A STUDY OF METHODS FOR SEPARATION AND IDENTIFI- CATION OF COMPLEX AROMATIC HYDROCARBONS OBTAINED IN THE CATALYTIC DECOMPOSITION OF XYLENES BY JOHN DANIEL MALECKI THESIS FOR THE DEGREE OF BACHELOR OF SCIENCE IN CHEMICAL ENGINEERING COLLEGE OF LIBERAL ARTS AND SCIENCES UNIVERSITY OF ILLINOIS 1922 UNIVERSITY OF ILLINOIS January.. .2.6., 19122. THIS IS TO CERTIFY THAT THE THESIS PREPARED UNDER MY SUPERVISION BY John. ..Daniel. Male.cJci ENTITLED. .. A.. S.tudy...oi... Methods.. .ior .. Separation.. . and. ..Identifi,ca.tiQn.... of Complex Aromatic Hydrocarbons Obtained in the Catalytic Decpmpp s i t i on of .Xylene s IS APPROVED BY ME AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE DEGREE OF Bachelor-of •Science in .Chemical.. .Engineering Approved HEAD OF DEPARTMENT OF ' : ji 'it w i / V I l' < i *4 r V ' oa. f t . .j f o f T f o f f ; cr '7 f f / r < 1 1. '■'■'• t i-'t-if.'}. t%c.- ' / .t T .» -^r ■£ T 1-4 "V f'»'7 lo ao rsoTTroco i't ■ * ■ j**f ■ Acknowledgement. T . This investigation was carried on in the chemical laboratory of the University of Illinois, during the collegiate years 1920-1921. The problem was one arising directly from an extended investigation carried on by Dr. M. J. Bradley , and was carried out under the general supervision of Professor S.W.Parr. I am grateful to Professor Parr for his kindness and the inspiration he has given me in the course of the v/ork,and in no less degree to Dr. Bradley to whom I am indebted for the materials used, as well as a large part of the apparatus. His suggestions and assistance have likewise been invaluable to me in the investigation. \\W>fi\»'". H-TM ' ■ ' ^V< ‘ife'-’-; ‘ 'i, •', ! ■ '\ir ?■ V / ' -• ■ • y ' ■ 'y, ■ ;Vi j ■ ■ ,,', wv. ’■ \ < V; ,'■, » 0> ' i5' j :U ’ ♦'* - . . V , I ■ r . V A • ** V.’- •• ''•Ki0i "i , . 'J'V^ i’Xi'O.f S' '^ •■-•''■' ', ' . ■ - ' ■'• ,' ■■ >1 ' >•■ ■ 'ySB ■•■•■ ':,' ■ .774^3..^$ '■' ', \'j- ‘ . # „* '/ '■ . * r ‘ ''' ' <’' ’^,.1 ' - , • " ' ,G II < 1 ** ^>v. >c> ' I ‘ V * 4 ifPP ' “ 'T«. ' ; j •I-? i - '. ■ /. SCtftA!; '■'S^ ™ './..■ 4' :•■■ '•’V 'V r « ! ' :r *1,1 ■ ' ■ '-V V f ./ *■ " ' ■ '■ "''Sir - ■ ^ ru,jcTtr^>tiz Sh^ iii V • • , ' ■ ' , ' •* ■' ■.'■ : , ' ' ' ' 5'- ' " ' * I V ^ - W" w mm *4 mm^rn ■>/ '■ 1 * •# V ^ . „■ (■ ' ■ V ' * '• ' .'• >1 Ac i I > |f . . ; . „’■/"-■. j_ ' '‘'?i .I.:';.? v-v '■'•Y-.-v . 1 , - ,• i i « .f.' - w. -‘ii ‘ ^ ' ;'l* At /• .. *rf_ - ; .•' L' VMd '. Jy f JV ..V^iW J _ '5i n y x>/ i«* “J it *r $ s' . ■L , ■" '" . y'r,/.' ' . V'-C in.^b afA.;.' ^ .i (."■'. licnq •Xv/.-i ‘<.-C '-r- O'l. T c-t YV’- , y-- ' “ - " ... -"-•./ '* ' 1 , "f !- - ” :*•' V i'* , 'j‘,' c//.Y*;» t ■' -y; j. ' 4 Y ’ • ''* f '■' ■? -•.. f.r.i :^ ' :. '':j : .■ ■..,: • '/a Oi» . •' ■ . 5-'.-i • 5j •? .T.p •«<.» :.'.1 cxricj;-j'.«r7 ■'w>n- fsl^xiin .-v If i/ 'I y. •r< '■- V C' '.' ‘'i. 1,/ '-. W , i' V V ‘ ' -r-V-n #. 7. the Identity of the compound in question. Some investip;ator 3 offer as proof a melting point or boiling point, while others give some methods of isolation with complete identification data, but these are rare. The early investigators lacked some of the p.dvantages offered by the systematic methods of identification of organic compounds now in use, although these methods are still far from satisfactory in dealing with this type of mixtures. Former methods of ident- ification were probably as complete as circumstances would permit, but many erroneous assumptions were made which were later correct- ed by succeeding investigators. As an example,Berthelot, the pioneer investigator in this field, claimed to have obtained chrysene in his re searches, and it was later proven by Schmidt and Shultz that this was not so,a,nd that the compound referred to by Berthelot which melted af'about 200°" was a mixture of the methyl anthracenes, and not chrysene which has a melting point of 250°. Zanetti and Egloff '^distilled the products they obtained from the thermal decomposition of benzene with a Glinsky column, and merely stated that the fraction boiling from 250-275° contained diphenyl, and that the fraction distilling from 200-250® might be considered as napthalene. The higher boiling fractions they extracted with alcohol and obtained products melting at 86® and 196® which they called the m & p diphenyl benzenes and triphenylene respectively. Clark^ Ogives methods of seperation of anthracene, carbazol and phenanthrene in detail. Although nitrogen compounds are absent in the materials investigated in this Virork,both anthracene and phen- anthrene may be present so the article was studied in detail. In i i s 1 \ 8 . their process the solids are allowed to settle out, filtered from the accompanying liquids as far as possible, and then sub.iected to solvent methods. Coal tar naptha was used to remove phenantlirene and coal tar bases until a limiting purity is obtained, then acetone are and pyridene used to isolate the anthracene; these solvents dissolving out the two principal impurities left ,ph'enanthrene and carbazol , finally the anthracene was heated until it passed into solution, filtered and crystallized. The purity of the product was claimed to be sufficiently high for commercial purposes. The phenanthrene was recrystallized from gasoline and then from alcohol with bone black, and obtained as a pure white solid M.P.IDO®. Rittman, Byron and Egloss^^ arrived at three fundamental con- clusions in their investigations of the thermal reactions of hydrocarbons in the vapor phase; a) The course of the reaction is, in general in the direction of decrease in the size of the molecule when the degree of saturation is unchanged, b) Reaction may occur in the direction of dehydrogenation with either an increase or decrease in the size of the molecule, c) Reverse reactions are negligible. They distilled their products fractionally , and arbitrarily named each fraction according to its corresponding coal tar fraction, the solids being assumed to be napthalene and anthracene. An investigation on the formation of anthracene from benzene and ethylene gives a method for the quantitative determination of anthracene by oxidation of the entire anthracene fraction to anthra- quinone,and subsequently purifying the product. This is not necess- arily an index of the anthracene present as there are several closely related hydrocarbons besides anthracene Virhich yield anthra- i 9 . quinone on oxidation. A very complete Identification of compounds produced under conditions similar to those utilized in makinj^ the substances investigated in this problem is found in the report of the work of Cook and Chambers?^ They hydrolized their product and steam distilled, obtaining in the first fraction benzene, toluene, and xylene. The heavier products, which steam distilled less readily, were fouind to consist of an oil boiling between 280-290° which they proved to be unsymmetrical diphenyl- ethane having a boiling point of 286^ by oxidation to benzophenone and subsequent formation of the oxime M.P.140°. Still higher boiling compounds from 360° up, they recrystallized from alcohol and found to have a melting point of 181*^. This they oxidized to anthraquinone and assumed to be 9-10 dimethyl-anthracene-hydride. A fraction boiling from 290-300^ was oxidized in an acetic chromic acid solution to its oxida-tion product and subsequently converted to the oxime which melted at 163° indicating the presence of unsym- p-ditolyl-ethane B. P. 294-295°. They also obtained some brown needles which they were able to oxidize to anthraquinone and assumed to be 2-7 dimethyl-anthracene. D. T. Jones'^^^ gives no identifications on his products, but he states that he removed the methyl napthalenes from the oils by treatment with picric acid. He gives an interesting discussion on the formation of napthalene from phenyl-butylene and on the formation of phenanthrene from o-ditolyl. Bone and Coward‘S give no identifications in their article on decomposition of hydro carbons, and G-.W.McKee^ ^ uses a polarizing microscope in his work on the products of the decomposition of • • s 10 . benzene. C. Smith and W-Lewcock^"^ investigated only the presence of diphenyl in their problem. Haber^ in his researches on pyrogenetic reactions of aliph- atic hydrocarbons , does not mention that he obtained any of the higher degree aromatic compounds. Libermann and Berg^T in their researches in I 878 mention that they obtained anthracene, a yellow, butterlike mass which they oxidized to anthraquinone with chromic acid in acetic acid solution, and then converted the anthraquinone into alazarin. They obtained napthalene,but did not mention further identifications, although they claimed chrysene to be present with other aromatic compounds. Ostromisslenski and Burshanadse^ ^ evidently carried out their hydrocarbon decomp- osition with a view of obtaining hydrogen for balloons, and no compounds were mentioned. Smith and Shultz^'^ gave a few valuable hints in their article relative to their researches. They carried on benzene decompositior by methods similar to those employed by Berthelot, and they first distilled fractionally their high boiling compounds. The fraction boiling under 100*^ they considered as benzene, and they noted that the thermometer ran right up to the point where diphenyl began distilling. They collected the diphenyl and identified it. Cuts were then made at 250^,300®, and finally at 360%fter ivhich they examined the fractions separately. That coming over above 300*^ was a bright yellow solid, and the residue in the flask was a brownish powder. They identified their compounds by physical con- stants , solubilities, reactions with picric acid, likewise placing great reliance on the crystalline structure of the compounds under consideration. 11 . COI-CPOUND CRYST.STR. M.P. B.P. ALC.SOL. PI CRATE diphenyl benzene leaflets 205° 383° difficultly none ^ diphenyl " 70.5° 254° soluble ^ " isodiphenylbenzyl needles 85° 363° " " triphenylene " 196*^ 36o°up slightly yes benzerythrene plates 307-8° High insol. none Benzerythrene is in the residue after extracting with alcohol. Trlphenyle^e was described as a yellow compound, and v/as mentioned to be identical with the compound that Berthelot designated as chrysene. The oils proved unmanagj,ble,but they identified a compound with the formula(C 5 H 5 GH 2 in the high boiling oils. » 4. Outline of Present Investl-g^atlon. The fact that in the distill- ation of low temperature tars, there is a relatively large fraction ' which is of little or no commercial import^ance. This fraction consists principally of the three isomeric xylenes, which have too high 'h boiling point for a motor fuel^and are of no importance otherwise. If this fraction could be converted into other more useful substances by thermal decomposition, use could be made of large quantities of this oil which now has no practical applic- ation. Following this line of* reasoning, the vapors of the mixed xylenes were passed through a furnace containing various catalysts, and under various conditions of temperature and pressure, which are exhaustively described in the thesis of Dr. M.J. Bradley^ A wide variety of substances were produced, but we are concerned only with the substa.nces formed which have a higher boiling point I i \ t • } \ 12 . than the origional xylenes, or roughly , above 150°, the other products condensible and noncondensible being identified in the course of Dr. Bradleys work. In this Investigation the products obtained were first fract- ionally distilled; cuts being made at about every twenty degrees. These fractions were then extracted with solvent Sj or otherwise treated by methods adapted from the scant literature on the subject and by other methods which in the course of the invostiga,tion seemed to give promise of good results. In the distillation of the hydrocarbons they came over, in general, in the oiTder of increasing specific gravity; the very light oils distilling first, the heavier oils coming next with increasing amounts of solids in suspension, and finally the very heavy , syrupy oils and large amounts of solids are driven over. These fractions grade gradually one into another, and no separation is possible by fractionation , except in a general way. Unlike petroleum distillation products , these substances cannot be used as distilled, but they must first be subjected to purification, as it is the pure compoimds which are important in the case of aromatic hydrocarbons. Aliphatic hydrocarbons do not differ markedly from those in their immediate neighborhood, and for all practical purposes fractionation effects ample separation. It ca,n easily bet seen that the difficulties be equally great in separating a-methyl napthalene from its accompanying oils as separating pure hexane from gasoline. Many of these aromatic hydrocarbons are used pure in the manufacture of various types of dyes, so a commercial means of separating them from these synthetic mixtures obtained from I < 13 . coal tar fractions v^rould be of considerable value to the dye industry. Separation methods which are worthy of trial are steam dist- illation, centrifuging, fractionation, separation by specific gravity difference, and separation by behavior towards reagents and solvents. A sample of the total run was examined for the elements, and another portion was subjected to the class reactions for organic compounds . Steam distillation was employed continually , as was distillation under diminished pressure. Identification of a compound after isolation was effected by determining its physical constants and appearance, looking up in the literature the compound corresponding to the one in question most closely, and finally making a derivative of the compound and obtaining its physical constants and comparing them to the literature values. Melting points are the most common physical constants employed , although boiling points and specific gravities are used to some extent. Only rough approximations were made as to quantitative results, / as in many cases the runs were combined in groups to form a single sample, thus making quantitative results of little value. After the preliminary tests which were applied to the crude tar, the bulk of the tar was distilled fractionally in an elect- rically heated distilling flask, which is fully explained in detail , together with the manner of making same, in the thesis 20 prepared by Charleton . This general separation preceded the special methods which were then applied separately to the liquids and solids as described fully later. The last traces of xylene f ) I I I V » r i i: t; I f 14 . were removed by steam distillation and subsequent refractionation. The solids were treated with solvents except in a few exceptional cases. The napthalene runs were steam distilled; the napthalene thus recovered was fractionated, and the products not volatile v/ith steam were examined in detail. The data on these runs will be grouped together later. No combustions were considered necessary, as all of the compounds are known and can be found in the literature, so the obtaining of the two physical constants and the making of a derivative was considered sufficient identification. \ 15 . II Experimental Work. 1 . Elementary a nalysis of xylene and tars for elements. A pre- liminary ignition of the xylene and the resultant crude tar on platinum foil gave no Inorganic residue in either case, although there was a heavy deposit of carbon from the tar; so it was assumed that inorganic matter was entirely absent. The test for the organic elements was made in the usual manner. A piece of clean metallic sodium was placed in a two inch test tube, heated, and several drops of the substance in- vestigated, dropped in, care being taken to have it react with the sodium vapors. The tube was then ignited to a red heat and plunged into ten cubic centimeters of cold distilled water. / This is then boiled, filtered, the solution obtained being ready for the tests for sulphur, nitrogen, and the halogens. In testing for sulphur, a few cubic centimeters of the solu- tion was acidified slightly with acetic acid, and a few drops of lead acetate were added. No sulphur was found in the xylene, but in the tar a very small amount was detected. A second trial confirmed this. It is hard to account for the appearance of sulphur, but in all probability it was a.bsorbed from the pumice stone and the other catalysts used. In testing for nitrogen a few cubic centimeters of the stock solution are boiled with five drops of FeSO^and one drop of FeCl^ for two minutes. It is then cooled, and just acidified v/ith dilute HCl. The xylene gave a clear yellow solution on acidification, but the tar sample gave' a light green solution which on filtering through a hard filter left a slight blue 16 . stain on the paper, Indicatinp; a trace of nitrogen. This also was confirmed. For halogens the solution is acidified with nitric acid, boil- ed a few minutes , cooled , and a few drops of silver nitrate added. The solution remained clear in both cases indicating a total absence of halogens. 2. Fractional distillation. During fractionation , cuts were made at much more frequent intervals than would be maide on a commer- cial scale for better observation of the products obtained. Many of these fractions were combined again later, when it was found that they v;ere of the same composition. The distillations were carried out either in an electrically heated distilling flask or in a flask immersed in a sand bath, which latter method, though being cruder, y/as fully as efficient and had all of the advantages of the electrical outfit. An air condenser was used throughout , as the substances were all so heavy that they condensed almost immediately in the side tube of the distilling flask. As far as could be observed there was no decomposition due to cracking until the temperature was well over 300°, after which there was some indication that cracking occurred. In the mixed runs the first fraction was collected between the temperatures 150° and 210°. This was a rather large fraction, but was found later to consist largely of xylene and a very small quantity of a higher boiling oil. The next fraction was taken from 211° to 240^ so that any napthalene which might be present would be included, after that regular cuts were made at 260° , 280° , 300° , 330° , 360° , and until nothing further came over. - V .• V fvJS it ■ ■ :d ^'>-•■'•..■1 /-V ■ T- . . • . , . » ♦; ■ .,; _ ^ ' '-iw f.\'. •■ ■ Tfj ^ '-I A - - ...Tf .* ..- v.c"i A.'^b '.o, TT^-r.j^ ' ■■ * ■ 'V./v-r •.*, ■ . o',- ■„' > ,: ■ ’.iV ' ■ <1 vn ' • . . u ;• ■A’Tr»* i'jx’rr . .'tA i,.5ju’''iin I u'\ ' -t • > 4 «. • >- *1 I t'PAm' { . I :' cr-’ ‘'. ac X * <-< ► I /i- rs\i" 5 ^ r* . • s ',.'■'••• ; •♦ ' ■ 4 . . tt, . ^ ^ 1 - W 0. . . A./ - ; .' A- 4- ^ •- ■ • 'v- 5 i»/4 .;.i: 1 .■; w,-. WXA^a,' : >’ :. : ir ■ . ^. .e t'i x.v» '^?.; ' 'o a 'iw J’..'J 1 .- ■ '. -' i '• t <- : ' 1 '■’If. . . t i .. . j\ ■< ■ ■ iiy I -; V ' -^: t^: :v>y Or, ■ T .' ' ■ i'. ■■ V. ■;/..; v: . / r,. ^ I c <.■’'■ i?.t' n!- ; ’•;p ^ ,. -■ >’ ■ ' • . . • . ,. ■ -n ; ':.t f»rft;; at,-": ^ ■' • ■ - c ' 'r-:". -J ... . ■'•’•' ^}gi* ■• f. 5i o ' t '-i ..a-' . r.r tS-C ,...: / ' ' ■/’: ...^ *'■ . . ^ ■ •■'/. '<• / A, AW GAw v'r.G V- ■ c • . V - •. ax ^.' : . • ■' ■ fV, . ' ' f ■ > .,: ..i- -;•<.; -. '■ T *' 1^’ . ".•!T . vf-:.v , .‘' /- . A., -7^ ■fK ';.xli. ^ A ,\' • ix' 'v.;. ' 1 - ■ ', H-F -'.A',-,,-- ' : ; ' G "■, ' 1 1? .. '. ;. ■■ Li^fi V •■< :/ ' ' .. 4 ^ i ,..* ♦ i \i . . !• -'ff'.i ,i<^r • /.Iv G *-) !f?. ( /■ ' 0 ' C r.7)' A .:a‘ r. .va^ • > ' irtf.i:- I , . ' , . ' a , v-C. 17 . The two ethylene runs were fractioned still more carefully; as there was a greater volume of them, and a greater variety of compounds were expected^ The distillation was carried out in the electric stills, cuts being made at 150°, 175°, 200°, 225°, 240°, 255°, 300°, 550°, 400°, 500°, and until nothing further came over. The residual coke was saved for future reference in all cases. 3. Approximate determination of physical constants. The boiling points were, of course, noted in the course of the distillation. Melting points on the solids produced , taken after there solids were pressed free of the oil, proved to be of little value even approximately , for the oil which remained in them lowered the true melting point from twenty to sixty degrees as was later found out. Specific gravities of the oils in the ethylene series, which were considered representative of all of the substances obtained, were next acertained by means of an accurately calibr- ated pyknoraeter(l c.c.). These determinations were carried out under a constant temperature of 15.5°C,and the weight of the tube was checked three times during the course of the determination. The specific gravities found v/ere as follows. Origional xylene 0.8664 Tar (as obtained) I.O 808 Fraction 145-175° O. 89 II 175-200° Fraction 0.9057 200 - 225 ° ” 0.9605 225-240° " 0.9888 240-255° ” 1.0087 255 - 300 ° " 1.0282 18 . 300-350° Fraction I.0625 I 350-400° " 1.1082 Some of these liquids undoubtedly had solids in solution, thus introducing an error in the determinations, but this error is slight, as the solids involved have practically the same specific gravity, when in the liquid state, as the liquids with which they distill. The boiling range of the original xylene was the next thing taken into consideration. The specifications were that it should boil between 140° and 1 43°, actually the range was as follows. A common distilling flask and condenser were used, the bottom of the flask being well covered with glass beads to insure smooth boiling, one hundred cubic centimeters of xylene were used. 1 st drop -- 125° 1 St 10 c. 2nd 10 c . c. -- 138-139° 3rd 10 c. c. -- 139-139.5° 4 th 10 c. c. — I39.5-I400 5 th 10 c. c. -- 140-140.5° 6 th 10 c. c. — 140.5-140.5° 7th 10 c. c. -- 140.5-141° 8th 10 c. c. •- 141-141.5° 9 th 10 c. c. - 141.5-142° 10th 10 c. c. .. 142-144° No residue. This distillation was carried out more to satisfy all doubt as to the limiting values of the boiling range of the origlonal. 1 19 . so that it could confidently assumed that all of the substances obtained were actually synthesized and not contained in the original . This was specially true of the small amounts of substance boiling directly above xylene which were obtained. 4. Senaral Appearance of Compounds. The original oil was a moderately thick, black, fluid tar much resembling crude tar oil in appearance. The fractions from xylene to 200® were almost water white and had an odor like xylene. From 200®to 250® the oils were just faintly yellow, but their viscosity was greater and they smelled distinctly of napthalene( after long standing small amounts of napthalene settled out). From 250*^ to 300® the oils were a reddish yellow by transmitted light and light green by reflected light, in other words they were fluorescent to a slight extent. From 300®to 350® the oils were increasingly viscous and strongly fluorescent, the solids which occurred here were a bright golden yellow in color. The fraction boiling from 350® to 400® was a very heavy, dark green oil, and the solids were still the bright yellow spoken of before. The final dist- illation products were a thick paste, not unlike vaseline, which could not be crystallized even by an ice-salt bath, but became almost glassy in hardness. Some reddish solids were present in small quantities in this oil, and the final distillation product was a red solid merging to a green solid, and finally a mass of what proved to be free carbon with small amounts of the latter solids, which came over just before coking occurred. 5. Determination of Behavior Toward Solvents. First the fraction boiling from 300-350® was taken for separation as it was a large fraction and obviously consisted of at least two substances. 20 . A method of hydrocarbon separation recommended in the literature, 1 ft was by fractional recrystallization from alcohol and it was thought that it might be of service here. This straight recryst- allization from alcohol , although it finally yielded a substance with a definite melting point (after twelve recrystallizations) was finally abandoned as worthless, as the amount of solvent used was altogether out of proportion to the results obtained, and as it was later proved, the same substance could be obtained in a purer state by a much simpler process. This solid had a melting point of t 97^, and on oxidation yielded a gummy mass which was impossible to obtain in crystalline form. A single recrystallization from hexane brought the melting point up to 202° where it remained. The alcohol, on attempting to recover it from the mother liquors was found to be unfit for further use, owing to a small amount of oil, which was volatile in alcohol vapor, being present. Diluting the alcohol gave a milky emulsion, but efforts to recover this oil by fractioning the alcohol proved fruitless due to the volatile nature of the oil and the small amounts of it present. Abandoning special methods for the moment, a portion of the origional run was tested with certain solvents to effect a group separation if there were more than one group, or if only one, to determine the position of this group. There were no water soluble constituents, so phenols are absent. Except for particles of free carbon, the mass was entirely soluble in ether . Dilute hydrochloric acid had no solvent action, and dilute potassium hydroxide was likewise without effect. This definitely eliminated phenolic substances of all kinds. Sulphuric acid had no action 21 . 3 in the cold, but dissolved the whole mass on heating. This last fact indicates the presence of aromatic or unsaturated aliphatic hydrocarbons. It is thus seen that the solubility reactions as generally employed gave nothing but negative information. It eliminated the possibility of acidic, basic, phenolic , and in fact everything but aromatic hydrocarbons , unsaturated aliphatic hydrocarbons, and halogen derivatives of these. The preliminary analysis showed an absence of halogens so the search v/as confined to the first two types of substances mentioned exclusively. To be doubly sure that there were not small amounts of sat- urated aliphatic hydrocarbons which had escaped detection before, three ten cubic centimeter portions of the run were now tested for saturated aliphatlcs. One of these portions was the original tar oil, another was from the light oils, and the third from the heavy oils. They were placed in three Babcock testing bottles and thirty cubic centimeters of concentrated sulphuric acid v/as added to each. They were then heated on the steam bath at a temp- erature of 100® for one hour. After cooling somewhat, the bottles were filled with the same concentrated acid up to the graduated neck and centrifuged for five minutes. If any saturated aliphatlcs were present they would appear in the necks of the bottles and could be read off directly in fractions of a centimeter. Careful examination of the liquid in the necks of the three bottles showed no oily layer, and saturated aliphatic hydrocarbons were thus shown to be completely absent. Alcohol extraction in the cold appeared to effect a fairly good separation of the oils from the solids in the high boiling 22 . mixtures, and leave a mass of greenish yellow crystals of a fair degree of purity. On recovering the alcohol by distillation, it was noticed that on chilling the concentrated liquors at inter- vals, a dark brown, pitchy mass separated out. A considerable amount of this substance was obtained, and it was noticed that the same yellow crystals came out of solution on standing. Distilling this oil under diminished pressure gave a clear, heavy oil, which likewise gave a batch of yellow crystals on standing. The net result of these attempts seemed to show that separation of the liquids from the solids by fractional precipitation from alcohol was possible, but incomplete and tedious. The combined residues were then tested with carbon tetrachlor- ide. It was entirely soluble, and no separation was effected. Hexane, or petroleum ether was next tried. This would remove phenanthrene if any of this substance were present. The Hexane dissolved out all of the liquids and left a mass of bright green crystals. These were later purified still more by recrystallizing from acetone, and identified. On recovering the hexane the same pitchy mass was left as a residue. In recrystallizing the solids present in the materials studied it was Invariably found that , regardless of the solvent used, the first batch of crystals were greenish in color, and the second batch pure white, or nearly so. This seemed to indicate the pres- ence of a persistent impurity whose relative solubility in a majority of solvents was less than the accompanying substances. The amount of this substance was never great enough to change the melting point of the compound by an amount that could be detected with the thermometer used, as two samples obtained from the same 23 . solution, one colored and the other colorless, both had exactly the same melting point. In spite of this it was assumed that two different substances were present, and this actually proved to be the case in one instance where the a and b methyl anthracenes were found in intimate mixture. Many other solvents were tried on small samples, benzene, xylene, acetone, acetic acid, absolute alcohol , carbon disulfide, chloroform and ether among these, but mostly without re suit, or with results which were no better than those already tried. A few combinations met with success and will be mentioned again in connection with identification. The results from the applic- ation of general solvent methods were negative, as far as isolat- ing any pure compounds were concerned. The group solubilities and elementary analysis eliminate everything but aromatic hydrocarbons and unsaturated aliphatic hydrocarbons. Testing a portion of the origional run with neutral potassium permanganate in dilute alkali solution gives a permanent green solution with no reduction of the permanganate. This eliminates the unsaturated aliphatic compounds, so the substances may be assumed to be purely aromatic hydro carbons, and subsequent treatment should be guided, to a great extent, by experience and observation during the course of the investigation. Separation of liquids from solids. After fractional distill- ation, the fractions were allowed to stand for some time; so that an equilibreum might be reached between the suspended solids and the liquids. It was found that if filtration was attempted soon after distillation , solids continued to settle out of the liquid afterwards, and subsequent filtration was necessary. The .V ;,)^l ir ^ .• •i'- !.' .■ 'j • '''■/ c .••r, '.R ■ : --i ‘.V ri ' ■■ .'v* KSi I' J - » -■• ,oi‘ ^.... .... 1 1 . ' ■ --f Ur-:- ‘-V^rr ■' i-.’. : •V: c.f .-i ••': r/* ' r>- ‘'J I v». ! : ' ‘ v’r. liii’ij'i ; vf..-...' •^.;- -i. ’ /:v 3 •. _ '. ..V . r .‘.i toacc- , -■u-'il -iv";.’ ...U • ’V . : ^ Ritzier - :. ::■• '*•.•• ■'•?• -t-'C*‘v •>f. *'• ■ "■ ' • •i r :,.■•/ x;:'^ AO V o-,. . • '> ■•■' ■ i ■‘.■- ^ 'i.-" •.oXX"; ’ '1 ' v'''X ■•; A/ ‘v: - .v “V- •;'^£ .Mo ir ':^;^'*.''^o.{ ■I.:’ V fc|f- < . .' |r- ; . J ♦ r« ' ** ‘ ' i " \} ’ •’ ' rtt iS Jrr^CjiXr ‘ ™ ’ 1 . 7 ' £ 00 ^ .i:- tw ,. t .-rXU‘ •■r-rj?V .-; / * « ;•;■■ tel' \ vJ » ,L j^I. t - ■ ' ^ V'I'* ‘T.>\ &c. ..J v.Oi i ,'. . 0 '/: viOti >■•. * ■ .X ^ -' ~ ' ’ ■ ■ . >•".. . X# /I'X 1 t' :. r."‘ '”S'lw 'l' , 'r»0.OJ.'\v ■' .'* . Ty i-i fi'i V 'r,j '■X .7 reXvJ-OY*. • ' <' ^ "i- '■* CJvi7'' -n ^ v.X , v c:* • « ■ :X ‘•'{J" 'JO X X~U i-v . fi.« r. Y.,-.;.t JV- I < jc: - '■ •.*■ — i . jsit . -j- - ... . •;• . .1 '•i’--.'* • ^r*' Uoi:’: • Tj .,:> rjiff'.-.' ..tv-* ;.■; As-o i-f • ; v” . -i, v‘ .if ^ f 'M .Uw*:* Y.r ■•!. ”< i . »r"d.i OiY A. i'O JriO'Jl ’fj/i- ;V*'. t ^ • I n ^ k. ^ i, Oi!0i • i* ■■■" * t i -'-ri '■;. ■ 3 r. -'t> - ‘ - - -.\ • jr sr.'JtMU''’: W‘ - 24 . crystals were larger on standing, and contained less occluded liquid. The scheme followed was to allow the fractions to stand for a minimum of two weeks after distillation, and then to filter under vacuum at room temperature, and Subjecting the liquids obtained to a freezing mixture of ice and salt for several hours and repeating the filtration. This was done for all of the fract- ions in which the liquids were capable of passing through common filter paper at the vacuum available. Melting points on the crude substances first filtered off, and those subsequently frozen out and filtered off, showed considerable variation, the latter being usually lower than the former. Recrystallization showed that they were the same substance, and that the initial apparent diff- erence was due to the varying amounts of occluded oils. The products of the fractional distillation were thus roughly divided into two parts, the liquids and the solids, which, in general, agreed quite closely in every way to similar fractions obtained in the process of coal tar distillation. 7 • Isolation of Compounds Possible by Solubility Behavior. There are several compounds v^hose presence was suspected, and later proven, that are capable of removal by solubility methods. These separations by solvents follow conventional lines, although there are a few slight diversions followed which seemed to give good results. The isolation of anthracene by solubility difference in coal tar naptha and acetone was followed with success. The solids filtered from the fractions boiling from 290° to 360° were extracted in the cold v/ith hexane, then with hot hexane after which it was filtered and washed with the same solvent. The 25. solids were dissolved in hot acetone, crystallized out, and filter- ed. These crystals were recrystallized three times from benzene, after which the melting point v/as constant at 215°. It was pure white with a light green tinge by reflected light. It was identi- fied later by oxidation to anthraquinone. Phenanthrene, supposedly , is in the hexane mother liquors, and by concentration it may be obtained in an impure form. This was done, and the resulting mass dissolved in alcohol. This solution was concentrated and the first and second batch of crystals obtained discarded. Finally it was evaporated down to small volume and chilled. A mass crystallized out,ancj^had a melting point of 98 °. Recrystallization from benzene raised this to 99 ^, it was a white solid without fluorescence. Pyrene, if present, is in the solids distilling above 360°, and may be separated from these by extraction with carbon disulf- ide. The carbon disulfide solution is evaporated to dryness, the residue dissolved in hot alcohol and a hot alcoholic solution of picric acid is added. The picrate fomed is repeatedly recryst- allized from alcohol and finally decomposed with ammonia. The resulting compound is pyrene. The identification was somewhat difficult, but was finally accomplished successfully. Chrysene is left behind on extraction with carbon disulfide, and it may be crystallized from hot glacial acetic acid. This was done to the black particles left by the carbon disulfide and a very small amount of very green crystals were obtained which o melted at 250 exactly, but further identification was impossible due to the small amount obtained. The large quantities of the solids which were intermediate ^ ^ i f4v V t '. , * .rM U^vli^ ,b ‘♦'ITi •'’ J J'^/'.J . ;t. ♦'(Av; .. „ J (i-i ' . o«c <»• ' ' v; ' .;> : -t ' •/;!} 3fi , ( ‘Or /. r * * t ti ^ < - -» » ^: • W. -' .A i.. j -■ *', ‘'’fri" ' :. ^., , .. .:r n * ■ • •. r.;il -f ■. •'< ■.t Lv, e.-a. *. f . ‘ - •. ;: .i*' ;;.• .'.'tlo'' . 'L '■ •: t .>toc8 C I r'.'ii '«■*'*■ • • -■: • ■ ‘■■•i . , J^IS' <■' ■M 1 ■ I- .-i- < . .<5. >\ . . .' ./ X a 5 \"v 7 t **^i • *u * '■ ? ■'•< S; r% ■• ‘■' "■ *’ •?.♦ ♦‘.1 ■ ' ^ T, W r» *j • < ',M ^ ' V !-''** ’ ' ' ' f. « ; i . , \'''l , V vV J i i.t ? iJir : ■•: 1.' '* ' ■ ' . '/ J 'V^'Lh r. . - '^v i'- f j'/‘-; _ '•;•■ ‘» "' ‘ i.i . . • .'■ . .on'V li) < s ' i ^Lt . ' ^ j ■' ** •; •; i i>(. I* lUt- ).:■ j ■*;. .' *:> j;.iCMa -■* S ' . •v'-: ‘ .I 7 C , ,bj^-.,Ji V ilJpA* ■ ’I *, ,; . ■•<••; -ra tV'-jf n vz-Uait^ ./;. ■ .» -fCi- . • ■. .=' .* • -t’;. /.;i6^A risalf n ■ ■" ■ :■ . :. r- v-'j. v ’ - • ; . , . ' ; .;.■, ' J >' -> y't:/ (irL, i‘*\C * V'il''!: • ' ' ■• *> CT.C'/i :r^- : .•s. i-?-: . ' Oi. : r*:-. : c ■ ■ •■ f.r r ' •v ; ,«■,.' » V* ‘ 4. . . ^ ■■ ’ ' ■* ' i ' 1 ■,-T< -J'T&i'.' *'• ' * 1 ▼ , • * • • « • ' V P V <• V’ ■• i’.' - j ; 1 iOi!- . .•.: . ‘ ' ;ca ;.5 1 ret . A. - ' . « ■ Vi C t* i ^ tj . t0>~ J i •• * V * ^ V* ; 4 i' ' '' t.li 26 . in solubility between anthracene and phenanthrene were next examined and recrystallized from acetone. The first batch of crystals were discarded to remove the last traces of anthracene, and the second batch taken and recrystallized from alcohol until a constant melting point was reached at 202°. This same substance was obtained from other solutions in considerable quantities. It wa,s noticed that the substance seemed to have a transition point at about 110°, and a considerable portion of it sublimed to the sides of the tube. A survey of the literature left little doubt that this was b-methyl anthracene , and the presence of this hydrocarbon was fully confirmed later. 8. Ste am distillation of liquids and solids. In the scheme of separation the substances were first distilled, and then steam distilled. The fractions were distilled in a current of steam in their regular order, the oils coming over being water white, and of a specific gravity noticably less than before steam distillat- ion. This was due to a high boiling residue which was relatively non-volatile in steam which remained behind in the flask. In the lowest boiling f raction^which origionally had a yellowish color^ this high boiling constituent was present to some extent, and an increasingly great amount of it collected as the distillation progressed. This was of considerable interest, and as it was noticed, a good separation of the heavy non-volatile liquids was effected in this manner, which was possible in no other way, as they were dispersed in considerable quantities throughout the whole range of the preliminary distillation products. The oils thus obtained were carefully dried and fractioned again. This time they distilled over practically colorless, and many of the ‘VO f--- . 'Ti*',- '■'■'’tV ■;i‘ ‘j’X«|k»i f rj ■ 1 ^ / . < *;. . i' . ', ■ . .'' n ■ ■ y,. ^ V - • '■■•.» • fK »■.* ' # . . . - • A ■* , . , . ■ • . • • >^ . V '■; :m ./.V ^ ■ .• ■;i^; >■:: j c :/i.rt^-.\. . : • • ^ . r. , , . *. ; i ‘ j •.•-;.} 'k :'.u '•■ :y aJ • ■.‘ri4';;idi/n ■ i.'* ■ ■ ' : '■•■y ^ . J t r^>T;‘:rr.V.v, ; • » ■ ' • ' ■ J ! ■ V ' ’ ■yr.-'j - • . V •..' 5k Ilf '•• -■■ ■ J .'T ,)5. -- : ’ /Mr?: f vj:9; ■/ ^ , ':-ii.J-j.^ ; V. ' .. ■ ' »'■ T, filijki-:’ ■ 1;.. '- ■ r, i 1 .. : j ■- <■-!** . ... . .! ' ifmt' -7 S '■ -VI. » f - .'V> •' r tJx - -S'..' * •T.:v‘. vl'V '>••.* ;*? -7 V',;. .■' ' r'- \ J / - L > ;Y *♦»' il-V rr • 'fV -4 r *. b:- • '' A , •* / ■* .* V -' ^ t ■ \>v V ‘.••Kr,t: ^ t ? ) f > - . ■‘V.iJ ' ■ / ' ■; '•‘.'■•i'.' •• V ■'■C ’ . , :‘ ' ■- ''<' •.';:( o: /V/; •w J r^\: r\: w • 'fiti i. ;■*■: D".r. Ji -:jjf ■ c ■ fv^'.cD' ^ ..' * •* ' fif - O X 1. 1-V - *i .-i ■» I - 7 -:■ \ .. Vo- V x.F.?:': :■•, * •> . < "-• ,vV^.‘iog^. j- 4 .; :• - . r. .' . ■ ' • £.•• , >• '.i ’ /J . V -V- . ;r if*'..-’lO o;- ir- •'.•■' • r* i -V'* :.- v-ivcrt, • J ' rr ••■.*• •■■ » .• •?;.•»■ ir^- :; » ■’ f i <•. ,07 v‘ 3 ^i TvViVC XH'CS • , 'll ' ■•’ 'j-V'.r: ;' f ■ ' .'■^, ‘ I* •. .j' I. IL -.■• ‘ >■ , • ' J Cfl ■ ..C '•■ ,' ! ■ ^; ':C. '':j ' ’ ,,r.':o ’"'li'- ••'•.at ■•.•>• ..•;rrV, ' .: ; rv ■ 1 , V’-*5 Vo o -'*■' r., -V ! ■(.; >'. 'f :■*■' ••n*# '.: , .••!'/ 'j > •• : 'v ; ■.: ';.o- \ 7 . - b f'aLXCJ Jill *■' .., V.. .,f ‘ .' ■ *. ; • * ' V ■ , ■ 'iv' 'Jj I . rfv J V 4 :-.«K • ' c; y#'* Vx' • •» •<■'-■. ■ It' a-?'- T. ; •/> S' ' ''.c jrtG'ii'.r ^ *■ •.* ^ * V ' L'> J.^; i-' . ;/ . ' '•V • . :;- ; ' -• J ■>:... ?} ■■ \ \ f ■•-.':■ . 'i -'j r',t- ./. ->.i '.: r. . ‘X '•. r;c X ' .••>/' i 1 i •%. _ s • /■- u .,,i ,^:.v.X. ' .>Txta rj; :-;:v;oul» 3 ^ r‘ ^ { " ^ ,*?.u . ;'AT . . F V ■>/, ,>v.' V ;■ 'V • . 1 .:^ ;- ? ::|.3 ' --f •■ ./ >’ ;i f n ‘ , . t ' \ ."S'... .' T- Iw 'j/, '•> .f ■• ■^: v^lT:v .1'- o ,** ■ ' ' ’ ' V! i ;•• t . . r^,\i '..MMT . . ( -C V., f . - ^ . .. ■Ui.-^ I / ‘ * •. * I ' nU-: .X ' , •• t ^ T. •*. t .. V ■ • ■• • •• - '- .*• u^-v.. ■■:••'* 'JC' nrxT. . o, -,; : 0 '* UmO '/X .U/i.X'C'iq > » *tf <* • . ,, , ■ ; V f '(-Vr .',. , • .* ;-r.*i Xr‘ ■: 0 .'XJ-iO* U»*'- i ’yi;/.-. .c-u.,. --j;*,' 1 ’: if: . ''"j. '■ 'i! 33 dilute potassium hydroxide , and gave the transient red color, disappearing on shaking. On subjecting the oils boiling from 251°to 271^^ to a freezing mixture, a mass of white crystals appeared, which were filtered quickly and clay plated. This was diphenyl, (C 5 H 5 ) 2 , M.P. 70°. A single recrystallization from a few c.c. of alcohol gave a product which melted sharply at 70® and was practically odorless. Bromination in the presence of iron was tried, but no derivative- could be separated. A small amount was dissolved in carbon disul- phide and bromine added in the cold. Concentration of this solution gave a brown solid which had a melting point of 302-5® and it was assumed to be the p-brom derivative of diphenyl, M. P.310®. The a and b methyl napthalenes,C^^H^CH^ B. P. 1 40- 1 42°, 1 41 - 143°, were next identified. In the oils which boiled from 240®to 250® it was noticed that on placing in an ice salt solution the oil congealed to an opaque, white mass, which, on attempting to filter, dissolved again into the accompanying oil and passed through the filter. This occurred repeatedly despite all pre- cautions to have both the filter and solution as cold as possible. The slightest current of v/arm air dissolved the substance. The oil was ref ractioned,and only the fraction boiling between 240® and 245® was taken for consideration, as this was assumed to contain the major portion of the a and b methyl napthalenes suspected to be present. B-methyl napthalene melts at 37 ^ while a-methyl napthalene does not solidify until -22° is reached, but their solubilities one in the other are ao complete that this difference in melting point could not be utilized in separating :V 1 ) , : ■ \J., 1 , . -r ' ' '£• ■' . ’>'. j ..> rt.'^x \ S i ' s.r»-. ’V r.r^. JB.A'VI .: • • (■...• 1 “• '. ‘"i.v • {;... .,. 2» ^ • ‘. ‘KV : > I ;(.■ . . .. .'vu- •.'C '■• ’ji « ..- vT.: l' -lO -•4 l'^D''-i*. 34 . these substances. Their boiling points are practically identical, so they cannot be fractioned from each other. The picrates of each of these substances melt at the same point , 1 15* 5^ , and a picrate made of the mixture was seen to melt somewhat below this temperature, 1 12-1 13®. The two oils have the same specific (3;ravity ; 1 . 001 ,and a specific s^ravlty determination made on the mixed oil gave exactly this value at 15*5®Centigrade. The two isomers were then assumed to be present in a mixture which was inseparable by ordinary means. In the middle fractions of oil, the fractions boiling from 270® to 280° was suspected to consist largely of the mixed di- tolyls ,CH3CgH^-C5H^CH^,275“280® . These substances all oxidise to isophthalic acid, so their presence could be collectively proven by this oxidation. It was not so simply carried out, since these oils seem to resist oxidation by chromic acid in glacial acetic acid to a remarkable degree, but after prolonged boiling under reflux, a product was obtained, which on recrystallization from benzene gave a melting point of 295®. This was soluble in hot alkali and was assumed to be the isophthalic acid derivative. The remainder of the oils were assumed to be a mixture of the dimethyl napthalenes, and careful fractionation gave a portion boiling from 262-264® , which gave a picrate in alcoholic solution consisting of bunches of orange colored needles which melted at 138 °. The picrate of 1,2, dimethyl napthalene,B.P. 262-4®, is said to melt at 139®, and this hydrocarbon was present in considerable quantity with its other i somers , which were not identified. Phenanthrene ,C-j 4 H|Q,M.P. 100® ,was assumed to be present on account of the presence of dltolyls from which it may be formed 35 . by simple reduction , but as with anthracene, only an extremely smal] amount was isolated. It was obtained from the hexane mother liqu- ors after the anthracene and methyl anthracene was removed. It had a melting point of 96 ^ and after an attempt at oxidation which yielded only a pasty mass, the picrate was made from benzene solution and had a melting point of 141-143°. The melting point of the picrate given in the literature is 145®. Attention was now given to the high boiling solids which came over Just before coking. These were orange colored crystals merging to a dark green toward the very last. These green cryst- als corresponded in every way to chrysene as described. in the literature , and as this substance is relatively insoluble in carbon di sulfide , the mass wa,s dissolved in the smallest a,mount of this solvent possible. It all passed into solution, but on standing for several days a heavy black residue settled out of the carbon disulphide solution,. which was filtered off. There was but a trifling amount of this, but it was dissolved in a few cubic centimeters of benzene and recrystallized. A small quantity of transparent crystals were obtained which on filtering, colored the filter a deep green. There was Just enough for a melting point which was 255°. Chrysene, Ci3Hi2»M.P250® was present in extremely small quantity. The carbon disulphide solution was then evaporated to dry- ness, and taken up with alcohol, from which the solid obtained was crystallized several times. Pyrene ,C^ gH-j q,M P. 148°, was what this solid proved to be, its melting point was 145-7° and a picrate made melted at 220°, which checked with the theoretical M.P.222°. A pure white substance which had a blue fluorescence, distill- .0 iMK tf t^ia> ! w riii;ww>i 36 . ed over with the heavy oils during steam distillation. This had a melting point of 140-141° when impure, but on recrystallization from benzene the melting point rose to 2 13^, where it remained stationary after another recrystallization.lt was assumed to be a very pure form of anthracene , which had appeared at two other points in the investigation, and had been identified , so no fur- thur identification was made, as the amount wa,s too small. These are all of the substances identified in the straight xylene runs and undoubtedly they make up a major portion of the total. There are, without doubt, many compounds present in small quantities which entirely escaped detection. The oils from which the diphenyl was obtained might contain ethyl napthalenes and methyl diphenyls, but no further attempt was made to identify any compounds at this point. The higher fractions, from 280-2go? from the results of previous investigators , was thought to be largely diphenyl ethane, and there was a possibility of mesitylene being present in the small quantities of oil obtained from the fractions distilling between xylene and napthalene, although this possibility was lessened by the course taken by the breaking down of hydrocarbons in pyrogenetic decomposition. According to the most generally accepted theory the high benzene homologues are the first to decompose, so mesitylene would not be formed here at all, or if formed momentarily , would be decomposed as fast as formed in the furnace. The latter is probably the case. 14. Hydrocarbons in the Xylene-Ethylene Runs. The hydrocarbons formed in this series of runs, in general , comprised the same sub stances, and were separated according to the same scheme as those in the straight xylene runs. Napthalene was formed in 57 . considerable quantity, but its identity being obvious, no tests were made upon it. Diphenyl was produced in about the same quantity as in the former case, and was identified as previously. The oils were considerably less in quantity than in the straight xylene runs, but the solids were present in greater proportion , thi s v/as especially true of the substances which distilled over 360°. Over a gram of pure chrysene was isolated from the fraction distilling just prior to coking. It was a yellowish white solid upon recrystalliz- ation from benzene, and melted sharply at 250^, Pyrene was present as before and a small quantity of a substance melting between 244*^ and 246° was found with the pyrene and separated from it by difference in solubility in alcohol. This was oxidised in glacial acetic acid with chromic acid,a,nd a product obtained which melted at 178° This proved the original compound to be 2, 3, dimethyl anthracene. Anthracene was present , though as before, in surprisingly small amounts, and the a and b methyl anthracenes were found in large quantity. The a and b methyl napthalenes were in the oils, as was 1,2, dimethyl napthalene, which was fractioned and identified by making the picrate as before. All indications show that the two runs were essentially of the same composition. The xylene-ethylene product contained the solids in much larger proportion than did the straight xylene products, so it was possible to find the 2,3 dimethyl anthracene which was probably present in small quantity in the straight xylene r^on as well. Chrysene was also present in larger amounts as well as a very small quantity of a hydrocarbon which was obtained pure as small . granul ar . bl a.ck crystals which looked like powdered 38 . graphite. This had a melting point of 255°and in glacial acetic acid solution it was a dark purple .without fluorescence. It was insoluble in benzene and carbon disulfide , but soluble in glacial acetic acid, from which it was recrystallized. It v/as not identi- fied, but was thought to be chrysogene.a compound about which very little is known. 15 . M anipulation of Napthalene Run Products. These samples were principally napthalene , so the first thing to be done was obvious- ly to get rid of this napthalene entirely. Steam distillation suggested itself immediately , and the five samples were steam distilled until the condensate came over clear. The residues left in the flask in every case looked like pitch with consider- able amounts of free carbon in suspension. After trying several solvent s , ether was fina,lly decided upon as it ' seemed to be the most efficient in dissolving the tarry constituents in the product. After evaporation of the ether extract a very viscous reddish oil was left behind in every case, but the combined oils from the five runs did not exceed three cubic centimeters, and it was impossible to identify it. There was a solid in suspension in this oil which was thought to be benzerythrene , as its appearance and the manner in which the product came from the 4 furnace, as a'*bright red powder", all bear out this assumption . It was not identified. The solids left by the ether extraction were black or dark brown in all cases, and consisted principally of amorphous carbon. A suggestion of crystalline structure in one of the samples led to an extraction with benzene, and after filtering off the carbon and concentrating the solution, a mass of pure white plates VT ® A - • V ^ ^ ■ : ui^^' .“ r,cj^'/afc "s ^-r. j- ;^*i.'Xc»r ►>;•.. X i: -\^Lp_ ill r> /. : f. »* ’ '.'-i.tn'’'' v^';^i n-. .^i*i:^^c^i. cvZ-xr-*} t . ‘ ■ ■• ;-‘ t .’ ;.‘; ,7 ■ ©I U '. . vrrl, -.'C f \ * * '.,^, l>.t4v> »»l ‘;^ ■ ' S_.si;r;St:i . : . < •V > , ^ ' ■ . f u ^ ■ -i, c ' ■ £y .'.' ! ' ( •■ .• ■'• w >-' - Is. V 5 ^ is- r ^ If, » ■ :;;V ^>;. 4 ■ .- V4 * ..a. - Jf'u ' ■ ' >■ • c *. ■ ' r - ' ,o V ■•: ■ * ■ -1 V t :’■: Miiv.o.. ^ SiS .A. 'J f '.Av* : i ,i K''- - ■.« ij ; r a ' X' ' 'j' ‘ ■' ■ -.s-f..;' 7 ^ ,- • y:iV:l ?.' r^f-: ..*■ /{f-4n : •: '• ’. r ' Six,' i'l' f-'fly : *;' t'l ■ / ■ J’i' -fl'X-' 1TC«; jt * «■ * K.' V .=^1 r ':t£: r, ■j i'’ ^'K . !;•(’ •■ ■ ’ h !: iJ: ] .T: •• -■ • ■ ■(, • •■ - Ih-ul '*• fey i.ni 'y'^^ rr - *• * < *<• p . ' , '. ■' % 4 ' *i ' ■‘ v' 0 £; •£v •'.• ''• M ’v ^ » '-jrv' . .i:-v >.. 'fij, !.; :ri:Sn '* < ii b«j t »' ' ■’'O' f * f • Wv- ei . V. ’ ^ ■’ '■' ^• ", • ■>• 7 .vry'X^ :■ £c!jl 5. i;^ «. LK J' * 'a , ‘ '. ■ ,^.4/ r-v - / ■ •> c^‘ '>• . ' <■ ■' ) ? / '- ;■ bi: fi ' ' f- r'/. : • , >. ►:'£ • '‘•;:;• • 'icv ';"•.’ X'" .. 'V , . ■■•t* T'''' ''•i, ’ )T.. >4 * - *{,t ' • ■’ r '/K, ...v>.*'-: , j.’. t,. ; ■ • ni {mc<*xd '.' '-■ if iji'\ '^ 4 .;'^f‘>,y' ■■ i,'r. ; -v! ■ ■ ■ i it. 39 . crystallized out. These were filtered off and identified. The napthalene which had been steam distilled was then dried and distilled under atmospheric pressure to see if any other substance had come over with the napthalene. There v/as nothing, as practically all of the napthalene boiled at the desired temperature, 2 12®. The products obtained seemed to be independent of the carrier o;as, except in the quantity produced, as all of the runs had the same composition. 16. Hydrocarbons in the Napthalene Runs. The oil produced was so difficult to handle, and so small in quantity that nothinf?; was identified in it. The solid produced was present in considerable quantities, and was found to have a sharp meltinp; point at 189*^. It had the same melting point and appearance in every case so it was assumed to be the same substance produced in each run. It was first thought to be b-napthoic acid , qH^COOH , as phthalic acid is formed in a similar pyrogenetic decomposition, although the melting point was quite high for it. Another reason which supported this view of the presence of napthoic acid was that it was formed in the 002 '^^^'^ in quantities exceeding the amount produced by all the other runs put together, and its formation could be easily explained by the simple addition of a molecule of C 02 to a molecule of napthalene. A portion of it recrystallized from petroleum ether had a constant melting point at 184°, which was much nearer the melting point of b-napthoic acid (132®) than the sample obtained from benzene. This theory was broken down at once when it was found that the substance was insoluble even in hot concentrated alkali. r ( ) i 40 . The regular procedure was then resorted to. It was found to be a saturated aromatic hydrocarbon. Consulting the literature , there was finally found a hydrocarbon which seemed to fit the require- ments. It was b-b-dinapthyl , which crystallized in plates, melted at 187*^, and had a stable picrate which melted at 183-4°. The pic- rate was made in benzene solution and came down as an orange crystalline precipitate , which had the desired melting point, 132- 184°. It is seen that dinapthyl is the principal product formed in the pyrogenetic decomposition of napthalene ,in the same manner as diphenyl is formed in the thermal decomposition of benzene, a point which, heretofore, had not been touched upon in the treatment of pyrogenetic decompositions of hydro carbons. The a-a-dinapthyl was probably formed in very small quantities, but it escaped detection. 17. Approximate Yields in each Case. No attempt was made to get quantitative results in this work. Note was generally taken of the relative amounts of the substances present in each case, and for estimations of the amounts produced in proportion to the amounts of xylene run through the furnace, ref erence must be made to the thesis of Dr. Bradley^ in which the methods of formation of the tars are discussed in detail. Napthalene was formed to the extent of about 5 to 7 percent, in both the xylene and the xylene -ethylene runs, which is consider- ably higher than most investigators had found previously. The oils comprised about forty percent of the total in the case of the xylene run, and was about one-half that amount in the case of the ethylene run. I ' 5 I ■I 4 i 41 . Anthracene was very low in both runs , certainly not over two or three percent. The methyl anthracenes comprised the bulk of the solids in both cases , amount in/ 2 ; to about one third of the total in both cases. The hi? 2 ;h boilinc^ tarry constituents containing the pyrene and chrysene were present in small amounts; in the case of the xylene amountina; to perhans five percent. In the ethylene runs they were much higher probably near twenty percent. In the napthalene runs, as was mentioned before, the single CO^ 3 :’un contained the bulk of the product. The nitrogen sample came next with about one-third of this amount, or roughly, five grams. The hydrogen and one CO sample contained about four grams each, and the other CO sample contained less than one gram. 18. Methods of Separation Best Adapted to this Type of Hydrocarbon/ It was found, as all workers in the separation of decomposition product hydrocarbons seem to agree, that a preliminary fractional distillation was essential. Steam distillation was next tried with considerable success, as it divided the substances present into two classes , those volatile vrith steam and those non- volatile. In general, the substances having two benzene rings or under in each molecule distilled in a current of steam quite readily. Three ring compounds were found to be almost non- volatile,and above that completely non-volatile, as the clarity of the oils testified. In the preliminary distillation the oils as low as 190^ were a reddish yellow color with green fluor- escence, but after steam distillation they were practically colorless with slight opalescence. The anthracene may be removed from accompanying substances by solvents as described, or by direct oxidation of the fraction as obtained with chromic acid f 1' !:■ • \ ‘ ‘ ' 3 . 1 fc/i * ^ ^ V, ►«' I. ^ JA'-i J ^ . 1 42 . filtering and washing, sund then extro,cting acidic and uanlc substances by treating with dilute alkali and acid successively. The anthraquinone resulting may be reduced back to anthracene or used as obtained. Napthalene crystallizes out of its accompanying oils on standing, and may be filtered off direct in a fair state of purity. These are the only two substances produced which are other than of scientific interest, and the separation of the other sub- stances recommended is the same as described in the course of the study. In the case of the napthalene samples, the method of separation was more or less a product of the"cut and try'* system, as there was no previous work done along these lines to guide the methods used. The use of solvents was seriously contemplated without a preliminary steam distillation, but fortunately was not attempted. The preliminary steam distillation removed nothing but the napthalene; this was proved by the subsequent distillation of the napthalene obtained. Various solvents were tried on the tarry substances resulting from steam distillation, but none of them were as satisfactory as the common ether finally decided upon. The dinapthyl found was almost insoluble in cold ether, but quite soluble in benzene, from which it was later recrystallized . t9. Some Special Aspects of the problem. A carbon disulfide solution of the origional run, completely free of solid matter, v/as found to give a heavy, dark brown, floe culent precipitate on pouring into a large volume of petroleum ether. This, on 43. filtering was found to have a melting point of 1 00-1 05° • Having an opportunity to examine asphaltenes obtained from G-ilsonite by exactly the same method ; precipitation of the carbon disulfide solution with petroleum ether, it was noticed that it agreed remarkably with the natural asphaltenes in appearance and melt- ing point. Knowing that the natural asphaltenes have a high sulphur content, and knowing that there was but a trace of sulphur found in the run, it was thought that the substances obtained from these tars would not contain sulphur, but on running a peroxide fusion and determining the sulphur, it v/as seen to be present in practically the same quantities as the origional asphaltenes contained it. Either all of the sulphur indicated in the previous determination on the whole run was concentrated in these bodies, or they were the result of polymerization with the CS^used as a solvent. No further work v/as done along these lines, but an exhaustive study of these strange products might reveal something about the structure of asphaltenes. Wax- like substances appeared in small quantities , at various times during the investigation. They ranged from hard, white waxes to brown , resinlike bodies, and invariably resulted from solutions of oils and solids which had stood for some time exposed to the action of light. The coke resulting from the primary distillations ?/as in all cases very dense and hard. Solvents extracted nothing from them. 1 44 . . Ill Summary. The results of the fore,o;oing investigation may be s^omm- arized as follows; 1 .Hydrocarbons identified in the straight xylene runs. Name . Formula. M.P. (B.p. ) How Identified. Xylenes (mixed) 3.P. 145-150° appearance ,B- B* Napthalene CO M.P. 80° Pi crate 151° Diphenyl < M.P. 70° p-brom deriv 305® a- methyl napthalene 00 B.p. 241-3° Appearance and Picrate 112° b-methyl napthalene B.P. 240-2® appearance and Picrate 1 1 2® b-methyl anthracene COX'" M.P. 202® b methyl anth. carbonic ac.263° a-methyl anthracene O0 M-P. 202-5° a methyl anth. ^ carbonic ac .206 Anthracene 000 M.P. 212-14° anthraquinone 278° Phenanthrene exp M.P. 98® picrate 141-3° Ditolyls( mixed) B.p. 275-80® Isophthalic 295° 1 ,2 dimethyl napthalene B.p. 262-4° picrate 138° Pyrene M.P. 145-7° picrate 220° Chrysene 0X0 IJ.P.250° Appearance, M.P. . Hydrocarbons identified in the xylene -ethylene runs. Name Formula M.P. (B.P. ) How Identified. Xylenes (mixed) B.P. 145-150° appearance, B.p. Napthalene CO M.P. 80° smell, M.P. ,B. P. Diphenyl a-raethyl napthalene M.P. 70° B.P. 241-3° M.P. ,B.P. p-brom deriv. Appearance ,B. P. i I i 45 . Name Formula M.P. (B.P. ) How Identified. b-methyl napthalene CO'"^ B.P. 240-242° Appearance,B. P. a-methyl anthracene 000 M.P. 202-5° a- anthracene carbonic acid 206° b-methyl anthracene COO' MP. 202° b-anthracene carbonic acid 262° Anthracene coo M.p. 212-4° anthraquinone 276-8° Phenanthrene o9 M.P 98° Pi crate 143° 1 ,2 dimethyl 05""^ napthalene B.P. 262-4° Pi crate 136° 2,3 dimethyl anthracene M.P. 244-6° Quinone l80-2° Pyrene M.P. 145-7° Appearance, M. P. Chrysene 0X0 M.P. 250° Solubility , M.p. Hydrocarbons in the napthalene runs. Name Formula M.P. (B.P. ) How identified. b-b-dinapthyl \ a-a-Dinapthyl ocnoo M.P. 187-9° picrate 183° ooxo M.p 154° Not found, but a-b-dinapthyl CXTXO M.p. 79-80° assumed to be present In summarizing the separations used the following flow sheets gives the "best idea of the methods in a condensed form. LIQUIDS ( steam J-dlsti nation) ORIGIONAL TAR at 1 gt.i r>n ^ VOLATILE (distillation) i NAPTHALENE (filtered off) -i NON-VOLATILE ( inseperable heavy oils, none identif- ied. ) “I — 290-360° (extracted with hexane ) SOLIDS AIITHRA( :ENE (least o 360 to coke, (extracted with CSo) soluble ) Diphenyl, a and b methyl napthalenes, 1,2 dimethyl napthalene ,ditolyls , and xylene also obtained here and purified by special methods. snd b >thyl anth. ire sol) PHSNANTHRENE (most Sol. ) SOLlfebE (pyrene, dimethyl anthracene) INSOLUBLE (Chrysene) IV ?l!!l » Rr:-',.'- '■' ’ ri-j®-;' ■ ■■' '!;■•%’ -. %• ■'■' ‘ ;ta% i ji - c si. • •- 'v;' ^ :><;■'■ ' r: ■.'* . 'V.^-* . ' " ‘‘.^6 ■■ CiY> ' '.’•'ttti<>£'-“C:-^‘ . “>f I • . ^ ta yl.- t . •■ - - -/£<>■ fe .:■ . '^Tv 'Afslb'ii .%U '•' fftfapytitt'Xfl “v ** ■■■»'' / ■ ' '-^ .* ' * o-’.-^d ■ ■ ' . .: ' ViSV^ ' ,. ■ , •. ■' ■• ,; .-'i \ ■ f. W 1] , ■/' t-, -N . ■■■ •<;«•"' ^ ■ , ” u , . 'f ..', . . lit M ^ •"■' V J "J ■ •* \v> .'■ I’l 4 ^ '■ T .A ^ ' ■>*. ’ i i1 *. ■ ■"’ ' . r ". ■‘^ ' , • rP-rt?r; ejC5^i^!!54T- '»r^av & / t m£i. sirt^ fVtf‘ 9 w '.jSffi/c«.i 7 . >, ^ ..V ''"i i i . . ■ : •• ‘ .. < TN^ ■' ' -''^ . v-’«'fv. • /r.',’ . . ^ it' ■ ' ■ ,*-. M, . •. _ ^ a'mbt- a M lo ir. ytt^ • • • >.", X ■'i -vA. • * <* • m’’ , gaiypSai . , ,.; ‘-i?/ ’ *5i6t rjtte i.. . 'f'^pA^-Ti<5U;v , "- r ' HI ) {hoXJ A^!.t2vf- Xf^V *■' , . . ^cti' ■ • : ..i.3.;-i-.-H<*.v, f r j . . , , « ( "Aat / f>e^'»j'«Li:t'7; . .. . A3 ( \M ,.J, ; ';■', , ,.(■ '..w— ; ■'A'' -i W WiU j lJ ^ ! "XaIuOv;it‘ XV rw’.^.'.^u'i.x ' ■V‘. V ■ ' -rvMrj ' ' ’X ■'■:■ ^ .* 1 .'• .’ V'‘iv iL ,f '.Am •.Vt< .r 46 . The flow sheet showinf^ the scheme of separation used in the case of the napthalene samples is given below. ORIGINAL SAMPLE ( steamidistillation) VOLATILE WITH STEAJf (Distilled) NON-VOLATILE WITH STEAiM ( etherj. extraction ) (found to consist of entirely unchanged napthalene ) SOLUBLE (consists of a red oil, not identified) INSOLUBLE SOLIDS (benzenej, extr. ) Soluble b-b-dinapthyl . Insol. (carbon) ■ '•'’"Ti" '.iM, vTtJt;' - f • '’i < TL^ ' ’’J f*^.'’^- ' tjy ■, •w \''' ' '5 . ' ‘•<‘!' 'It '■ '■;• ■' . ' "% ' . 'Iffj^.t ‘ t'Vi t’ ,.r'^'iJ'<:!X'-f-. ;#E. frr4 • _ . .. „ ' ... . . '• .'... * f. ■ v> ‘ .ii--. . .. : __ ■>.'!?} f' ■(< ’V V ••■ y ••'.j(?.p; ■; y- • ■ '. ■>■ f. tl‘, l\ ! •' ' ‘ . ' ■ ,1 *' 'v:( 'V'l'^" ' ■ '*V'' ' • ■- ’’'V'r-. ”' •' ^'■'''•^* ' • 'V " - '• ; ”■' 'i " ^'kv ' ^ , »» ■ ri.} ' • X •* f.i .;4 no^ :trt ^ •% -M&j (l»ilUn!»M y . /.'i ’■■'• I'. T,r^i / V*’ ■ ... ^,' •' ■< ., r^. ■■; "V'- y •,,. •t ■,•■*, rf^' '^%:m i .'V ■:^^- •'*+ 5 ? .' ■ V. , . .;V aksi’ ■' I > i'’* • • . . ^ -IS ■ ->r, ;y;^ > j; y . i;'-..’ ■ ’ ■' -' 'iSiii ‘ ■ y ''-yf-'? - *. kV ' ^ mmH ■'• . '. . >; !'• '-■ .-i r-iVjL v, irV- ■ V V* ‘ ^i' '"’.'i^:'. .'* i'-'>’'^’r''fyv y s*' '■ <>y, '.• ';t‘ y/’'. ^ ‘ :v- •■■‘■':'-l /;'i:‘;K™i^ f- i ,. .■ . -,; V , ■ ... ' "A.'''^\ y‘y ‘1 j' • ‘‘'' ‘■'-■y'vV.'i^ ' ' f- ' ■< * r. 5 .‘'’‘^*!rw L i» w -^ — r * *' •■~II ■»w w ^i»*> **' ■ ■!< M l I'H iii 47 . IV. Blbllog^raphv. 1. Beilstein. Orp;anische Cheraie. 2 . Richter. Lexicon der Kohlenstoff Verbindungen, 3. Millikens Organic Analysis, Vol. I. 4. Allens Cominercial Organic Analysis. 5. Clarkes Organic Analysis. 6 . Qualitative Organic Analysis(Chera 21) 7. D.T. Jones. Jour. Soc. Chem. Ind. ,36,P3, I 917 . 8 . D.T. Jones. Jour.of the Chem. Soc. 107,0-1582,1915. 9 . F. Haber. Ber. ,29,P 2691,1896. 10. M.P.E.Berthelot. Bull. Soc. Chem. Paris. ,7, P 113,122,210,217, 274 . ,1367. 11. G-.W. McKee. Jour Soc. Chem. Ind. ,23,0 P 403,1904. 12 . V. N. Ipatieff . Jour. Russ. Phys. Chem. Soc ., 59 ,P 681 , 1 907. 13 . J. Ostromisslenski and J. Burshanadse. , Jour. Soc. Chem. Ind, 29 , P 682 , 1910 . 14. C. Smith and W. Lewcock. , Jour. Chem. Soc. , 1 01 ,P 1453,1912. 15 . W.F.Rittraan,0. Byron, and G. Egloss. , Ind. and Eng. Chem, 7, P 1019, 1915 . 16. J.E.Zanetti and G.Sgloff., Jour. Ind. and Eng. Chem. ,9 ,P350, 191 7. 17 . C. T.Liebermann and O.Berg. Ber. ,11,p 723 , 1878 . 18. K.V. Charelshkov. , Jour. Russ. Phys. Chem. Soc. , 38 , 1293-94, I 388 - 92 . 19 . Manson S. Bradley. , Thesis, 1 920, U. of 111. 20. Edward E. Charleton. ,Thesis, I 9 I 8,U. of 111. 21. Clark, J.M. , Jour. Ind. and Eng. Chem. , Vol 11,#3,Pp 204,1919. 22. Cook,0.W. 5 ind Chambers, V. J. , Jour. Am. Chem. Soc. 43,2, P 334,1921. r f * ^ ff I.' i I • r ; ;.c i *■> ■' ' . D i t ~ ! 48 . 23. Zanetti , J. E. and Kendall, M., Jour. Ind. and Eng. Chem. Vol. 1 3 , #3 Pp. 208-1 t ,Mar. 1921 . 24. Rittman,W.F. , Dutton, C. B. and Dean,E.W. Bureau of Mines, Bull. 114. 25. Cobb,J.W. and Dufton, S.F. ,The Gas World,Vol.72,Pp485, 1920. 26. Bone and Coward. Jour. Chem. Soc. ,93, Ppl 197, 1908. 27. Smith and Schultz., Annalen. , 203 , 1 1 8. 49 . Vivita. The writer of this thesis received his early education in the grade and high schools of Chicago. He entered Loyola University in the fall of 1916 and continued there until February , 1 918. He then transferred to the University of Illinois, where he has been continuously up until the present time.