THE SEPARATION AND IDENTIFICATION OF ACIDS IN LOW TEMPERATURE TAR BY DELLA DARLE JUNKIN B. A. University of Michigan 1912 THESIS Submitted in Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE IN CHEMISTRY IN THE GRADUATE SCHOOL OF THE UNIVERSITY OF ILLINOIS 1921 Digitized by the Internet Archive in 2015 https://archive.org/details/separationidentiOOjunk UNIVERSITY OF ILLINOIS THE GRADUATE SCHOOL June 2 192.1- I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY JQella.Darle Junitin ENTITLED The Separation and Ident.i fin n.t.i on of Acids in Low Temperature Tar. BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE DEGREE OF _ I n Charge of Thesis Head of Department Recommendation concurred in* Committee on Final Examination* *Required for doctor’s degree but not for master’; i ' TABLE OF CONTENTS Acknowledgment Page I, Introduction 1 Purpose of investigation 1 Historical 2 II* Experimental 1, Fractionation of tar ^ a. Extraction and purification of acids 6 b. Fractionation of acids 7 c. Physical nature of acids 10 2. Extraction of acids without fractionation 10 a. Dehydration of acids 11 b. Effect of steam distillation 11 c* Effect of vacuum distillation 11 d. Use of solvents 11 e. Choice of fractional extraction solvent 12 f. Choice of precipitation solvent 12 g; Purification of solid acids 13 3. Tests to determine nature of acids 14 III. Summary and Conclusion 20 IV. Bibliography 21 . , . . . ACKNOWLEDGMENT This investigation was carried out in the Chemical Laboratory of the University of Illinois 1920-21 under the direction of Dr. Layng. The writer takes this opport- unity to express appreciation for the assistance given by various numbers of the department, but more especially to Dr. Layng for his cooperation and timely assistance which has meant so much to the progress and development of this work. . THE SEPARATION AND IDENTIFICATION OF ACIDS IN LOW TEMPERATURE TARS. I, Introduction The purpose of this investigation is to secure, if possible, the acids in low temperature tar in their original form or as nearly so as possible, and thereby gain some knowledge of the mature of the constituents of the tar and perhaps of the original coal. In the process of low temper- ature carbonization, there is some decomposition of the higher boiling consti tuents , but the amount of this is small in comparison to that produced by high temperatures. Because of this there should be found in tar some of the constituents of the coal not wholly decomposed among which many acids undoubtably occur. The percentage of these acidic bodies known to be high in low temperature tar, their separ- ation and identification may give some idea of the nature of those compounds which decompose on application of heat to give bodies having lower boiling points and some which later polymerize to give the higher boiling liquids and solids in tar. By the low temperature carbonization^^ is meant the maintenance of such temperature that the formation of naphthalene and fuel oils is not begun, as a result of the decomposition of the higher boiling substances, ( 2) This transition temperature v ' occurs between 530 de- grees and 560 degrees but by careful distillation a very slight decomposition can be made to occur which increases . * * ' - 2 - with rapid, rise in temperature resulting in a thickening (3) which Schneider' ' attributed to be due to polymerization! and thus was he able to conclude from his work that the high boiling substances passed off undecomposed ^ 4 ^ from the coal in the coking process. Professor Parr (5) says that this secondary decomposition must be prevented, and may be, by maintaining a temperature of less than 750 degrees, otherwise the "oxygenated constitu- ents of the coal" will react with the phenol soluble substances whicn form the binder for the coke, ( 6 ) Already there have been found over three hundred substances in tar, of which about one hundred fifty have been estimated and over ninety separated and purified. Among these are. neutral basic, and acidic classes, containing oxygen, nitrogen, or sulphur, either in the ring or in the side chain and comprising five and six membered rings exemplied by benzene, phenol, pyridine, thiophen, and coumarone. Of all compounds contained, tar yields but few in large quantities, such as benzene, phenol, and naphthalene. (7) The higher homologues of phenol occur in large quantities in low temperature tar both as polyhydric phenol and their esters which form resins. The latter remaining undistilled if a fractionation is made at 240 degrees and form a perfect material for road making. Various methods have been employed for the purification of the constituents v ' of tar among which are the formation of salts of organic and inorganic reagents, with their sub- ■ . ' - •'■i ‘) Fischer elso maintains that the percent of phenols - 10 - can "be doubled by cracking tars at high temperatures. Having found no satisfactory temperature for taking fractions the cuts were made at those temperatures known to be about the boiling points of several substances previously found intars, ie, phenol 186 degrees, cresols 200 degrees, xylenols 210 degrees and 225 degrees. Color tests with FeCl-z were made on these members and £ positive results were secured but in no case was an individual member isolated. Comparative tests of the members of the liquid acids were made with those of the solid acids secured by the second method of extraction. This consisted in treating the crude tar with a 50% solution of potassium hydroxide. Successive extractions with the alkali finally gave an aqueous extract that contained but a small part of the acids, but still was not colored on further admixture with the tar. It was observed at the same time that all of the alkali being used was not recovered from the tar and upon examination it was found held in a black viscous emulsion below the remaining hydrocarbon layer. Because of the extreme viscosity of tnis product, removal of hydrocarbon by use of a separatory funnel was impossible. To remove hydrocarbon, suction was applied and from the remain- ing alkaline solution the acids were liberated by means of sulphuric acid. Having washed out any potassium sulphate re- maining, the acids were subjected to a combination of steam and vacuum for drying. Drying by the usual process was avoided because the viscosity of a.cids caused local superheating and • ' * - i }tf , ■ i. - ■ ■ . ... *'• - 11 - resulting d'-composti on . To avoid this the acids were warmed on a steam bath and transferred to a three liter flask having a long neck. Dry steam was passed into the war m acids and at the same time a vacuum of 600 ram. was maintained. After about 300 cc. of distillate had been collected the steam supply was shut off and the vacuum simultaneously increased to 200 mm. By this method complete dehydration was accomp- lished in a short time. The small amount of acids appearing in the distillate were low boiling. The yield of acids thus obtained was 36$. A sample of the dried acid was subjected to steam vacuum distillation as applied to the drying process with the same results, ie, any acid appearing in the distillate was of the lowest boiling fraction and a yield of a few cc. per liter of steam. This low yield of acids which were transferred in the steam showed that the percent of phenol in this sample of tar was very low. Extraction With Solvents. If the acids in the tar are to be separated and examined as they are in their original form, these methods are useless. By the time a sufficiently large yield could be obtained for satisfactory observation, the nature would be so changed that their resemblance to high temperature tan acids would be closer. Ligroin and petroleum ether were tested out as precipit- ating agents. Petroleum ether is more satisfactory, there being a heavy loss of all the low melting solids in the . - . . ♦ — 1 £— ligroin, whereas, it was from the cold petroleum ether solu- tion that these could he obtained and resulted in giving the nearest pure of any of the extracted products. Moreover, the high boiling point of the ligroin made sufficiently rapid evaporation by vacuum an impossibility. In no case, could there be produced a coat of ice on the outside of the filter flask and this was essential to securing the last traces of the dissolved solid, acids* The solvents tested for their extracting ability are the following, given in the order of their decreasing power to dissolve the acids: phenol, sodium phenolate, alcoholic KOH, CS 2 » CHCI3, ether, benzene, ligroin, alcohol, and petroleum ether, of which the fractional extraction solvents chosen were phenol, carbon disulphide, chloroform, and ether. A sample of the crude acids was extracted with ether and any undissolved residue was filtered and later treated success- ively with CHCl^, CS2» and phenol until none remained. The ether solution was poured slowly and with constant stirring into about ten times its volume of petroleum ether and a fine- ly divided light brown precipitate formed which filtered readily into a suction flask containing more petroleum ether previous- ly cooled by rapid evaporation. In the filter flask another precipitate appared and this being filtered and the filtrate as above gave second precipitate. This process was followed until no more precipitate could be secured even by using a large excess of petroleum ether at zero degrees, giving in each tr ial a precipita.te lighter in color and more readily - - 13 - soluble in petroleum ether on warming. The chloroform, carbon disulphide, and phenol extracts were likewise treated with petroleum ether as previously, the resulting precipitation in each case darker than the preceed- ing and much less soluble in the petroleum ether. Difficulties in Extraction ar d Precipitation. Carbon disulphide is a better solvent than ether, but for the separation of those acids which are soluble in petrol- eum ether at room temperature and precipitate on cooling, CSg is a failure. It can not be as readily evaporated as ether by vacuum and holds all of the more soluble bodies in solution. If the CS^ extracted be poured rapidly into the petroleum ether, the very rapid precipitation of the solute occluded liquid acids resulting in a tarry mass. This difficulty is greatly lessened if the petroleum ether is stirred rapidly and the CS£ solution added slowly. This trouble is not en- countered in the ca.se of the ether extract because the sub- stances dissolved by ether are more readily soluble in petrol- eum ether. If a filter paper is used for two or more filtrations there appears a resinous mass on the paper which causes very slow filtration. This is coused by the high boiling acids clogging the filter and then dissolving any solids to be filtered. A purification of the precipitated acids was made in the same manner as that previously done in the extraction and . . . . . « -14- this resulted in the removal of all viscous liquid acids and a very readily filtration. The residue that was soluble in phenol only, presented difficulties due to the fact that when the phenol solution was poured into the petroleum ether it formed a viscous mass adhearing firmly to the walls of the beaker and giving no precipitate in the petroleum ether. When a small amount of ethyl ether is added to the solution, it reduces the solvent action of the phenol, and then a fine black precipitate is liberated which is readily filtered. Appearance of the Purified Products. A gradual darkening in color occurrs with the rise in melting point and decreases in solubility of the solid acids separated by means of organic solvents. But one slight var- iation occurred in this generalization: those solid acids insoluble in the cold petroleum ether solution and having a melting point between 80-85 degrees were more distinctly orange in color than either the preceeding or following members. This characteristic became even more pronounced on repurifi cation. The lowest melting point of any solid obtained was 60 degrees and each succeeding sample gave a rise of from three to five degrees until that of six members wa,s taken. It was in these latter samples that the presence of mixtures could be observed, by the transition temperature occurring ■ - if . jH . I m\u ' - 15 - on application of heat. This became more pronounced on using those sa pies most easily precipitated in petroleum ether. The darker colored samples showed several transition points but no definite melting point. The phenol soluble extract had a melting point of 111 degrees and was the highest obtained. In every case there occurred a decided change at the melting temperature. The lightest colored products became dark and viscous did not return to their former color or powdered condition on cooling. At a few degrees above the melting point decomposition set in and the product began to increase in size. After the filtration of the acids the petroleum ether was allowed to evaporate and tne resulting precipitate was powder- ed. All of the lower melting solids were powdered with, con- siderable ease and had a tendency to adhere strongly to the spatula giving an easily compressed cake. Those having higher melting points possessed this property to a small extent there being a considerable decrease with increase of melting point and increase in hardness. The phenol soluble product wae very compact and was reduced to a powder only after considerable pressure had been applied. . * . ■ * ■ - 16 - Teste on Extracted Solid and Liquid Acids. Numerous tests were applied to the extracted acids to determine their chemical nature, among which were solubilities, ignition, combustion, nitrogen and sulphur determination, and condensation with hexame thylene tetramine. The solvents used were alcohol, turpentine, benzene, ether, chloroform, carbon disulphide, phenol alconolic KOH, and pyridine . All light colored acids precipitated by the cold petrol- eum ether were soluble to a greater or less extent in all solvents. Turpentine and alcohol showed the poorest solvent power. The light brown products were soluble in all reagents with the exception of turpentine a no. alcohol. The darA products showed a decided decrease in solubility in all solvents. The last extract was soluble in phenol and alcoholic KOH only. The pyridine test was made that the asphaltic character of the products might be observed. Samples were dissolved in pyridine and some water added. On being agitated a soapy foam was produced and this remained for sometime. The addit- ion of K SO4 caused the precipitation of the solute. The solubility tests in general show that all of the sub- stances are decidedly acidic in character, and that those lightest in color are more readily soluble in all reagents, and also that an increase in depth of color is accompnied by ♦ ♦ . . -17- a decrease in solubility. The ignition of the liquid acids showed no characteris- tics other than that they were of the aromatic type. Cn burning all of the solid acids, however, there was a decided change. Shortly after melting they began to increase decid- edly in volume and charred. This volume assumed remained constant until most of the carbon had been burned off. To obtain the percentage composition, the determinations of carbon hydrogen and oxygen were made in a combustion furnace, nitrogen was determined by the Kjeldahl method, ana sulphur as by the NagOg KCIO 3 fraction ation using id 1 • v v ( 30 ) a Parr explosion bomb. The following gives the results obtained using samples of the lightest and darkest acids: TABLE III Color of Samples C H C N S Dark (CSg ext. ) 78.6 4.3 13.6 1.5 1.11 Light. (ether ext.) 78.9 5.4 13.5 1.0 2 1.04 From the data thus obtained there was no indication that these acids were of very great difference in molecular structure. Due to the fact, hov/ever, that these are decided mixtures an accurate determination is impossible. To determine the position of the hydroxyl group in the molecule, ie, to ascertain if these are either cargoxylic " ' ' — - ■ ' -18- acids or alcohols, or of true phenolic nature, Redinanol condensations were made using hexamethyline tetraraine in alcoholic KOH solution, with the following results: TABLE IV Sample B .P. M.P. Result Residue liquid • V 200-10 light brown resin none liquid 210-20 light brown resin none li quid 220-30 brown resin none li quid 230- 60 dark brown resin none liquid 260-300 aark brown resin none li quid 300-340 black resin light yellow ppt. in water solid 65-85 none original solid 94-96 none original solid 104 none original crude tar acids black resin asphaltic powder. The foregoing table shows that there are two types of acidic substances isolated by the two methods used. Those separated by distillation are of phenolic type, and may contain a s mall amount of the lower melting solid acids which precipitate out on condensation of the liquid acids and those bearing decided acidic properties but not of phenolic type. The residue left from the crude tar acids after conden- ' . * . - 19 - sation was the same in color and character as that obtained by the extraction methods. In this way hexane thy lene tetra- mine may be considered as a satisfactory reagent for extract- ion of the higher acids. Tests on Gilsonite. Comparative tests on some asphaltene extracted from gilsonite were made, with a view to determine similarity to the solid acids obtained by extraction with one exception that there were no low melting products secured, every test applied was identical in its outcome to those applied to the acids, leading to the conclusion that these solid acids are in reality asphaltenes. (No combustion on asphaltenes from gilsonite was made). - . - - 20 - SUMMARY OF INVESTIGATION 1. Low temperature tar cannot be fractionated without a large amount of decomposition of the higher boiling consti- tuents giving a porous coke having a large volume. 2. Subsequent fractionation of the tar acids causes a decrease in yields of the higher boiling constituents with a corresponding increase of the low boiling fractions. 3. No satisfactory cutting temperatures can be select- ed for low temperature tar acids, as there is no decided increase in yield at any one temperature. 4. Tar acids may be classed as to liquid and solids: those soluble in common organic reagents and those insoluble in all save phenol and alkali; those possessing the hydroxyl connected to the ring (truly phenolic), and those of carboxylic or alcoholic natures; and those truly acidic and asphaltic . 5. Free carbon determination in low temperature tar cannot be made as in high temperature because there is found in low temperature tar an acidic substance insoluble in benzene and CSg. 6. The asphaltic bodies in low temperature tar decompose on heating giving an increase in the volume. 1. Freeman -21- BIBLIOGRAPHY Petroleum Times 3 , 543-4. 2. Fischer and Schneider Ges. Abhandl. der Kenntn. Kohle. 3. Schneider __3, 36-56. Ges. Abhandl. der. Kenntn. Kohle. 2 , 4. Schneider 133-44. Ges. Abhandl. der Kenntn Kohle. 2. 145-50. 5. Parr and Layng Mining and Met., 1920 Bo. 158 Sect. 4. 6 . Spielman Gas Journal 143, 65. . j 7. Lewes The Carbonization of Coal P.192. 8. Loco Git. 9. Maclaurin Gas Journal 139 , 143 • o i — 1 Lunge Coal Tar and Ammonia 1_, 577. 11. Fischer and Schneider Ges. Abhandl. der -^enntn. Kohle. 18. Ibid 3_, 200-12. 3 ., 1-38. 13. Benson and Canfield J. Ind. Eng. Chem. 12 , 443, 14. Cooper By Product Coking P. 138. 15. Lunge Coal Tar and Ammonia 1, 219-20. 15. Mulliken Identification of Pure Organic Compounds, 17. Baekland _1, 90-106. J. Ind. Eng. Chem. 1, 149-61. 18. Redman and Wei til J. Ind Eng. Chem., 6 _, 3-15. 19. Baekeland J. Ind Eng. Chem.,__5, 506-11. 20. Baekeland L. Ind. Eng. Chem.,_l, 545-9. 21. Sage J. Soc . Chem. Ind., ^30, 558. 22. Schne ider Brennstoff Chemie 1 , 70-72, and 80-85. - 22 - 23. Marcus son Z. Angew. Chem., 32, I 385-6. 24. Hubbard Dust Prevention Bull. Dept. Ag. , Ho. 34, P. 34. 25. Hubb ard J. Soc. Chem. Ind. 30, 201. 26. Robertson Chemistry of Coal P.21. 27. Church J. Ind Eng. Chem., 3 , 227-33* 28. Gluud Ges. Abhandl. derKenntn. Kohle. ,&2 , 236 56. 29. fischer Ges. Abhandl. der Kenntn. Kohle. , 2_, 22-35. -30 White Gas and Fuel Analysis P. 206