“THE PRIMARY DECOMPOSITION PRO- DUCTS OF PURE CELLULOSE AND CELLULOSIC RESIDUES OF WOOD AND COAL” BY HARLEY LILLARD KNAUER THESIS FOR THE DEGREE OF BACHELOR OF SCIENCE CHEMICAL ENGINEERING COLLEGE OF LIBERAL ARTS AND SCIENCES UNIVERSITY OF ILLINOIS Digitized by the Internet Archive in 2016 https://archive.org/details/primarydecomposiOOknau /922 K 72 UNIVERSITY OF ILLINOIS THIS IS TO CERTIFY THAT THE THESIS PREPARED UNDER MY SUPERVISION BY ENTITLED. IS APPROVED BY ME AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE Approved :. ACTING HEAD OF DEPARTMENT OF ..CHEMISTRY. ACEH0V7LEDGMEKT I take this opportunity to thank Dr. T. E. Layng for the valuable sug- gestions and assistance which I received from him during the course of this inves- tigation. . Table of Contents I. Introduction. 1. Purpose of the Investigation . 1 2. Historical . 1 2. Outline of the Investigation ....... 4 II. Experimental . 1. Apparatus •••.••• 5 2. Operation 6 3. Determination and Analysis of Products. ... 7 4. Preparation of Materials 8 III* Results. 1. Tables and graphs 10 2. Discussion of results 16 17. Summary and Conclusions .19 THE PRIMARY DECOMPOSITION PRODUCTS OF PURE CELLULOSE AND CELLULOSIC RESIDUES OF WOOD AND COAL I. Introduction 1. Purpose of the investigation. The purpose of this investigation has been to determine the primary dec- omposition products of pure cellulose and the cellulosic derivatives of wood and coal at succesive temperatures from 250° - 4i50° C. 2. Historical. Coal, according to the most prominent investigators is made up of three main classes of constituents, namely, cellulosic, resinic, and nitrogen-bearing protein. When coal is carbonized it gives off gases such as: COg, CO, H^, CH 4 , etc., which vary in amounts according to temperature, rate of heating, and kind of coal used. There has been quite a lot of theorizing among coal investigators as to what the primary products of decomposition of each of tnese main constit- uents is when the coal is carbonized. And on this question they have disagreed, sometimes very widely. First we will take up a discussion of the work of Cross and Bevanl, who are probably the most prominent investigators of cellulose today. "By the dir- ect action of heat upon celluloses, aromatic hydrocarbons and phenols are obtain- ed in small quantities. The main products are: (1) gases; CO^, CO, and CH 4 , ( 2 ) liquids; water, acetic acid, furfural, methyl alcohol, and small quantities of hydrocarbons and phenols, (3) solids; paraffins, and aromatic hydrocarbons in small amounts, and residual charcoal. The proportions of each vary with the conditions of distillation, chiefly, rapidity of heating, and maximum temperature attained. In the recent investigations, in which these conditions have been carefully regulated, cellulose has been used in a broad sense, and the destruct- 1. Reasearches on Cellulose. Cross and Bevan 2 ive distillation of cellulose has included wood.” They obtained 35$ of the orig- inal sample as charcoal, about 50$ as distillate, and about 15$ as gas; 50$ of this gas being C0 2 , and 35$ of it 00. The aqueous distillate contained 6$ acetic acid, 30$ methyl spirit, and the remainder, tar. Finely divided filter paper was distilled in quantities of 1 lqg. from a copper retort by Erdmann and Schaeffer^ - . The heating was continued about two hours. The distillate was passed through a train of vessels cooled by air, water, and liquid air, in the order given. The paper contained 35$ of fat which gave rise to the palmitic acid found in the distillate. The uncondensed gases contained: co 2 , CO, CH4, Hg, and N 2 . The portion condensed by liquid air con- sisted of COo,, unsaturated hydrocarbons, and acetone. The tar amounted to about 4$ of the paper. The aqueous portion of' the distillate contained H 2 S, HCHO, fur- furaldehyde, maltol, hydroxymethyl- furfuraldehyde , and valeroacetone. Klason, Heidenstam, and Norlin2 distilled cellulose to 500° C. leaving a char resembling bituminous coal, acetic acid, acetone, and only traces of methyl alcohol. The gases were 15$ of the cellulose, and consisted of C0 2 , CO, CH4, and C^Hg. No hydrogen nor any of the aromatic hydrocarbons were produced. Bantlin 3 carbonized pure cellulose, and observed the following data. The evolution of gas begins at about 140° - 160° C., and at this point C0 2 pre- dominates. This condition continues to a point somewhere between 350° C. and 4000 C. when the proportions change, and the carbon dioxide begins to fall off. At the 5000 mark the evolution of C0 2 is practically ended. Klason, Heidenstam, and Norlin observed a point of exothermic heat at 270° C. Bantlin did not ver- ify this statement. In the table below is given a summary comparison of the work of Klason, Heidenstam, and Norlin, and Sacharoff and Bantlin. Each used a temperature of 5000 C. 1. Ber. 43 2398 2406 (1910) 2. Cham. Abstracts 2: 3280 (1908) 3. Jour, ftfr Gasbeleuchtung 57 1914 pp. 33 - 41 . 3 Products KLason, Heidenstam and ITorlin Sacharoff Bant 1 in Solids and liquids charcoal 38.82% 32.05% 32.97 % water 34.52 31.17 31.67 tar 4.18 5.46 3.25 acetic acid 1.39 4.27 3.28 aldehyde 5.14 8.91 5.82 ketone 0.07 0.50 0.11 Oases co 2 10.35 12.71 11.26 Ethylene series 0.17 0.17 0.24 h 2 0.01 0.02 CO 4.15 3.27 4.78 ch 4 0.30 0.35 0.27 0.48 0.68 100.00 100.00 100.00 An analysis of Bant 1 in’s gas was as follows: carbon dioxide 14.98% oxygen 1.64 unsaturated series 2.72 aromatics hydrogen 4.95 carbon monoxide 47.21 ethane 3.30 methane 16.23 nitrogen 8.97 100.00 H. Soilings and Wm. CotrtA found that cellulose shows a strong exothermic reaction starting at 345° C., hut this is very much weakened in dehydrated cell- ulose and lignite and absent in coal. ”It is presumably connected with the loss of hydroxyl groups and consequent condensation in the residue.” Burgess and Wheeler 2 , from the results of their work on coal, drew the following conclusion: ’’That coal is made up of two distinctly different parts or classes of compounds of different degrees of ease of decomposition, the least stable yielding paraffin hydrocarbons and no hydrogen, and the other decomposing with greater difficulty, and yielding hydrogen and the oxides of carbon”. I. Jour, of the Chem. Soc. 107, pp. 1106 - 1115 2. Jour, of the Chem. Soc. 97, pp. 1017 - 1035 (1910) . • . 4 Porter and Taylor^, two American investigators, do not agree with the above idea that the rapid increase in the yield of hydrogen marks the end of the decomposition of the less stable, paraffin yielding resinic constituent and the beginning of the decomposition of the hydrogen yielding cellulosic constituent. They believe that the first decomposition of any type of coal as the temperature is raised is the breaking down of certain oxygen-bearing substances, related to cellulose, from which water of decomposition, and the oxides of carbon is produced. Other decompositions producing paraffin hydrocarbons, both liquid and gaseous, begin at an early stage. Whether or not such decompositions become the predomin- ating type below 500° C. depends upon the character of the coal, and, as a rule, the higher the oxygen in the coal, the less will be the proportions of hydrocar- bons in the tar and volatile matter. In the case of a sub-bituminous coal the carbon dioxide, water yielding reaction predominates up to 450° C. Secondary decompositions occur easily at 730° C. and above, and the liquid hydrocarbons are broken down to give hydrogen, methane, ethane, ethylene, other hydrocarbons, and carbon. They also believe that the cellulosic derivatives decompose more easily than the resinic and yield water, oxides of carbon, and hydro carbons, giving less of the first the more saturated they are. The resinous derivatives decompose on moderate heating so as to yield principally the paraffin hydrocarbons, and prob- ably hydrogen as a direct product. 3. Outline of the present investigation. It has been the object of this investigation to obtain the primary pro- ducts of decomposition without having any secondary decomposition take place due to the products passing over surfaces heated to a higher temperature than that at which they were evolved. A study of the products of the destructive distillation has envolved the following points: (a) Preparation of pure cellulose, and as nearly as possible the 1. Tech. Paper Ho. 140, U. S. Bureau of Mines. . ' 5 cellulosic derivatives of wood and coal. (b) Determination of tne we ignis ana volumes ox one prouucis of tne ux3oxflaoxun. (c) Ultimate analysis of the starting materials. (d) Analysis and calculation of the total yield of the gases ev- olved. (e) Regulation of the rate of heating, and the temperature attained during the caroonizatiun, ana the noting of the temperature at which any active chemical reaction tabes piace witnin the material being carbonized. II. Experimental 1. Apparatus. A diagram or tne apparatus which was used in this investigation is shown in Fig. 1. The retort was of Pyrex tubing, and nad a sligntiy inclined delivery tube near the top, and a tube near tne Dot tom for passing- nitrogen into the appar- atus. The retort was closed at the top by a rubber stopper carrying a Pyrex tube sealed at the lower end. This tube extended down almost to the bottom of the retort, and in it was placed a 510 degree thermometer. The rubber stopper was protected from the heat by two aluminium discs placed around the tube which held tne thermometer near the top of the retort. The bottom of the retort was covered witn glass wool to prevent the loss of sample through the tube at the bottom. The test from 4500 C. to 750° C. was made in an iron retort already in the laboratory. It consisted of a capped, two-inch iron pip© containing a ther- mocouple tuDe extending down aimost to tne bottom of the retort. The retort was fitted with a l/8 inch pipe delivery tube at the top, and a l/a inch tube at the bottom for the introduction of nitrogen. The pipe was cut in two, and fitted with a flange union which was bolted together with l/2 inch bolts, and sealed by an asbestos gasket. The retort was found to be gas-tight by testing same under , . . water 6 The inside and outside temperatures were measured by a chromel-alumel thermocouple which was standardized against molten lead and tin. It was found to be sensitive and accurate to 4° C. The retort was heated by a ni-chrome wound, electric resistance furnace which was insulated and packed with asbestos, and sil-o-cell. An external res- istance was used to regulate the temperature. Nitrogen gas was passed into the apparatus from a 12 liter aspirator bottle. It was dried by passing through a sulfuric acid wash bottle. The 100 c. c. distilling flask at the end of the side arm delivery tube collected the water, tar, and oil. Next in line was a calcium chloride tube with glass wool in it to catch the tar fog. After that came a calcium chloride tube wnich dried the gas. A mercury manometer next indicated the pressure in the retort. A U-tube half full of glass beads and 10$ sulfuric acid served to remove any ammonia and to indicate when gas was coming over. The gas was then collected in a 12 liter aspirator bottle over saturated salt solution. The Dottle was graduated to 25 c. c., ana. the gas was measured under atmospheric pressure after each run by means of a second aspirator bottle which served as a levelling Dottle. 2. Operation. Before each test the apparatus was swept out with nitrogen, the air ex- hausted uy a vacuum at the delivery end, and the nitrogen let in to take the place of the exhausted air. The sample used was 33 l/3 grams. The collecting bottles were hooked on by means of a rubber tube, and the current turned on. The level- ing bottle v/as kept at a slightly lower level than the collecting bottle through- out the run so as to maintain a slight vacuum, and thus remove the gases as soon as they were formed. The retort was heated at practically the same rate each time, the inside alv/ays being about 75° C. cooler than the outside till the final temperature was reaoned. When the final temperature was attained it v/as held by . * ■ . . 7 means of the external resistance till all the gas was expelled from the material in the retort. The complete run required from 2 to 3jr hours depending upon the material used, and the final temperature attained. At the end of each run the apparatus was swept; out with one liter or dry nitrogen to insure tne removal of the last traces of evolved gas. The stopcock on the collecting bottle was then closed, the bottles leveled and the reading taken. The inside and outside temp- eratures were read by thermometers. A piece of asuestos uoard with a hole through which the retort passed covered the top of tne furnace. The portion of the retort extending out of the furnace was wrapped with asbestos. 3. Determination and Analysis of Products. Oil, tar, and water were determined by weighing the flask before and after tne run. The residue was determined by weight. Gas was collected and measured as described above. It was analyzed in a modified Orsat apparatus of the type in use at the University of Illinois. First COg was removed by 40$ KOH, 0g was taken out by passing the gas into alk- aline pyrogallol solution. The ethylene series was taken out oy bromine water, the aromatics by absorption in fuming H 2 SO 4 , hydrogen and carbon monoxide by pass- ing the gas through a copper oxiae furnace heated to 300° C. The H 2 was oxidized to HgO, and determined by the amount of contraction, and the CO was burnt to COg, which was absorbed in the KOH pipette. Finally the metiiane series was burned with an excess of oxygen in a slaw combustion pipette, and the volume of methane and ethane determined by the method of Parr. 2 x the contraction - CO? m volume (of methane and ethane) 3 CO? n I vol. (n - 1 ) x vol. = ethane vol. - ethane - methane. Nitrogen was determined by difference. Due to the small amounts of oil and tar obtained no analysis was made of them. ■ 8 4. Preparation of Materials. (a) . Preparation of pure cellulose. 300 grams of cotton batting v/ere purified, and. freed, from resins and oils by soiling in a 5 liter flask with. 5 $ CaOH, and then neutralizing with HOI. The product was then thoroughly washed and dried. This is after the method of Cross and Bevan. The pure product was stored in a sealed jar under nitrogen. (b) . Preparation of Cellulosic Residues of Coal. 200 grains of finely powdered coal were extracted for 40 -50 hours with one liter of 10$ alcoholic K0H tinder a reflux condensor. The flask was heated by a water bath. The contents of the flask were then allowed to cool, and the residue was filtered off, and washed with alcohol, water, and ether in a Buchner funnel. It was then transferred to a large evaporating dish and the entrapped K0H neutralized with dilute HC1. The product was again filtered* and washed with alcohol and water. It was then dried at 105° C. under nitrogen, and stored in a sealed jar under nitrogen. The filtrate containing tne extract was then concentrated by distilling off most of the alcohol and water. It was then neutralized and precipitated by dilute HC1. The extract v/as filtered, dried, and weighed. This method of ex- traction saponified, and dissolved the resinous material from the coal leaving the cellulosic materials behind. It did not, however, alter the appearance of the coal. The Utah coal residue from diphenyl ether extraction was obtained from J. R. Smith who was working in the same laboratory at the time of this investi- gation. The oak shavings were prepared by planing down an oak board into small shavings. They were then thoroughly air-dried. The 4b0 degree cellulose was prepared by heating in the retort the pure cellulose to 450° C. until all the gaseous products were expelled. It was at once transferred to the iron retort, and carbonized up to 750 degrees C. . ' 9 The following are the ultimate analyses of the materials used in the in- vestigation: Pure Cellulose (as rec'd) 450 ° Cellulose (as rec'd) Utah Residue (dry) Zeigler Residue (as rec'd) Lignite Residue (as rec'd) Oak Shavings (as rec'd) Carbon 46.82 71.52 71.71 69.45 67.08 49.04 Hydrogen 5.83 5.52 5.84 4.45 2.14 5.85 Oxygen 43.02 14.31 13.96 14.66 14.12 35.38 Nitrogen 0.00 0.00 1.32 1.40 1.62 0.11 Sulfur 0.00 0.00 0.45 0.71 0.86 0.56 Ash 0.48 1.14 6.72 6.18 3.59 0.38 Moisture — a «. a £. 100.00 4.20 100.00 0.00 100.00 3.15 100.00 10.56 100.00 8.68 100.00 B. T. U. 7,092 12,465 12,965 11,765 10,050 8,062 . . . . III. Results. Table I. Carbonization of Pure Cellulose Degrees centigrade 250 300 350 400 450 10 Percentages COg 59.14 53.20 70.40 60.80 54.05 °2 13.41 10.88 1.15 1.15 .90 CH 2n .42 1.48 1.73 1.95 Aromatics H 2 27.45 35.50 25.22 34.73 37.20 CO . CD IV) .15 1.50 ch 4 .93 1.44 5.40 100.00 100.00 100.00 100.00 100.00 Table II. Degrees centigrade 250 300 350 400 450 Volume gas in c. c. C0 2 418.0 1014.0 1785.0 1875.0 1825.0 per 100 gr. of sample. q 94.5 207.0 37.3 35.6 30.3 °n H 2n 8.0 29.0 53.4 65.6 Aromatics H 2 201.0 677.0 639.0 1073.0 1253.0 CO 20.8 4.5 50.5 CH 4 24.9 44.5 182.0 713.5 1906.0 2566.0 3086.0 3406.4 Table III. Degrees centigrade 250 300 350 400 450 Liquid Distillate, grams 8.4 24.9 31.5 33.1 36.1 Charred Residue, grams 72.2 54.0 46.6 42.1 37.6 . . • , Table 17 Carbonization of 450° Cellulose Temperature 450° - 750° C. % Composition COg 16.58 °2 3.62 7.10 arom. Hg 0.00 38.00 ce 10.88 ch 4 22.72 CEH6 1.10 7ol. per 100 g. 2240 488 959 5130 1470 3060 148 Wt. of water and tar per 100 gram of sample 5.0 grams Total volume of gas 13,495 o. c. Table 7. Carbonization of Zeigler 111. Coal after Extraction with 10 % Alcoholic KOH Gas Analysis (nitrogen free basis) Percentage Composition Degrees C. 300 350 400 450 o o ro 84.30 77.60 66.45 65.30 02 2.82 3.87 0.79 2.74 c n E 2n 1.30 2.00 3.77 4.18 arom. 0.32 h 2 1.08 7.46 6.43 5.75 CO 5.30 4 . 48 7.02 5.82 ch 4 1.30 2.00 6.43 7.38 0 2 h 6 3.90 3.59 6.79 8.93 Totals 100.00 100.00 100.00 100.00 Table V. (cont’d) Volume of gas in c. c. per 100 grams Degrees C . 300 350 400 450 C0 2 1185.0 1550.0 1860.0 2200.0 °2 39.5 79.3 22.1 90.5 c nH 2n 18.2 41.0 105.7 138.0 Arom. 8.9 h 2 15.1 153.0 180.0 192.0 CO 74.3 92.0 195.0 192.0 ch 4 18.2 41.0 180.0 244.0 °2K 6 54.6 72.6 190.0 295.0 Totals 1404.9 2028.9 2791.7 3329.5 wt. of water and tar 4.0 5.0 5.0 5.5 per 100 grams Wt. of residue in gr . 90.0 87.0 85.0 84.0 * Table VI Carbonization of Castle Gate Utah Coal after extraction with 10% alcoholic KOH. Gas Analysis (nitrogen free basis) Percentage by volume Degrees C . 300 350 400 450 C0 2 65.43 55.90 54.00 49.90 °2 11.40 23.93 4.24 3.73 c n H 2n 0.42 1.48 6.52 5.37 Arom. 0.43 1.14 0.61 0.67 H 2 18.24 12.96 9.03 8.20 CO 2.28 2.00 3.30 3.18 0H 4 1.80 2.59 14.60 18.40 c 2 e 6 9.50 10.70 Totals 100.00 100.00 100.00 100.00 Vol. of gas in c. c. per 100 gr. Degrees C, . 300 350 400 450 00 2 393.0 672.0 973.0 1100.0 °2 68.4 287.0 76.4 82.1 C n H 2n 2.5 17.8 118.0 118.2 Arom. 2.6 13.7 13.8 14.7 H 2 109.5 155.6 163.0 181.0 CO 13.7 24.0 59.4 70.0 "ch 4 10.8 31.1 263.0 405.0 °2H 6 171.0 236.0 Totals 601.5 1201.2 1867 . 6 2209.0 wt. of water and tar per 100 grams 3.0 5.0 8.0 8.0 Wt. of residue 92.5 90.0 86.0 84.5 . . 14 Table VII. Carbonization of Air Dried Oak Shavings Gas Analysis (H2 free basis) Percentage composition Degrees C. 350 400 450 0 0 tV3 64.10 53.40 52.70 °2 0.92 0.80 0.31 CnHsn 1.71 1.80 1.84 Arom. 0.26 0.26 0.31 H 2 19.55 27.40 26 . 33 CO 13.20 12.69 10.71 ch 4 0.39 3.75 7.16 °2= 6 0.62 Totals 100.00 100.00 100.00 Vol. of gas in c. c. per 100 gr. Degrees C. 350 400 450 C0 2 5620.0 5280.0 6325.0 350 400 -jQ °2 80.6 79.3 37.2 c n H 2n 149.7 179.3 221.0 Arom. 22.8 25.8 37.2 1607.0 2720.0 3160.0 CO 1157.0 1258.0 1287.0 ch 4 34.2 372.0 863.0 C 2 H 6 74.5 Totals 8771.3 9914.4 12004.9 V/t. of water and tar per 100 grams 26.5 37.0 42.0 V/t • of Residue 70.2 61.0 51.4 15 Table VIII. Carbonization of a Weathered Western Lignite after extraction with a 10/£ alcoholic KOH solution. Gas analysis (U 2 free basis) Percentage composition Degrees C . 350 400 450 co 2 86.50 80.90 73.80 °2 2.49 1.92 0.74 ^n^2n 1.03 1.62 1.93 Arom. 0.29 0.31 0.59 H 2 2.64 5.20 4.10 CO 7.05 4.00 8.30 oh 4 2.15 3.42 =2*6 3.20 7.12 Totals 100.00 100.00 100.00 Volume of gas in c. c. per 100 gr. Degrees C . 350 400 450 co 2 3635.0 3885.0 4430.0 °2 104.0 99.2 44.4 C n H 2n 43.3 77.7 116.0 Aram. 12.2 14.9 35.4 h 2 110.0 250.0 246.0 CO 296.0 192.0 500.0 ch 4 105.3 205.0 c 2H 6 154.0 427.0 Totals 4202.1 4775.1 6003.8 wt. of water and tar per 100 grains 14.0 16.0 19.0 Wt. of residue 82.5 79.0 74.2 . CARBOK X I OX'i OX 1 XXJX.X Cii»X XU X O l>— i • QAXIBOI; X i.jii.'j. 1 IOIJ 0^ /juii XltXuxR G OiiXi iiLSS XX)G^-* • CAKBODIZAT I OH OF UTAII GOAL RESIDUE. -iso d>6o © GARB Oil! 2AT I OU OR LIGATED RESIDUE* 16 Table IX Extraction of Coal with 10 % alcoholic KOH Zeigler 111. Coal 8.0% extraction Western lignite 11.0$ extraction 2. Discussion of results. The results of the work are given in the preceding tables and graphs. The gas analyses have been reduced to the nitrogen free basis in order to bring out the relative amounts of the various constituents more sharply and to elimin- ate the error incurred by determining nitrogen by difference. In the case of cellulose, the first evolution of gas occured at 150° - 165° C. This gas was principally carbon dioxide, hydrogen, and a small amount of occluded oxygen. Water in appreciable amounts began to come over about 190° C., which point marks the formation of water of decomposition. Bantlin observed the first evolution of gas at about 140° C. Near 300° C, the evolution of COg took a decided jump, and the hydrogen slowed down a little. It was at this point that the presence of ethylene in the gas was first noted. At 350° C. the carbon dioxide had slowed down to a gradual rise, while the hydrogen had begun to increase. Here the occluded Og was very small, probably due to the fact that it was combining with the free car Don. Bantlin observed this point near 350° C. where the CO^ of the gas grows less. At 350° C., carbon monoxide and methane were first determined. At 400° G, the methane began to rise, and there was pro- bably some ethane given off here but it was too small a quantity to be determined. At 450° C, the carbon dioxide was slumping a little, but had not nearly disappear- ed as was found by Bantlin; the hydrogen was growing less, and the methane and carbon monoxide were rising. The data on the test from 450° - 750° C. shows that COg had sunk 16.58%, and the CO was up to 10.88%. Methane was up to 22.72%, and the hydrogen 38.00%. This verifies the statement of several other investigators that CO and H£ begin 17 to dominate the field at this point. The rapid increase in the carbon dioxide between 300° C. and 350° C, corroborated the statement of Hollings and Cobb that there is an exothermic re- action at this poiht, and a change in the type of decomposition which occurs. The aqueous distillates were analyzed qualitatively, and found to contain water, methyl alcohol, acetone, and some tarry matter. A quantitative estimation of these products was not made due to the small amounts of eacn yield. Klason, Heidenstam, and Norlin found no hydrogen nor metixyi -.loonoi wnic-u. is ausoiuoeiy contrary to one results ox tais woric. Bantlin obtained a great deal more caroon monoxide, and about naif as much hydrogen. The test from 4t>00 C. to 730° C is significant because of the small amount of water or tarry matter evolved. It shows that the principal decompos- itions of cellulose occur below 500° C* In the carbonization of the coal residues, CO2 was the predominating gas throughout. In the Zeigler coal, CO and CH4 snowed no rapid rise. Hydrogen arose abruptly from 300° C. to 350° C, and then almost stopped. The ethylene series began to rise rapidly at 350° C. Ethane rose gradually throughout. In the Utah coal, residue Kg rose gradually. CH4 jumped up at 350° C. Ethane did not appear in any amount till 400° C. Ethylene series rose till 400° C, and then practically ceased. The aromatic hydrocarbons were so slight that they were not plotted on the graph. In the case of the lignite residue, the C0 o was unusually high, indicat- ing that the coal was highly weathered. Hg rose rapidly from 3500 - 400° C. CH^ appeared at 400° C., and rose rapidly. The ethylene series showed a gradual rise throughout. Aromatics were plotted in this case, and increased steadily to 450° C. CO and ethane increased slowly. The lignite coal residue gave off more gas than either of the other two. Nowhere in the literature is tnere any absol- ute data on the carbonization of a coal residue. Burgess and Wheeler state that the cellulosic constituent decomposes with difficulty, but it was found in tiie . 18 present investigation that they decompose with ease throughout the range of temp- eratures used. Porter and Taylor believe that the first decomposition in coal carbonization is the breaking down of the oxygen bearing ceilulosic derivatives from which water of decomposition, and the oxides of carbon are the main products. This is undoubtedly true. However, their statements would lead one to believe that they think most of the hydrogen produced in the early decompositions come from the resinic portions. Prom a study of tne tables and graphs it is evident that hydrogen is produced in appreciable quantities from the decomposition of the ceilulosic substances, and also some of the hydrocarbons. Carbon dioxide predominates, also, in the carbonization of the oak shav- ings. Here the quantity of gas evolved was much larger than that from any other material worked with. Hg, CH 4 , ethylene, and aromatics rose gradually. CO remained practically constant. Ethane did not appear until 450° C., and then only in small amounts. The quantity of aqueous distillate was greater in the case of the oak shavings, cellulose came next, and che yield of this product from the three coal residues was practically the same. In several instances in this work, one will notice that the amount of a gas given off at one temperature is about the same, and sometimes even less than that given off at a lower temperature. This is due to slight variations in the rate of heating to the different degrees of exhaustion of atmospheric oxygen from the apparatus, and in the case of the coal residues, to unavoidable oxidation of certain portions of them in handling, wasning, and drying. The figures given in this data are by no means absolutely accurate ex- pressions of the amounts of each substance which will be evolved in carbonization. Several tests were made to discover whether or not absolute checks could be made on a carbonization at the same final temperature, and it was found that only approximate checks were possible. Better checks, no doubt, could be made by us- ing a larger apparatus, and larger quantities of materials. 1 , 19 17. Summary and Conclusions. The principal gas in the carbonization of cellulose, and cellulosic res- idues up to 450° C. is CC> 2 . Hydrogen is produced in appreciable amounts from all these materials, but more so from the pure cellulose, than from its degrada- tion products. Paraffin, hydrocarbons, chiefly methane, are evolved from pure cellulose, and as we procedd down the list of cellulosic materials which have been altered more and more the methane gives way to increasing yields of ethane, and probably some of the higher members of the series. Aromatic hydrocarbons are not produced in the carbonization of cellulose. They appear in the gases from wood, and lignite, and then decrease in the gases from bituminous coals. Here tney appear as higher members of the series, and in the tar. This might indicate that cellulose in its cnange into the cellulosic derivatives of bituminous coal is arranged inco compounds which give aromatic hydrocarbons on carbonization. Prom the yields of aqueous distillates, it is evident tnat cellulose and wood cel- lulose are more hydrated tnan hie coal residues. The following conclusions have been drawn: 1. That tne cellulosic constituent is being decomposed throughout the coicing process, and is responsible for most of the hydrogen of the gas, the car- bon dioxide, and a large part of the hydrocarbons. 2. That there is a point at about 350° C. in which the decomposition of the pure cellulose rapidly increases, and new types of decomposition begin. 3. That the products of the more highly altered cellulosic materials are of the same series or type as those from the less altered ones.