1886. He received his early educa- tion in the public schools of Baltimore. In 1904, he graduated from the Baltimore City College, and in October of the same year he entered the Johns Hopkins University. He received the degree of Bachelor of Arts in 1907. He then entered the post-graduate department of the university, pursuing Chem- isty as his major, and Physical Chemistry and Geology as his minor subjects. During 1908-1909 he acted as laboratory assistant to Professor Renouf . UNIVERSITY OK CALIFORNIA LIBRARY THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW OK 18 SEP 22 MAR 28 1948