r A STUDY OF THE PRODUCTS OBTAINED BY THE SOLVENT ACTION OF DI-PHENYL ETHER ON FRANKLIN COUNTY COAL BY ARVID HENRY BERG THESIS FOR THE DEGREE OF BACHELOR OF SCIENCE CHEMISTRY COLLEGE OF LIBERAL ARTS AND SCIENCES UNIVERSITY OF ILLINOIS Digitized by the Internet Archive in 2015 https://archive.org/details/studyofproductsoOOberg ACKNOWLEDGMENT This work war. undertaken in the Chemical Laboratory of the University of Illinois. The problem was suggested by Dr. Layng and its completion was due largely to his timely advice and helpful suggestions. I wish to take this opportunity of expressing my appreciation to him. > « . TABLE OF CONTENTS. Page Acknowledgment i I. Introduction 1- 10 1. Nature of Problem 1 2. Historical 2- 10 II. Experimental 11 -13 1. Solvent Methods 11 -12 2. Carbonization Methods 12 -13 III. Results 14 -16 IV. Discussion of Results 17 -19 V. Conclusion 20 VI. Bibliography 21 -22 A STUDY OF THE PRODUCTS OBTAINED BY THE SOLVENT ACTION OF DI- PHENYL ETHER ON FRANKLIN COUNTY COAL I. INTRODUCTION 1. - Nature of the problem. A Knowledge of the constitution of coal, which will explain why some coals coke and others do not, has been sought thru various channels of research. The means employed in this search have been ultimate analyses, microscopic studies, fractional carbonizations, and solvents. Each onehas given us some infor- mation and as a result various theories have arisen from time to time, which attempt to give the desired explanation. It is the intent of this work to find, if possible, some differences in fresh and oxidized coals thru a study of the products obtained by the use of a solvent, which in this case is di- phenyl ether, to confirm the present theories or prove a new. The work will not be confined to the extracts alone but to the coals and the residues as well. 2. - Historical. In order to proceed with the problem, it is necessary to re- view some of the work that has been done by former investigators, which has led to our present theories. White and Thiessen^ , in their great classic went deeply into the problem of the origin and constitution of coal and thru their studies brought out authoritatively the idea that coal consists of two different constituents, which they called cellulosic and O resinic. Cross and Bevan studied the action of sulphuric acid on cellulose and coal and they came to the conclusion that there (2) could "be two possibilities for the formation of coal. The first possibility is the lignifying of substances which originally con- sisted mainly of cellulose and the second is the combining of the cellulose and the resinic parts, found elsewhere in the plant and formed perhaps thru the oxidation of the hydrocarbons. Anderson-^, inf a continuation of the work done earlier by himself and Roberts on the treatment of coal with carbon dioxide and nitric acid at 31 £$^C., came to the conclusion that u whether or not different coals contain bodies that are in a genetic or homologous series, such bodies form at least only a part, and it may be but a minor part, of the mineral. Furthermore, that a con- siderable part of the organic matter consists of a complex com- pound comparatively rich in nitrogen and that, above all, resin- ous matter is present to a small but fairly constant degree'.' . 4 Parr and Francis conclude from a study of the literature as follows: (1) That coal was not derived from cellulose alone, but from a combination of cellulose and aromatic bodies. (2) That the main portion of the coal is a definite unit, associated with a small amount of bitumen or resinous matter. (3) That the struct- ure and the properties of coal are greatly modified by the re- moval of certain portions of the coal by one means or another, mainl# solvents. David 17hite5 explains the coking qualities of coal as de- pendent on the presence or the absence of certain gelatinous algae The more of the algae present, the closer the percentage of hy- drogen and oxygen comes to that of bitumen, that is higher hydro- gen lower oxygen and therefore better coking. He concludes that . i : • ■ . j . o' ”0 ' - ' i . . ; : . *• •• • * •; . - * • . . • • . 1 ' . . • . * - . - . . * ' l t • * > > 1 ( 3 ) coals high in volatile matter and showing high bituninization will coke and the degree of this bituminization will be indicated by the ratio of hydrogen to oxygen in the dry coal. From this assumption or apparent fact he has developed the theory of the H:C ratio as being an indication of the coking power of the coal. Wheeler , working in conjunction with Burgess and Clark on the carbonization method, has advanced several important facts in conjunction with the carbonization theory. There has developed certain contradictory evidence from time to time, which will be discussed later. Their researches brought out the following: (1) There is a well defined temperature between 700° and 800° C. which coresponds with a marked and rapid increase in the quantity of hydrogen evolved. (2) Evolution of hydrocarbons of the paraffin series ceases almost entirely at temperatures above 750°C. (3) Ethane, propane, butane, and probably higher members of the par- affin series form a large percentage of the gases evolved at tem- peratures below 450 C. (4) The rate of CO evolution is uniform thruout a distillation at any one temperature, and is maintained up to the end. While the rates of evolution of the other gases fall off, the CO rate increases with the temperature. From these results they argue that coal contains two types of compounds of different degrees of ease of decomposition. The one , the least stable, yields the paraffin hydrocarbons and no hydrogen. The other, the one decomposing with greater difficulty, yields hydro- gen alone ( or possibly hydrogen and the oxides of carbon) . They likewise claim that probably the difference between one coal and another is determined by the proportion in which the two types of (4) compounds exist. Porter and Taylor? working with what they considered re- presentative coals of the United States found disagreements with the results and the conclusions of Wheeler. Their results can he critized from the fact that they used old coals, which later have been shown to have been oxidized and therefore altered, and also from the fact that their analyses of the gases show no oxygen. Their results are as follows: (1) More than two thirds of the or- ganic substance of coal is decomposed at temperatures below 500°C. (2) The first decomposition occuring in any type of coal as the temperature is raised is the breaking down of certain oxygen- containing substances related to cellulose, the products being cheifly COg, GO, and water. (3) Coal probably breaks down more or less at all temperatures but the temperatures at which coal com- mences to give off volatile matter in appreciable quantities is at 250 c C. or lower. (4) paraffin hydrocarbons predominate at tem- peratures below 400°C. (5) Thermal decomposition of the volatile matter occursvery easily at temperatures around 750°C. (6) Dis- tillation at temperatures above 750 J C. yields gases in which hy- drogen largely predominates whether secondary reactions are pre- vented or not, but secondary decomposition of the volatile matter will increase the total gas yield at the expense of the tar. (7) Water of decomposition is produced above 250°C. but in greater amount below 530°C. than above. This water vapor may react with tar vapors or gases in passing out of the retort during high tem- perature carbonization. Porter and Taylor take issue with Wheeler on the proposition that the appearance or rapid increase of hydrogen between 700 C. ( 5 ) and 800°G. marks a really defined decomposition point or critical point at which, after the decomposition of the resinic material, that of the more stable, the cellulosic, commences. Th y think that the marked increase in the H, ;;i may be due to secondary decom- position of the volatile constituents of the coal. They propose the following hypothesis for the constitution of coal: All kinds of coal consist of cellulosic degradation products, more or less altered by the process of aging, together with the derivatives of the resinous materials in different proportions, also more or less altered. These substances are many in number and closely graded into one another in their nature and composition. They all under- go decomposition on moderate heating, some , however , decompose more rapidly than others at the lower temperatures. The less al- tered cellulosic derivatives decompose more easily than the more altered derivatives or the resinous derivatives. The cellulosic derivatives decompose on moderate heating so as to yield water, carbon dioxide, cenbon monoxide, and hydrocarbons, giving less of the first three products the more mature and altered they are. The resinous derivatives on the other hand decompose on moderate heating so as to yield principally the paraffin hydrocarbons with probably hydrogen as a direct decomposition product. The more mature bituminous coals with good coking qualities contain a large percentage of resinous material and their cellulosic material has been highly altered. The younger bituminous coals consist chiefly of cellulosic material much less altered than thatein the older coals. They undergo a large amount of decomposition below their fusion point and for that reason do not coke. * . : , ( 6 ) Lewes® points to the fact that if, as Burgess end Wheeler claim, the cellulosic constituent decomposes last; then, since it is the cellulosic part that is high in hydrogen and low in oxygen, an analysis of the coke donned at 450°C. should show higher oxy- gen and lower hydrogen than the original coal, due to the elim- ination of the resinic material which is supposedly high in hy- drogen as compared to oxygen. He tried this "but discovered that , though; the coke was lower in hydrogen than the coal, it was also 9 loY/er in oxygen. White , in his studies of resins, found that some resins have excess oxygen over hydrogen. Lewes view of the mechanism of coking is as follows: Water evaporates and then recondenses in that part of the coal, which is cool enough. Hygroscopic water comes off first and then hunus bodies commence to decompose and they give as their first product water. As the temperature rises to 3$>P°C . or above, the coal be- comes semi-fluid. Decomposition of all the coal bodies commences and increases rapidly in vigor wi th each acbssion of heat. The fluid tar from the humus, the slightly viscous tar from the re- sinoid bodies, and the rich heavy tar from the hydrocarbon gases go forward as vapors with the advancing gases away from the heat and towards cool places. It must be evident then, that as the heat spreads to the zone in which the water has condensed, that this water again evaporates and renders a large amount of latent heat, so cooling it down and leading to the condensation of the tar vapors. When in turn this is reached by the advancing temperature and is heated to 350° or 400°G. , it is only the more volatile portions that distil and a deposit of pitch is left, which binds the half formed coke. As the temperature rises still higher, the • ■ . . • - ’.V ■ • . • . ( 7 ) residues and the tar still further decompose until the hard coke is left. That tar, which is least volatile and leaves the largest amount of pitch, is the most valuable as far as coking is concerned. If all distillation of the tar out of the coal could be prevented, then we cpuld coke our so-called non-coking coalG. Dr. HargerlO states that it is generally held that the coking property of coal is due to the presence of resinic material. In addition to this the better coking coals have a class of bodies ( which have a definite melting point and are not affected by air or by heating) in sufficient quantity to renddr the whoiLe mass liquid and so will form good coke. Anderson, Roberts, and Prof. Lewes considered these as hydrocarbons bodies and in strongly coking coal it is the presence of such bodies which cauee coking even though the resins are oxidized or saponified. The assumption of such bodies, according to Dr. Harger, is entirely unjustifiable in the absence of data to prove their existence. Parr and Oli^ 1 -'- in the resume of their work developed certain facts. First, the formation of coke depends on the presence of certain constituents having a melting point which is lower tha h the temperature at which the decomposition or carbonization takes place. Second*, these constituents can be readily oxidized and when oxidized their coking properties are sadly impaired. Third, coals containing an excessive quantity of the coking substance produce a light porous coke. Fourth, by the use of temperatures between 400°C. and 500°C. , all of the resulting products are of a type distinctly different from those obtained by the usual high temperature pro- cedure. < ' ( ♦ . • < T ~i * < «. * ■ • • if < < ' ( 8 ) Bonel2 mentions the work of Dr. Smythe, who, in 1851, worked with different solvents for coal. He employed benzene, chloroform, ethyl alcohol, ethyl ether, light petroleum, and acetone. He pre- cipitated the resins from the solutions with petroleum ether. Benzene, the most efficient solvent, extraxted about three per cent of the coal. Prof. Bedson'" - ' in 1899 used pyridine in-; a Soxhlet extractor as a solvent for coal. His results show that there is practically no correlation between the amounts of pyridine extraction and either their gas-making qualities or their yield of volatiles. The extracts were rich in volatile constituents and in an assay test gave a sort of coke, whereas the residues from the treatment gave no indication of any coking property whatsoever. 14 - Dr. Karger used sealed tubes for extraction work and noticed greater action. Prom the fact that ultimate analyses of the por- tions of the coal, after an extraction with pyridine, show a larger per cent of nitrogen than the pure coal, he claimed that the action of pyridine must be other than that of a pure solvent. P.F.Reinsch 1 ' did some work in 1885 and at that time concluded that by means of solvents we could isolate from coals two different kinds of substances, distinguishable by their behavior toward al- kaline solutions and the subsequent behavior of the extract with mineral acids. Wahl^ worked with some coals from Spain, using py- ridine as a solvent, and secured from six to twenty six per cent extractions, while the coal would lose from five to six per cent Ip of volatile matter. Lewes"' gives as a conclusion to his work with pyridine and his treatment of the residues and extracts with sodium ( 9 ) hydroxide the following: That the resin constituents condition the coking of the coal and that they are of two kinds. The one is easilj oxidized and saponified with alkali and it is soluble. The other ie not easily oxidized, is non- saponifiable, end forms a compound with pyridine. This latter class he thinks may be the hydrocarbons from decomposed resins to which Or. Harger , as stated above, took such 18 exception. Frazer and Hoffman did a great deal of work on the constituents of coal that are soluble in phenol. They treated ex- tracts with sodium hydroxide and then with organic solvents, such as ether and alcohol, and suceeded in dividing the coal ,into various portions. No definite or absolutely accurate information was given. 19 Parr and Hadley ' worked with phenol as a solvent and give as a result of such work the following; (l) The extract is the vital constituent concerned in the coking of coal. (2) It conforms to the principle found before of having a definite melting point, \yhich is lower than that of decomposition. ( 3 ) The cellulosic (residue) has the greatest avidity for oxygen. They found that neither the extract or the residue gave real coke, but when mixed in the original pro- portions they would coke. 20 Cherry working on the effect of oxygen on coal came to the conclusions: (l) That in oxidation the first constituent to be affected is the cellulosic portion. (2) That oxidation of the c cellulosic portion alone is sufficient to destr'V the coking pro- perties of the coal. ( 3 ) That a reaction occurs at the fusion tem- perature between the oxidized cellulosic portion and the resinic portion, whereby the resinic material is so altered that it cannot lute the mass. .... . . . : < . . . , ( 10 ) Pi W. A. Bone reports that the absorption by coal of oxygen is slight at ordinary temperatures but that nearer 100°C. it increases and is attended by the formation of two oxides of carbon and steam. These products result from the decomposition of some unstable body previously formed by the absorption of oxygen. Fischer and Groppel found that low yields in extraction could be increased by pre- heating the coal for a short time. By heating, the structure of the coal is broken down so that the tar forming products can be com- pletely extracted. The temperature necessary for this action wa# found to be from 550°C. to 600 ~C., followed by a papid cooling. Fischer and Gludd^ tried the extraction of coal with liquid SO-, and found that it destroys the binding power and leaves the coal as a fine powder. By use of ozone they they converted the coal into a ninety two per cent soluble substance; 24 Hobart in his studies of the effect of oxygen and carbon di- oxide on coal, noticed that the coking power of the coal was des- troyed by the oxygen, but could not find any effect by the carbon dioxide. He comes to the conclusion thatare several compounds in coal, each with a definite melting point. If these compounds are oxidized, then either (l)the melting point is raised due to the formation of higher molecular weight compounds so that the decom- position point occurs before the melting point or (2) the oxidation decomposes and breaks up the coal into unstable substances of in- definite melting point. He thinks that the cellulosic constituent takes the oxygen in the first place and gives it to the resinic parts when the temperature, favorable for this reaction occurs. . . •• f . . ■ • ' « 4 • . . 0 , . ( 11 ) II. EXPERIMENTAL A Franklin county coal was chose#; for use in the investigation One portion was taken r£o he used as the sample of fresh coal and another portion was artificially weathered and used as the sample of oxidized coal. W. A. Bone and F. Hohart, working independently, noticed that oxidation ot the absorption of oxygen by coal would pF progress readily if carried out at a temperature of 100°C. There- fore the sample to ,be oxidized was placed in a bottle, which was supplied with oxygen under a slight pressure, and kept in an oven at a temperature of 110°C. for several weeks. The method of extraction to be used was then determined upon* Several ideas w r ere tried out, including that of using a Soxhlet ex- tractor with a regular as well as an alundura thimble. The use of a large Erlenmeyer flask connected to a reflux condenser with the coal and solvent in the flask and heated in an electric furnace to a temperature of 258°C. proved to be the simplest, roost rapid, and efficient procedure. It was therefore adopted as the method to be used forthwith. Fifty grams of coal and three hundred cc. of sol- vent were taken for each run. A mixture of litharge and glycerine w was used to seal the apparatus. A rubber tube connected the top end of the reflux condenser to a pair of bottles containing alkaline pyrogallol, and this was so arranged that no air could enter the apparatus without passing thru the solution and thus being freed of oxygen. The action was permitted to continue for twenty four hours in each case. The solution was then filtered end the residue re- turned to the flask for another extraction under the same condition* This was repeated three times for each sample taken. ( 12 ) The accumulated extracts from the three runs on each sample were then distilled to a small volume in an ordinary distilling flask, transferred to a large-sized side-arm test tube and taken to drynessin an atmosphere of nitrogen at a temperature of 260°C. The residues were then washed repeatedly with alcohol and ether. •- The washings were at first handled separately from the main ex- tracts hut later were all combined and taken to dryness as one. The residues were dried in an oven with an atmosphere of nitrogen at a temperature of 110°C. The proximate and ultimate analyses of the coals, residues, and extracts were made using standard methods of analyses. The heat of combustion was determined in a Parr Adiabatic Oxygen Bomb Calorimeter, the total carbon in a Parr To&&l Carbon Apparatus, and the nitrogen by the Keldjahl- Gunning method using a .098 N acid and a .104 N KOH solution. The hydrogen and the Oxygen were determined by calculation from DuLongs formulae. The sulphur was determined by the fusion of the sample and the subsequent pre- cipitation of the sulphur as BaSO^. The carbonization experiments were carried out on ten gram samples. The retort was made by sealing one end of a pyrex tube, one inch in diameter and ten inches long, and attaching a side arm to it very near the top. This was connected to a tar and water receptacle, which in turn led to a CaCl 2 U-tube, which would re- move the last traces of moisture from the passing gases. This led to a KOH bulb, in which inp place of KOH was a solution of CdS04 acid in H 2 SO 4 . This would remove the H 2 S and the NH 3 . The final peice in the train was the aspirator bottle to collect the gases. Z 7 / 0- 3 .^ ' ' ' ( 13 ) Thru the cork at the top end of the retort a small tube, sealed at one end , was inserted. This extended to the bottom and the ma- terial to be carbonized was placed around it. One of the thermo- couples waB placed in this tube and accurate measurements of the temperature could be secured at any time during therun. A very small metallic tube was also put thru the cork, thru which nitro- gen was passed to sweep out the system when it was desired. The thermocouples were standardized against Zn and Al, using a Weston millivoltmeter as a recorder of the induced current. Increases in the weight of the tar and water receptacle and the CaCl 0 tube during a carbonization gave the weight of tar and the total water. By subtracting from this weight the amount of water of hydration, which was calculated from theoven drying loss and the amount of coal taken, the weight of the tar and the water of decomposition was secured. The increase in the weight of the CdS© 4 bulb gave thw weight of the H 2 S and the The apparatus was swept with nitrogen before starting and, v/hen the desired tem- perature had been reached, it was again sewpt inorder to .carry away all the gases evolved. Nitoogen was passed thru concentrated sulphuric acid to remove moisture since it was stored above water. Gases were analyzed in a modified D’Orsatt apparatus. A solution of ten per cent KOH in absolute alcohol was tried as a solvent. The method above was used except that a greater pro- portion of solvent to coal was taken, a temperature of 80 'C. was used, and the reflux condenser was water cooled. . • , . , . . < ■ 1 . ' . < ■ * . « ' • < . < III. RESULTS The fresh and the oxidized coal had an equal and constant quantity of material removed by the di-phenvl ether during three repeated extractions. Three percent wqs removed in the first ex- traction, seventy five hundredths per cent in the second, and three tenths per cent in the third. This totaled approximately four per cent for the three operations. An extraction of the coal with a solution of ten per cent KOH in absolute alcohol was also tried and it was found that four and one half per cent of the coal was removed. This extract after being dried and weighed was tree.ted with di-phenyl ether, which removed sixteen and eight tenths per cent. The original extract gave in a crucible test a coke structure, whereas after th§ action Of the di- phenyl ether it would not take on the coke structure whatsoever. Table I. Proximate analyses of products. Moisture free basis. 1 Fresh Coal % Oxid. Coal * Res. Fresh Coal % Res. Oxid. Coal % Ext . Fresh Coal % Ext . Oxid. C oal % V. Matter 26.39 33.16 37.23 38.25 52.84 62.40 F. Carbon 55.75 58.97 54.47 53.40 46.28 36.65 Ash 7.84 7.87 8.30 8.35 .88 .95 Sulfur 1.36 1-33 1.00 1.09 .40 .40 Unit B.T.U. 14588 13,580 14,550 13.530 15,190 15,210 Pyritic Sulfur • 59 • 58 .32 .29 .00 .00 — 1 — r r . Quality of coke good powder bad powder semblance same I t < t n (15) Table II. Ultimate analyses of fresh coal and its products. Moisture free basis. Results also in % coal. Residue. Dry ^Coal Extract . Dry ^Coal Total of 7° ■ y Coal % r Yield 96.00 4.00 100.00 Ash 8.30 7.97 .88 .03 8.00 7.34 Sulfur 1.00 .96 .40 .01 .97 1.36 Nitrogen 1.53 1.47 1.58 .06 1.53 1.63 T. Carbon 73.45 70.45 83.47 3*34 73.79 72.30 Hydrogen 507 5.15 5.51 .22 5.37 5* 64 Oxygen 10.35 9.95 8.16 .32 10.27 10.72 B.T.U. 13.205 12,670 14,998 600 13,270 13,314 Ultimat Moi e analys sture fr Ta' es of Ox ■ee basis bl|- III. idized coal and its products. Results also in % coal. Residue Dry ^Coal Extract Dry $Coal Total % Coal % Yield 96.00 4.00 100. 00 Ash 8.35 8.00 .95 .03 8.03 7.87 Sulfur 1.09 1.04 .40 .01 1.05 1-33 Nitrogen 1. 51 1.44 1.56 .06 1. 50 1.64 T. Carbon 70.89 68.30 84.35 3-48 71.78 70.43 Hydrogen 4.80 4.51 5.38 . 20 4.71 5.12 Oxygen 13.36 12.68 7.36 .29 12.97 13.61 B.T.U. 12. 260 11,710 15,039 600 12,310 12,487 ( 16 ) Table IV. Results of heating the coals and ' their products to 600°C. \ Fresh Coal Oxid. Coal Res. Fresh Coal Res. Oxid. Coal Ext. Fresh Coal Ext • Oxid. Co al Wt. (dry) gms. 9.6 9.83 9.64 9.89 10.00 ' 10.00 Time of test min. 90 90 90 90 90 9° Coke %. 7.15 7.42 6. 64 6.79 5.80 5.10 Tar & Hp$ decomp. % 1.55 1.22 2.07 1.96 4.30 3.90 H 2 S & NH 3 % .04 .04 .03 .03 .02 .00 *Gas cc. per 10 gms. 885 1000 800 1200 1300 1250 CM 0 0 63 141 47 158 46 87 246 °2 41 39 77 62 264 c 2 h 4 " 29 21 14 12 31 52 c 6 h 6 6 2 19 3 4 10 H 2 218 268 187 308 135 180 CO " 42 79 41 213 22 21 CH| " 410 449 364 430 780 465 c 2 h 6 76 ** 49 ** V- y. 7VT 181 Factor n 1.162 ** 1.12 ** ** 1.27 Composition of gas C0 ? % 7.2 14.1 5.9 13 . 2 3.5 7.0 o 3 % 4.7 3.9 9.6 5.2 ■ 20.5 19. 7 1 c 2 h 4 % 3.3 2.1 1.8 1.0 2.4 4.2 c 6 h 6 % .7 .2 2.4 •3 •3 • 9 CM w 24.7 26.8 23*5 25.7 10.4 14.4 CO % 4.3 7.9 5.1 17.8 1.7 1.7 ch 4 % 46 . 6 44*9 45 . 6 43.0 60. 8 37.2 CpHA % 8 . 6 ** 6.1 ** 1 ** - 14,5- — ^4 Figured to yiei & logins. dry ana n 2 free •i i c ,r ‘ » 1 KOH Extract KOH Residue — ^ Residue after di-phenyl ether extraction Ash 2.76 14 . 36 V. Matter 52.96 48.20 39.92 Coking coke structure no coke no coke U6a) Table V. Analyses of products from alcoholic KOH extractions. ( 17 ) IV. DISCUSSION OF RESULTS. Decided differences in fresh and oxidized coal are apparent from a study of the results and these are in accord with the find- ings of former investigations. The calorific value is less in the oxidized coal. This can he explained on the ground that the oxid- ation is in effect a partial combustion and the noted decrease in heating value is generated in the C04I during the oxidizing period. It is also seen that the weathered coal will not coke. The amount of extraction is the same for the fresh and the oxidized coal. The amount extracted from the coal is not large and consequently differences in the solubilities of the two coals, if present, would not be noticed. In Table I, differences between the coals and the products of the coals can be noticed. The ash seems to have been unaffected by the solvent, and there are no indications of any other reagent activity having occured, so the di-phenyl ether must be a true solvent. The residues have lees heating value than their respective coals. THis follows from the fact that part of the pure coal substance has been removed. It is likewise noted that in heating value, the residue of the oxidized coal differs from the residue of the fresh coal by the same amount that the oxidized coal differs from the fresh coal, whereas the extracts from the two are alike. This confirms the theory that the oxygen affects the cell- ulosic constituent first and that if it does affect the resinic it must do so at a higher temperature. Cherry 2{l and Hobart advocate this idea. Di -phenyl ether affects the coking constituent of the coal for 1 the fresh coal does not coke after wi extraction with this solvent, . ■ ■ . . * - . c 1 « < . . . . , . . . (18) The Extracts from both coals coke. Again the resinic part seems to show that it is unaffected hy the oxygen. In Tables II and III the per cent of oxygen is eeen to he higher in the weathered coal than in the fresh coal. This same fact is true for the residues of the coals, hut is not true for the extracts. The oxygen must he con- nected with the cellulosic portion. According to White, the extract can he of higher oxygen content than of hydrogen. The analysis o-f the extract points to tjxe same fact. That weathering does not affect the extract is attested hy the fact the extract from the weathered coal does not show higher oxygen than that from the fresh coal. Q Lewes' argued that, if it was the cellulosic part that had the greater oxygen and the lower hydrogen, then hy coking ata low tem- perature and thud removing the resinic, the coke should analyze still higher in oxygen and lower in hydrogen. This same principle would he applied if the resinic were removed hy extraction. He found that the resulting product was lower in hydrogen hut also lower in oxygen. The analysis, when the resinic was removed hy solvent, shows the same to he true hut though lower in oxygen in absolute amount, it is higher proportionately. Turning to the carbonization data, it is noticed that the oxidized coal shows a large increase in the per cent of CO 9 , app- roximately the same amount of hydrogen, and a decrease of paraffin hydrocarbons. This is in conformance with the theory as originally set forth at the start of this work. The fact presents itself that large differences in properties occur in the residues rather than in the extracts. It is seen that a larger percentage of C02 is present in the oxidized residue than in that of the fresh, about the same per cent of hydrogen, and also the same per cent of par- •• . . . < • ■ • . . , < •• ' * -X . * < , . . ( t . ( 19 ) affins. The extracts show almost the same percentages of each of these constituents. Forter and Taylor? argued that the resinic part decomposes largely to give the paraffin hydrocarbons and that it may give some hydrogen as a direct decomposition product. The ex- tracts do show tfery much larger percentages of the paraffin hy- drocarbons than their coresponding residues and they do ha foe a fairly large hydrogen content. Wheeler 0 claimed that the resinic would not give hydrogen as a decomposition product. We note also that Wheeler, Porter, and Taylor agree that the carbon monoxide comes from the cellulosic portion. The data shows that though it is not great in either case that it is greater by far in the gas from the residues than in the gas from the extracts. The coke de- rived from the*tar in the carbonizat ion test was very much like a g pitch coke and so the pitch structure idea of Lewes fOr coking has some basis. It is also to be noted that in the distillation of the residues tar is still secured. This would indicate that the resinic material has not been all removed and if this is so, then the differences already so pronounced between the residues and extracts, would be greatly increased if the extraction were complete. The alcoholic potash extract of the coal coked. After this extract had been acted upon by di-phenyl ether it did not coke. This indicates that the KOH solution removes at least some of the coking constituent and also some other material. This added material is no all mineral for the ash content of the extract is low. :l 5 * extract. , - ' . . . ■ . ( 20 ) V. CONCLUSION. 1. Di-phenyl ether removes a part at least of the coking con- stituent of coal. 2 . Di-phenyl ether is "better in some respects as a solvent than phenol or pyridine. It is easier to work with an ci to wash from the residue than is phenol. It does not form addition products with the coal substance as does pyridine. It does not extract a great deal of the coal and so its usefulness as a solvent is limited in this respect . 3. Oxidation destroys the coking property of the coal. 4. Oxidation decreases the heat of combustion of the coal. 5. Oxidation affects the cellulosic constituent and the effect of oxidation is therefore best noticed thru a study of the residue. 6. Differences between the fresh and the oxidized coal in both ultimate andlysdbs and carbonization work, occur similaarly in the residues from the fresh and the oxidized coal. The extracts from the two are fundamentally alike. 7. The use of a double extraction with di-phenyl ether and a solution of 10 % KOH in absolute alcohol does not indicate that it will give a great deal of information regarding the constitution of coal. ♦ - . . * . • < • . . ' ( 21 ) VI. BIBLIOGRAPHY. 1. White and Thiessen, Bull. 38 , Bureau of Mines. 2. Cross and Eevan, Phil. Mag. 352,1882 (abstracted in Bull No. 24, U.I. Exp. Sta. page 4.) 3. Anderson and Roberts, Jour. Soc. Chen. Ind. , 1013, 1918. 4. Parr and Francis, Bull. No. (The modification of Illinoi 24, U. s coal I . Exp . by low . Sta. temperature he at ing) c , • David White , Bull. No. 29, Bureau of Mines, 1911. 6. Burgess and Wheeler, Jour. Chem Soc . V9 7, page 1917, year 1910 h (i 11 11 11 11 V99, " 649 ” 1911 tt 11 11 it 11 11 V105 " 131 " 1914 Clark and Wheeler ti 11 11 V 103 " 1705 ” 1913 7. Porter and Taylor, Tech. Paper No. 140, Bureau of Mines, 1916 . 8 . Lewes, text ’ The Carbonization of Coal* 1914. 9. White, U.S. Geol Paper 85®, pp 65-83- 11C. Dr. Harger, Gas World V60, page 195 » 1914. 11 . Parr and Olin, Bull. No. 79, U.I. Eng. Exp. Sta., 1915* 12. W.A. Bone, text ’ Coal and Its Scientific Uses’. 13. Prof. Bedson, J.S.C.I. V27, 147, 1908. 14. Dr. Harger, J.S.C.I., page 389, year 1914. 15. P.F.Reinsch, J.C.S., No. 48, page 876 , year 1885* 16 . Wahl, Compt. Rendu, No. 154, page 1094, year 1912. 17- Lewes, Progressive Age, No 29, page 1030 , year 1911. 18. Erazer and Hoffman, Tech. Paper No. 5, Bureau of Mines, 1912. 19. Parr and Hadley, Bull. no. 76 . , Eng. Exp. Sta. , 1914. 20. Cherry, Theses for B.S. Degree, Uof I. L 92 C 21. W.A. Bone, text ’Coal and Its Scientific Uses’. 22. Fischer and Groppel, J.S.C.I., No. 38 , 400a, 1919* ( 22 ) • CM CM Fisher and Gludd, Ber . No. ^9, page 1469-71, 1910 11 it ii n ii No . 49, page 1460-9 1916 i« H it ii No. 49, page 1472-4 1916 24. Hobart , Theses for M.S. Degree , U. I. 19^1.