V 
 
 * 
 
 
THE THERMAL BEHAVIOR OF ILLINOIS COALS IN THE LOW 
 TEMPERATURE CARBONIZATION PROCESS 
 
 BY 
 
 BENJAMIN RACZKOWSKI HARRIS 
 B. S., College of the City of New York, 1917 
 
 THESIS 
 
 Submitted in Partial Fulfillment of the Requirements for the 
 
 Degree of 
 
 MASTER OF SCIENCE 
 
 IN 
 
 CHEMISTRY 
 
 IN 
 
 THE GRADUATE SCHOOL 
 
 OF THE 
 
 UNIVERSITY OF ILLINOIS 
 
 1921 
 
Digitized by the Internet Archive 
 in 2016 
 
 https://archive.org/details/thermalbehavioroOOharr 
 
y\i<v 
 
 UNIVERSITY OF ILLINOIS 
 
 THE GRADUATE SCHOOL 
 
 June 4 , 
 
 OC 
 
 I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY 
 
 SUPERVISION BY Benjam i n Raczkowski Harris 
 
 ENTITLED "The Thermal Behavior of Illinois Co a ls in t he 
 
 Low Temperatu r e Ca rbo nizati on Proc ess " 
 
 BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR 
 THE DEGREE OF faster of Science in Chemistry 
 
 In Charge of Thesis 
 
 ~XP 
 
 Head of Department 
 
 Recommendation concurred in* 
 
 Committee 
 
 on 
 
 Final Examination* 
 
 *Required for doctor’s degree but not for master’s 
 
 A qq 
 

 
 
 
 
 
 - 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 . . 
 
 t 
 
 
 y 
 
 
 
 
 
 
 
1 
 
 ACKNOWLEDGMENT 
 
 The problem was initiated by Professor S. W. Parr. I am 
 greatly indebted to him not only for criticism and guidance but 
 even more for the many ingenious instruments which he has devised, 
 especially the adiabatic calorimeter, without which the inves- 
 tigation could not have been possible. 
 
 Thanks are due to Mr. C. E. Hollister, the department mechan- 
 ician, who made all the special apparatus and originated many of 
 its features and to Mr. Paul Anders who blew the glass apparatus. 
 
 Finally, I take pleasure in expressing my appreciation for 
 valuable assistance, in the manipulation of apparatus, rendered 
 by Messrs. J. M. Lindgren, H. D. Carter, F. B. Hobart and others. 
 
 It goes without saying, that I deeply feel my obligation to 
 the great host of painstaking and devoted workers who have gone 
 before, who have laid the cornerstones and erected the structure 
 of our exact sciences. In particular, to that inspired sage and 
 seer, Michael Faraday, I humbly bow my head in reverence and 
 gratitude J 
 
 In this spirit, I offer the modest tract which follows. 
 
2 
 
 CONTENTS 
 
 PAGE 
 
 Acknowledgment . 
 
 1* Introduction 3 
 
 a. History 
 
 b. Statement of the problem 
 
 2. Critical Discussion of the Available Methods with Special 
 
 Reference to the Method Used in the Present Investigation 
 
 3. Design and Arrangement of Apparatus 10 
 
 4. Theory of Measurements and Calculati ons--Procedure ... 22 
 
 5. History, Preparation and Analysis of Samples 30 
 
 6. Experimental Da ta--Di scus si on . .33 
 
 7* Summary 39 
 
 8. Bibliography ..................... .40 
 
 9. Appendix 41 
 
 PLATES 
 
 NO. 
 
 I. " GRID " T5 
 
 II. " CARBONIZER "...... 16 
 
 III. " CARBONIZER ", disassembled 17 
 
 IV. Complete Set-up ........ 18 
 
 17 • Diagrammatic Representation of Electrical Circuits ... 19 
 VI. « CARBONIZER ", Cross section 30 
 
 . 21 
 
 n t 
 
 • •«•••*•••• • i-J 1 . 
 
 TT II." GRID ", Plan and Elevation . . 
 VIII Cross-sect ion Detail of #47 . . 
 

 
 
 
 
 .1 
 
 
3 . 
 
 SECTION 1 
 INTRODUCTION 
 
 The fact that coal undergoes thermal decomposition essential- 
 ly with evolution of heat has teen for some time well known to 
 chemists and coal technologists. Though the magnitude of this 
 quantity of heat has "been the subject of considerable speculation 
 and research, there is still much to be desired as regards the 
 reliability of the data available. 
 
 For a comprehensive history of the problem, the reader is re- 
 
 1 
 
 f erred to an admirable paper by Eollings and Cobb ? since only such 
 reference will be made to individual workers as will be necessary 
 for the intelligent consideration of the questions treated below. 
 
 Included herewith, (SEE APPENDIX), is a table, rather complete 
 it is hoped, of values to be found in the literature, for though 
 Hollings and Cobb do give an exhaustive discussion of past inves- 
 tigations, they have not assembled into a compact form the various 
 figures that are on record. It will suffice at this point to say 
 that the exothermic heat values, thus far reported, vary between 
 1 and l7o of the calorific value of the coal, the variations be- 
 ing due both to the compositions of the coals and the methods 
 employed. Further, such work as has been done since 1914, the 
 date of the Hollings and Cobb publication, will be mentioned so 
 as to bring the history up to date. 
 
 In a paper, misleadingly entitled, "A Comparative Method of 
 
 2 
 
 Determining the Heat of Carbonization of Coal,' 1 exothe rmi c i ty 
 
 is discussed but the paper can in no sense be considered as con- 
 tributing to our knowledge of the exothermic decomposition of 
 

4 
 
 coal. The destructive distillation of lignin, prepared from spruce 
 
 3 
 
 wood sawdust , has "been recently described. The reaction is dis- 
 tinctly exothermic as in the case of wood and cellulose. Ten to 
 15% of the calorific value is the magnitude assigned to the ex- 
 
 4 
 
 othsrmic heat of wood in "Fuel Production and Utilization" (1920). 
 
 5 
 
 A patent, recently granted to 0. F. Stafford, claims the utili- 
 zation of the exothermic heat as a feature of the process. To 
 these may be added two investigations not covered by Hollings and 
 
 Cobb, one before and one after their publication. Exothe rrni c i ty 
 
 6 
 
 was observed by Parr and Francis in connection with another in- 
 vestigation and studied at some length by E. B. Vliet (unpublished 
 
 7 
 
 reports, University of I 1 linoi s , 19 17 and 1918). 
 
 Of peculiar significance is the determination of this exother- 
 mic heat for the low temperature carbonization process developed 
 at the University of Illinois, depending as it does on the auto- 
 genous heating of the coal being carbonized; namely, the coal is 
 brought up to a suitable temperature when it furnishes its own 
 heat exothermically, the whole mass being heated gently and uni- 
 formly, without the violent superheating of and consequent de- 
 composition of products by the walls of the retort or oven. 
 

5 
 
 SECTION 2 
 
 CRITICAL DISCUSSION OF THE AVAILABLE METHODS 
 
 The methods for the measurement of the exothermic heat of de- 
 composition of coal that are available or, to be more accurate, the 
 methods that have so far been used may be conveniently divided into 
 four types. The names are arbitrary, serving merely for the pur- 
 pose of differentiation: 
 
 1. Temperature rise method 
 
 2. Method of heat balances 
 
 3. "Heat loss" method 
 
 4. Direct measurement 
 
 The first was apparently originated by Rollings and Cobb. A 
 crucible containing the coal to be tested and another similar cruci- 
 ble filled with a thermally Inert substance, e.g. coke, are placed 
 side by side in an electrical tube furnace. A thermocouple dips in- 
 to each crucible and the thermocouples are connected in opposition. 
 The furnace is then slowly and steadily heated, an inert atmosphere 
 being maintained in the tube and readings are taken on the differen- 
 tial pyrometer. When these are plotted against actual temperatures 
 of the coke, curves are obtained which indicate the thermal behavior 
 of the coal at various stages of its decomposition. Essentially 
 the same method, but with some modifications, was used by E. B. Vliet 
 (unpublished report, University of Illinois, 1918). The results 
 are purely qualitative and serve merely to show the zones of exo- 
 thermicity, endothermi ci ty and thermal neutrality, with only a very 
 crude indication of the relative intensities. 
 
 The second method pretends to be quantitative and utilizes data 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 • .. * ” 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ‘ 
 
 
 
 
 . . .. i 
 
 
 . 
 
 . 
 
 , • i. 
 
 « 
 
 
 
 
 
 
 i 
 
 *4 - - • 
 
 * 
 
 4 
 
 
 
 
 
 
 
 
6 
 
 obtained from large scale industrial car boni zati on. A balance is 
 struck between (l) the heat furnished to the retort or oven--by the 
 fuel burned to effect carboni zati on--and (2) the heat yielded up 
 by the retort, taking account of the sensible heat of ashes and coke, 
 heat lost from setting by convection and radiation, heat carried 
 out by flue gases and gases of decomposition, etc. The difference 
 thus arrived at--the second quantity being greater than the first— 
 
 8 
 
 is the heat given up by the coal during its decomposition. Euchene 
 9 
 
 and Barnum are the principal contributors in this field. The 
 method is obviously only very crudely quantitative, since many of 
 the items considered are estimations of an unknown degree of ac- 
 curacy. 
 
 The third method, which has by far yielded the most and probably 
 
 10 
 
 also the best information has been worked with chiefly by Mahler, 
 
 11 12 
 
 Constam and Schlapfer, Constam and Kolbe and S. W. Parr and 
 T. E. Layng (unpublished researches, University of Illinois). The 
 combined calorific value of gas, tar, pitch and coke subtracted from 
 the calorific value of the coal, from which they are obtained, gives 
 a difference which is the "he at loss," i.e., the heat liberated 
 by the coal while it is being carbonized. 
 
 It should be borne in mind, that in these quantitative methods, 
 as well as in the method employed in the present investigation, 
 to be described below, all determinations are made at room tempera- 
 ture and to arrive at the exothermic heat actually delivered at 
 the more or less elevated temperature of decomposition, a calcula- 
 tion would have tc be made according to Kirchoff's Law, taking in- 
 to account the specific heats of the reactants and resultants. 
 
7 
 
 Manifestly, such a calculation would he extremely difficult, if not 
 impossible, the determination of the various specific heats, let 
 alone their variation with temperature, being a feat in itself. 
 
 The direct measurement method was submitted to a brief test by 
 E. B. Vliet, (unpublished report, University of Illinois, 1917) 
 and abandoned as unfeasible. It consists in coking a weighed mass 
 of coal in an inert atmosphere inside a calorimeter. Heat is sup- 
 plied to the coal electrically and accurately measured; this con- 
 stitutes the energy input. The energy output is obtained from the 
 temperature rise and the water equivalent of the calorimeter sys- 
 tem. The amount by which the energy output exceeds the energy input 
 must obviously come from the coal and represents the exothermic 
 heat. The water equivalent of the system is obtained in the seme 
 way, coke being used instead of coal. In this case, of course, 
 there is no difference to be considered, the electrical energy in- 
 put, converted to calories is divided by the temperature rise; the 
 result is the water equivalent. 
 
 The weak point of the method is that the final result is a small 
 difference between two large quantities, the accuracy necessarily 
 suffering thereby. To put it another way, the quantity sought is 
 in great danger of being "swamped" by the quantities actually meas- 
 ured. A typical case will serve to bring out this point. Under the 
 conditions which obtained in the apparatus used in this investiga- 
 tion, an input of approximately 35,000 calories was required to raise 
 the coal to 500° C and keep it there two minutes. The input was 
 measured electrically (SEE SECTIONS 3 and 4 BELOW) with a probable 
 accuracy of + j?fo , so that the 35,000 was correct to +175 calories. 
 
 The output, with a moderately exothermic coal ? in such a case, would 
 
8 . 
 
 "be about 35,700 calories, measured by the temperature rise of the 
 calorimeter system with an accuracy not worse than +^$. The 
 difference is 700 calories, correct to +175 calories and its per- 
 centage accuracy is only £25$. However, the only other quantita- 
 tive method, is in a measure open to the same criticism, the heat 
 of combustion of the coal and its carbonization products being 
 large quantities and the difference relatively small. Moreover, 
 it involves errors of collecting, sampling and analyzing of the 
 gas, tar, pitch, and coke, something which is avoided in the di- 
 rect measurement method. Further, too much emphasis cannot be 
 laid upon the fact that the water equivalent is more than merely 
 a water equivalent,, it is at the same time a control of the most 
 desirable type, a better than which could not be wished for. 
 Various errors to which the determination would otherwise be sub- 
 ject are eliminated by the water equivalent determination being 
 an exactly parallel experiment to the actual test, practically the 
 only difference being that coal is replaced by a corresponding 
 quantity of thermally inert coke. Just one such error will be 
 considered in detail to show how it is removed by the feature a- 
 bove referred to. SECTION 3, below, will make it appear that the 
 wires that carry the current to the " GRID” which heats the coal, 
 might conceivably carry heat out of the calorimeter water into 
 the air, thereby diminishing the energy output and decreasing the 
 exothermic value. To be sure the wires were made 5 to 6 feet long 
 so as to afford sufficient contact with the water before emerg- 
 ing into the air, but still there might be some heat leakage. 
 Granting this to be the case--the leakage would at all events be 
 small--the significant point to note is that it would be about the 
 
9 
 
 same as in the water equivalent run; the same, in so far as the 
 current, time of passage of current and temperature gradient "be- 
 tween the "GRID” and air were the same. The two sets of condi- 
 tions did correspond very closely in most cases. Now, a careful 
 consideration of the relationships prevailing will show that un- 
 der the conditions specified, such an error and numerous other er- 
 rors, are absorbed by the water equivalent. 
 
 Minor errors not eliminated by the water equivalent are heats 
 of solution of NH 3 and H ? S and other water soluble products of 
 carbonization, escape of heat with gases of carbonization, etc. 
 
 A limitation of the apparatus used, though not of the method, 
 was that it was not cs.pable of treating coal at temperatures much 
 above 500°C. 
 
10 
 
 SECTION 3 
 
 DESIGN AND ARRANGEMENT OE APPARATUS 
 
 NOTE: Parts have teen numbered, in rotation but any one mem- 
 
 ber bears the same number wherever it appears in the 
 PLATES. Arabic numbers with Roman numerals as exponents, 
 referring to the PLATES, will serve to locate the parts 
 under discussion. 
 
 The leading objective in designing the apparatus proper for 
 the decomposition of the coal was economy of heat input so as to 
 render the ratio of exothermic heat to input as large as possi- 
 ble since it is that relation which in a large measure controls 
 the accuracy (SEE SECTION 2). In this connection, three factors 
 were of primary importance: first, low thermal capacity of the 
 
 heating element, i.e., a minimum of material to heat up with the 
 coal, second, good contact between the heating element and the coal 
 and third, a minimum amount of radiation and conduction of heat 
 to parts of the system other than those which it was desired to 
 heat, which in turn depended upon insulation and duration of 
 heating. However, effective insulation rendered the attainment 
 of thermal equilibrium in the system, at the start of an experi- 
 ment, extremely difficult. So that a compromise had to be struck 
 bet ween these two desiderata. 
 
 After considerable trial and tribulation the •'GRID,” PLATES 
 I and VII, was evolved. A slab of soft "Alberene” stone was 
 hacksawed to the shape indicated and 25 slots, 0.008 to 0.009 of 
 an inch wide, were cut into the ends with a circular metal slit- 
 ting saw. The stone teeth so formed were broken out at one end, 
 

11 
 
 (background of PLATE I) and replaced when the ribbon was in posi- 
 tion so as to, allow the ribbon to expand, when heated, longitu- 
 dinally rather than laterally which caused shorting. The actual 
 heating was effected by approximately 95” of chromel ribbon 
 wide and 0.005" thick, resistance per foot 0.442ohms, bent as 
 shown and crimped in the middle of each bend so as to lend suffi- 
 cient rigidity. The brass binding posts 70* and 71*, were slotted 
 with a fine hacksaw, the ends of the ribbon slipped in, the whole 
 drilled through brass and chromel and secured with a copper rivet. 
 One of the grids made had a total resistance of 3.22 ohms, another 
 3.32 ohms. AnAlberene 3tone plate, (not shown), 3 3/8" x 3 3/8" x 
 3/32" with a hole in the center to admit the thermocouple tube, 
 two diagonally opposite corners out out to make room 
 for the binding posts 70*, 71-'- , was used for covering the grid to 
 prevent the powdered coal from blowing out and to keep in the 
 heat . 
 
 The grid was used in conjunction with the decomposition ap- 
 , shown 
 
 paratus /in PLATE II, disassembled in PLATE III, cross-section in 
 PLATE IY, and hereinafter referred to for the sake of brevity as 
 the "CA1B0NIZER. " Elaborations were introduced as the necessity 
 for them arose and space will not be taken to justify them indi- 
 
 dually : 
 
 The carbonizer may well be looked upon as consisting of two 
 parts, the upper and lower halves, held together by six clamps 
 63**, 64**, etc., the junction being rendered gas-tight by the 
 gasket 72**, cut out of sheet rubber 0.025" thick. The upper 
 half carries: 
 
 1. the copper spiral tube, 67**, through which gases 
 

 
 t ' 
 
 
 
 
 
 *-» 
 
 « 
 
 
 
 V 
 
 
 
 1 ■ 
 
 :I J 
 
 
 
12 
 
 pass after the quenching of the grid (SEE SECTION 4). 
 
 2. the hard rubber member 65 1 * » V * which supports the 
 
 binding posts 52** ,53 **of the thermocouple, 73 v * and 
 
 50**, 51** of the leads 48 * * » * * * , 49**>*** which carry the 
 heating current to the grid. 
 
 3. the collar 76** which can be adjusted along the 
 tube 66 ** >^* by means of the set screw 69**. The collar 
 76** engages the arms 57**, 58** and 59** which hold the 
 carbonizer in position in the calorimeter can (not shown), 
 a copper can 9 n x 7” provided with 3 members soldered on 
 its inside on which the projecting ends, of the arms 
 57**, 58** and 59**, rest. The arms 57 and 58 were cut 
 
 in two so that the carbonizer might fit into the museum 
 jar l* v for evacuation and replacement of its atmosphere 
 with nitrogen. When in use, the severed halves of the 
 arms were held in place by the glass tubes 55** and 56**. 
 Attention should also be called to the copper-constantan 
 thermocouple 73 v * and its "Pyrex” protecting tube 41***> v ]r 
 which is embedded in the coal between two ribbons when the 
 apparatus is completely assembled. 75 III 1 VI i S a special 
 nut which fixes the brass tube 66 **>* r * firmly on to the 
 principal member of the upper half of the carbonizer. 
 
 With the exception of numbers 54**, 60**, 61** and 62 1 *, 
 which will be covered in SECTION 4, this completes the 
 description of the upper half of the carbonizer. 
 
 The lower half of the carbonizer carries on the outside: 
 
 1. the rubber tube 46 * for admitting nitrogen. 
 

 T 
 
 
 
 : : 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
13 
 
 2. the #16B. and S. copper wire leads 49 ** » 1 * 1 and 48 1 1 » 
 which enter at 47*** and 47a*** respectively, and 
 
 3. the tube 68 ,111 ,VI (provided with the 30 mesh cop- 
 
 per gauze 74 V*) for the escape of gases of decomposition 
 
 On the interior, the lower half of the carbonizer, presents 
 three M TRAN SITE” pegs, (only two of which are shown 42 ***>^* and 
 44 III, VI) t on which the grid rests. In addition to the pegs, there 
 are two flexible cables, each about l-jj” long, (only one is shown, 
 43*H), the function of which is to complete the circuit between 
 the binding posts 70 * and 71* and the terminals of the leads 48*** 
 49 III, on the interior of the carbonizer. 
 
 PLATES VI and VII show a few additional features and the ac- 
 tual dimensions. In PLATE VI the two halves of the carbonizer 
 are separated and the coil 67* 1 , the leads 48**, 49** and the arms 
 57** 58** 59** are not shown for the sake of simplifying the draw- 
 ing. PLATE VIII is a detail, self-explanatory, of the member 
 47lI,IH, (47a*** is similar to it), and shows the manner in which 
 the leads 48**>*** and 49**>*** enter the lower half of the car- 
 bonizer. 
 
 PLATES IV and V still remain to be described. The former is 
 a photograph of the complete set-up, ready for a run; the latter, 
 a diagrammatic representation of the electrical circuits. The stor- 
 age batteries shown in IV were originally used but later abandoned 
 in favor of 45 or 50 volts taken off the 110 D.C. circuit, 35*^>^ . 
 
 10 *V is a Parr Adiabatic Calorimeter, manufactured by the Standard 
 
 IV 
 
 Calorimeter Co., Moline, 111 .; 33 , the water heater for the cal- 
 
 IV IV o 
 
 orimeter. 6 , 7 Fahrenheit thermometers with a 65 to 90 range, 
 
 calibrated at the Bureau of Standards, (Beckman thermometers of a 
 
14 . 
 
 sufficiently large range were not available). 
 
 11 IV and 12 1 ^ , and a third not shown, are pyrogallol wash- 
 bottles for the purification of the Linde compressed nitrogen used 
 for displacing the air in the carbonizer. 4** is a reservoir for 
 nitrogen and 5 Iv a leveling bottle corresponding to it; 2*^ a dry- 
 ing tower intermediate between 4*^ and 1^ , the museum jar in which 
 the carbonizer was placed for evacuation. 9^ a Dewar flask cold 
 junction for the thermocouple 73^*; 8^^, a Siemens- Hal ske milli- 
 voltmeter used in connection with 73^*. The coulometer 31^^ , ,r 
 (SEE SECTION 4) is a glass jar 8^” x 2" containing: (1) 2 nickel 
 
 gauze electrodes, (6 3/4” x 1 3/4”, each provided with a copper 
 wire lead, riveted on), (2) about 230cc. of 15$ NaOH solution, pre- 
 pared from chlorine -f re e NaOH and (3) about 15cc. of carefully puri- 
 fied mineral oil (boiling between 180 and 220° C) to prevent froth- 
 ing. The coulometer is tightly sealed with a rubber stopper fit- 
 ted with (1) the dropping funnel 32^, for the introduction of 
 water and (2) a delivery tube (not distinctly shown) which is in 
 communi cati on with the gasholders 27 IV and 29 1- ^, of which 28 
 and 30 are the leveling bottles, and (3) the leads to the gauze 
 electrodes. The two-way stop-cock 27a^ permits 27 to be connected 
 with either 31 and 29 or 17*^, a water jacketed 200 cc. measuring 
 burette, for which 18 IV is the leveling tube. 
 
 Other features of PLATE IV will be referred to in SECTION 4. 
 

 
 
 
 
 
 -V - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 c % 
 
 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 • 
 
 
 
 
 
 
 
 
 
 
 1 * 
 
 . 
 
 * 
 
 • • - . 
 
 
 
 
 
16 
 
 PLATE U 
 


 jit j x v i d 
 
32 
 
 PLATE JY 
 

35 
 
 

CAK3CN/Z.T. R 
 

 
 
 
22 
 
 SECTION 4 
 
 THEORY OE MEASUREMENTS AND CALCULATIONS -- PROCEDURE 
 
 The electrical heat input could he measured in a number of 
 ways; after considerable testing of various methods, it was finally 
 decided to adopt the method which requires the measurement of (1) 
 the potential drop across the resistor, (2) the current, (3) dura- 
 tion of flow of current, Elt being the energy in joules, where E 
 represents volts, I, amperes and t, seconds. For the potential 
 measurement a Leeds and Northrup, TYPE 7650, potentiometer, 19^^*“ 
 was used. 20, 21, 22, 23 and 03, PLATES IV and V are accessories 
 to the potentiometer. I and t were measured by means of a water 
 coulometer, 31 IV,;f , reported to be correct to + i-$. A coulometer 
 was preferred to other instruments for two reasons: (l) because 
 
 it is an integrating instrument, a very important consideration, 
 since a more or less constant temperature was maintained in the grid 
 72^ by throwing the switch 13*^’^ in or out, the coulometer regis- 
 tering only when current was passing and adding up the various ele- 
 ments of It (2) because it gives directly the current-time product, 
 whereas if any other instrument were used for the measurement of 
 current, time would have to be measured with a stopwatch or chrono- 
 graph, thereby complicating the apparatus and multiplying the er- 
 rors, especially if time would have to be taken out everytime that 
 current was cut out. 
 
 It was at first intended to use a coulometdr of great er ac- 
 curacy , e.g., a silver or copper voltameter, but the current was so 
 large, 12 to 13 amperes, that cathodes of inconveni ent ly large size 
 would have had to be employed, in order not to exceed the low cur- 
 

23 
 
 rent densities which these instruments call for. 
 
 The mixed gas, from the coulometer, i . e . , the hydrogen and oxy- 
 gen collected in 27^ and 29^ and measured in 17^, was calculated 
 to coulombs with the aid of the following expression: 
 
 v ( b-p ) x 2,0543 == coulombs 
 T 
 
 where v is the measured volume, b the barometric pressure, p the 
 vapor pressure of water at the temperature T and T, the absolute 
 temperature at which the gas is measured. ( See Lehfeldt, "Electro- 
 Chemistry," pg. 6 , Longmans, Green and Co., 1904; the factor there 
 given, 1.8373, is incorrect). 
 
 Now, the product coulombs times volts is the number of joules 
 put in. This divided by 4.184 gives the calories put in and this di 
 vided in turn by the temperature rise, (Centigrade), gives the water 
 equivalent of the system in grams, i.e. when coke is the material 
 treated. In the case of a run with coal, the heat input as above 
 calculated was used as such and compared with the heat output -- 
 which is the product of temperature rise and water equivalent -- 
 to arrive at the exothermic heat. 
 
 To make it possible to multiply the total number of coulombs by 
 volts, it was necessary to keep the potential drop across the grid 
 constant throughtout the experiment, 5 to 6 minutes, for if the vol- 
 tage varied during the run, it would obviously have been necessary 
 to know how many coulombs corresponded to the respective potential 
 drops. Constancy of potential fall was maintained as follows: 5 
 
 or 10 minutes before the heating phase of the experiment, the plate 
 rheostat 16*^'^ was so adjusted, that the potential drop across the 
 external resistance 34^»^, a coil of #18 (B.andS.) ni chrome v/ire, 
 was of such a magni tude- -de te rrnined by a previous calibrati on-- that 
 it corresponded to a potential drop of 50 volts across the grid 72^ 
 
24 
 
 when the rolling contact 25*^ » ^ was about in the middle of the slide 
 wire 24* y > ter this coarse adjustment had been made, fluctuations 
 in the potential drop across the grid were almost entirely eliminat- 
 ed by varying the resistance 24 v , which is in series with the grid. 
 The run, as far as the electrical heating phase of the procedure is 
 concerned was then ready to begin, (the grid having previously been 
 loaded and placed in the calorimeter 10*^, as is to be described be- 
 low). Three observers were required for this phase of the experi- 
 ment. One manipulated the double-pole, double- throw , knife switch, 
 irjIV,V an d took minute readings on the mill ivol tme ter 8*"^. Another 
 maintained the jacket temperature of the calorimeter 10** T parallel 
 with the temperature of the can and a third manipulated the rolling 
 contact along the wire 24* 7 » v in such a manner as to maintain 
 
 a minimum amount of deflection in the galvanometer G2*^» v ,(the po- 
 tentiometer having been set at 50 volts before the start of the ex- 
 periment). The double-pole, double- throw, knife switch 14* V,V was, 
 of course, kept in at the left so that the potentiometer registered 
 potential fall across the grid. 13* v » v7 " was kept in until the grid 
 attained a temperature of 500° C and after that the switch was cut 
 out and in, for two minutes, sufficiently often to maintain the 
 
 O 
 
 temperature in the immediate neighborhood of 500 C. The thermocou- 
 ple readings do not in any way enter into the calorimetric calcula- 
 tions but serve only to indicate up to what temperature the coal has 
 been carbonized. 
 
 To make the arrangement even clearer it should be said that (l) 
 by means of 13*^ ,V current may be fed into either the external resis- 
 tance 34*^’ ^or the grid 72 V , the current passing through 16*^ direct- 
 ly to 34 V but through 16 V and 24^ before it reaches the grid and (2) 
 by means of 14* V,V the potentiometer may used to measure the po- 
 
tential drop across 34* / »'‘ r or the grid, as required. 
 
 23 . 
 
 After the heating has teen completed, only one observer is re- 
 quired. The switch 13 IV, V stays out permanently and, outside of meas 
 uring the coulometer gas, nothing remains hut to get the final temper 
 ature of the calorimeter system; this generally took 45 to 55 minutes 
 during which time, of course, the jacket temperature was constantly 
 equal to that of the can. It should also he noted that during this 
 
 interval, the residual gas in the carbonizer was cooling, contract- 
 
 ing and drawing up water through the gauze 74^1. When the millivolt- 
 
 *r r r O t -r 
 
 meter 8 registered about 125 C, the string 54 1 1 --whi ch protruded 
 through an opening in the cover of the calorimeter lG iV --was pulled. 
 This removed the rubber tip 60**, leaving the end 61** of the coil 
 67 A open to the water of the calorimeter can and allowing (l) the 
 
 escape of gas (trapped in the carhonizer) through the'gauze 40 ‘-^in- 
 
 to the coil 67^*, past the point 62 A *, along the coil and finally 
 out through 61 11 into the water and (2) the thorough quenching of 
 
 the grid, etc., which could not otherwise be effectively done. The 
 I V 
 
 thermometer 6 was then read until a constant final temperature was 
 attained, as stated above. 
 
 An account of that part of the procedure will now be given which 
 precedes the heating stage above described. 20 to 25 grams of 60 
 mesh, air-dry coal, weighed to tenth of a gram, were evenly distri- 
 buted in the grid, the grid having previously been placed in posi- 
 tion on the three pegs, 42*H 44IH in the lower half of the car- 
 bonizer and the leads 43*** screwed into the binding posts 70l, 
 
 71*. A steel spatula was drawn a number of times across the grid 
 in a direction perpendicular to the length of the ribbons; this 
 tended to pack the coal more densely and thereby increased the ca- 
 pacity of the grid. The alberene stone plate, (SEE SECTION 3), was 
 

 . 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 
 
 
 
 
 
 
 * . 
 
 ' 
 
 
 
 \ ■ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 »/ 
 
 c 
 
 
 
 
placed in position and empty spaces in the carbonizer were loose- 
 ly packed with approximately 12 grams of purified, shredded as- 
 bestos, weighed to +-| gram. (In cases where the material did not 
 cake, e.g. the oxygenated coals, (SEE SECTION 6), there was danger 
 of small portions of the powdered coal being blown out of the 
 grid and escaping decomposition. In such cases, special care was 
 taken to produce good contact between the plate and grid by stuff- 
 ing with asbestos. In spite of these precautions small amounts 
 were occasionally blown out and corrections had to be applied). 
 
 The surface 38*** *^"* was greased with vaseline, the rubber 
 gasket (not shown) placed in position, the surface 3 sTII>VI greased 
 and the two halves of the carbonizer assembled and clamped as 
 shown in PLATE II. The outer halves of the arms 57**, 58** and 
 the glass tubes 55**, 56** were removed and, after the rubber tip 
 60**, with the string 54** attached, had been slipped on to the 
 end of the coil 67** at 61** and the junction at 61* 1 coated with 
 paraffin, the carbonizer was placed in the jar 1*^, evacuated 
 twice to 25 millimeters (manometer 3*^) and alternately filled 
 with purified nitrogen from 4*^. The carbonizer was then removed 
 from 1*^, the parts 55, 56, 57 and 58 replaced and the whole put 
 into the calorimeter can (not shown), nitrogen passing into the 
 carbonizer through the tube 46**’*** while it was being handled in 
 the air. 
 
 Enough water was then added so that the can weighed 6700 grams 
 complete, 46** having been disconnected from the nitrogen train 
 and clamped tightly with 45**. The loaded can was placed in the 
 calorimeter 10*^. With the cover of the calorimeter raised about 
 4 inches, the thermocouple leads 77 * V , 78 1 ‘ and the grid leads 
 
' 
 
 . 
 
 
 
 
 
 - 
 
 * 
 
 - 
 
 . 
 
 
 
 
 ■ 
 
 
27 . 
 
 36*^’^, 37*^’^ were drawn through the opening in the cover and 
 screwed into the binding posts 53**, 52^, Sl^" 1 ' and 50*^ respec- 
 tively. The string 54** was then drawn through the opening to 
 make it accessible from the outside, the cover, thermometers and 
 stirrer placed in position, and the apparatus stirred for one 
 hour to allow the system to come to thermal equilibrium. The ini- 
 tial temperature reading was then taken on the thermometer 6*^ 
 and the heating stage, previously described, followed. So much 
 for the actual run. 
 
 For a water equivalent determination, the procedure preceding 
 the heating stage was somewhat different. The can was loaded 
 and connected at 50**, 51**, 52** and 53 * 1 as above but outside 
 the calorimeter 10*^". The coal was then coked, that is the heat- 
 ing stage was followed through, except that no attention was paid 
 to temperature rise or to getting the system into thermal equi- 
 librium at the start. As soon as heating was over, ice was thrown 
 into the can and nitrogen was passed through the carbonizer, by 
 way of 46 **, for about an hour, until the grid had again cooled 
 to room temperature. From then on the apparatus was treated in a 
 manner which coincided exactly with the actual run described a- 
 bove. The difference, then, between the water equivalent deter- 
 mination and the actual run was that the system heated up in the 
 former case was at the start in a condition almost identical with 
 the condition in which the system in the actual run was after the 
 heating had been completed. 
 
 After an experiment, the apparatus was disassembled, cleaned, 
 the coke scraped out of the grid and the grid dried over night at 
 
 O 
 
 125 — 200 C and, when necessary, gently ignited over a Bunsen 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 < > 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
flame 
 
 ZS. 
 
 A leaf from the note "book ie shown below to indicate how 
 the data were entered and calculated. 
 
 p 
 
29 
 
 RUN # S x 55b 
 
 Y/eight of flask coal 83.2 
 
 " 11 " - coal 59.0 
 
 " » coal 24.2 
 
 May 5, 1921 
 
 Weight of "bag asbestos 93 
 
 « " « « 80 
 
 11 " a s be s t o s 13 
 
 Temperature Record 
 
 
 (46” min. ) Final 
 
 82.71 
 
 Initial 
 
 66.81 
 
 Rise 
 
 15.90 
 
 Coulometer Record 
 C.C. temp. 
 
 199.8 22.8 
 160.2 22.8 
 126.4 22.8 
 154.2 22.9 
 
 The rmocouple 
 Record 
 
 VOLTAGE 50.00 
 
 Time in 
 minute s 
 
 Scale 
 
 Reading 
 
 
 Start 
 
 0.45 
 
 
 1 
 
 5.5 
 
 Room temp. 25 C 
 
 2 
 
 10. 1 
 
 
 3 
 
 14.0 
 
 Temperature at 
 5:24 was 72^7 
 
 3:24 
 F 5:24 
 
 15.5 
 
 
 18 
 
 3.8 
 
 
 22 
 
 0.9 
 
 
 33 
 
 0.6 
 
 
 46 
 
 0.6 
 
 pre ss , 
 
 744.9 
 
 CALCULATIONS 
 
 Calibration correction 0.00 
 Stem exposure ” 0.01 
 
 RISE (corrected) 15.91 = 8.8390 C. 
 
 640.6 x 724.1 x 2.0643 
 
 = 3237 coulombs 
 
 295.8 
 
 3237 x 50.00 
 
 - 38685 calories INPUT 
 
 4. 3 84 
 
 8.8390 x 4524.4 = 39991 calories OUTPUT 
 
 39991 
 
 38685 = 1306 calories EXOTHERMIC HEAT 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
30 
 
 SECTION 5 
 
 HISTORY, PREPARATION AND ANALYSIS OF SAMPLES 
 
 Six samples were worked with in all. As far as possible all 
 material was kept in stoppered bottles under nitrogen to prevent 
 weathe r ing . 
 
 SAMPLE #1 was a West Frankfort, Franklin County coal, obtained 
 from Harris, Dillavou and Co. , Champaign, 111. Good \ pound lumps 
 were picked out of a recent shipment, crushed in a Sturtevant mill 
 to ^ inch size and air dried. The air-dry material was reduced to 
 buckwheat size in a coffee mill and ground on a buckboard to pass 
 a 60 mesh sieve. This material was then used for the experimental 
 work. 
 
 SAMPLE #2 was a mixture of Ba(C 103 )., , anhydrous, 1 part and 
 ignited BaSQ^ 10 parts. 
 
 SAMPLE #3 -- Vermillion County coal, from the Sharon Coal Co., 
 Urbana, 111. Good 2 pound lumps were picked from a carload pile, 
 one week old and were treated like SAMPLE #1. 
 
 SAMPLE #4 -- Harrisburg, Saline County coal, secured from 
 Huff and Co., Urbana, 111. Good 2 pound lumps were selected from 
 a bin, which had been stored 2 to 3 weeks. Treatment given was 
 the same as in the case of SAMPLE #1. 
 
 SAMPLE #5 -- weathered Franklin County coal. About 220 grams 
 of air-dry SAMPLE #1, buckwheat size, were placed in an air-drying 
 oven kept at about 35° C. After 6 weeks the coal had lost in weight 
 2.3 grams and still gave a fairly good coke in a crucible test. It 
 was therfore ground to 60 mesh, placed in oven at 105° and raked 
 once a day so as to expose fresh surfaces to the air. After 10 
 
31 
 
 days' heating, it had lost only 0.5 grains, whereas according to the 
 moisture determination (SEE TABLE II, below), it should have lost 
 3.64$. Taking into account the 2.3 grams lost during the 6 weeks'' 
 exposure , it seems that the coal absorbed at 105° C about 2.5$ of 
 oxygen -- possibly also C0 2 and N 2 -- from the air. This oxygena- 
 ted material was submitted to tests in the carbonizer. 
 
 SAMPLE #6 -- weathered Vermillion County coal. In this case 
 a part of 60 mesh, air-dry SAMPLE #3 was placed directly in the 
 oven kept at 105°. A more careful recor'd of weights was kept and 
 is recorded below. 
 
 TABLE I 
 
 SHOWING ABSORPTION OE OXYGEN BY VERMILLION 
 COUNTY COAL AT 105° C 
 
 DURATION OP HEATING WEIGHT OF COAL 
 
 HOURS GRAMS 
 
 Start 
 
 269.7 
 
 2 
 
 254.4 
 
 12 
 
 255.5 
 
 24 
 
 256.5 
 
 38 
 
 256.8 
 
 60 
 
 256.8 
 
 84 
 
 257.6 
 
 108 
 
 259.1 
 
 132 
 
 599.6 
 
 154 
 
 260.1 
 
 176 
 
 260.5 
 
 204 
 
 260.6 
 
 228 
 
 260.8 
 
 According to the moisture determination, (SEE TABLE II), the 
 
 O 
 
 sample should have lost 16.02 grams at 105 and since it lost only 
 8.9, the absorption from the air amounts to 7.12 grams or 2.64$. 
 This is based on the assumption that the coal lost no more weight 
 during ten days' heating, either as hygroscopic moisture or by 
 oxidation of hydrogen, than it lost on one hour's heating at 105° 
 

 
 
 
 
 
 
 
 
 . 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 - - . 
 
 • # 
 
 
 
 
 
 : 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ 
 
 
 
 ■ 
 
 
 
 
 
 
 
 
 » 
 
 
 
 
 
 
 - • 
 
 
 
 
 
32 . 
 
 in a C0 9 atmosphere, the conditions under which moisture loss was 
 determined. Since that is undoubtedly not the case, the 2.64$ is 
 only a minimum, the amount of material actually absorbed being very 
 probably higher. The same criticism is applicable to the calcula- 
 tion made for SAMPLE #5. 
 
 The loss on air-drying was as follows: SAMPLE #1, 5.19$; #3, 
 
 10.48$; #4, 3.5$, 
 
 Table II gives the proximate and ultimate analyses of the 
 samples. All determinations were made in duplicate; single figures 
 indicate calculated percentages. 
 
 TABLE II 
 
 ANALYSES OF SAMPLES 
 
 
 SAMPLE 
 
 MO IS* 
 
 VOLA- 
 
 FIXED 
 
 ASH 
 
 C H 
 
 0 
 
 N 
 
 S CALS. 
 
 BASIS 
 
 NO. 
 
 TURE 
 
 TILE 
 
 CARBON 
 
 
 
 
 
 PER 
 
 
 
 
 MATTER 
 
 
 
 
 
 
 GRAM 
 
 
 1 
 
 3.67 
 
 36.21 
 
 52.54 
 
 7.72 
 
 71.614.97 
 
 9.01 
 
 1.57 
 
 1.26 7156 
 
 
 
 3.61 
 
 36.07 
 
 
 7.63 
 
 72.18 
 
 
 1.51 
 
 1.29 7166 
 
 
 3 
 
 5.98 
 
 37.26 
 
 49.48 
 
 7.43 
 
 69.824.84 
 
 9 . 20 
 
 1.83 
 
 1.07 6833 
 
 AIR- 
 
 
 5.89 
 
 36.95 
 
 
 7.49 
 
 69.57 
 
 
 1,79 
 
 1.03 6821 
 
 DRY 
 
 4 
 
 2.53 
 
 35.16 
 
 53.03 
 
 9.41 
 
 70.645.01 
 
 8.54 
 
 1.85 
 
 2.00 7122 
 
 
 
 2.46 
 
 35.00 
 
 
 9.37 
 
 70.73 
 
 
 1.98 
 
 1.90 7104 
 
 
 1 
 
 • » • • 
 
 37.50 
 
 54.52 
 
 7.96 
 
 74.615.16 
 
 9.35 
 
 1.60 
 
 1.32 7431 
 
 
 3 
 
 • • • • 
 
 3 9.45 
 
 52. 60 
 
 7.93 
 
 74.095.15 
 
 9.78 
 
 1.92 
 
 1.12 7258 
 
 MOISTURE 
 
 
 
 
 
 
 
 
 
 
 FREE 
 
 4 
 
 • • • • 
 
 35.97 
 
 54.38 
 
 9.63 
 
 72.485.14 
 
 8.76 
 
 1.96 
 
 2.00 7294 
 
 
 
 The analyses 
 
 canno t 
 
 be expected to be typical 
 
 of these 
 
 coal s , 
 
 inasmuch as the samples were selected and not representative. 
 

 
 
 
 
 
 
 
 
 
77 
 
 SECTION 6 
 
 EXPERIMENTAL DATA -- DISCUSSION 
 
 It was thought desirable to submit the accuracy of the method 
 to an absolute test by charging the grid with material which was 
 known to yield a definite quantity of exothermic heat. A suitable 
 substance was not easy to find for it had to fulfill a variety of 
 requirements, some of which were: 
 
 1. Sharp decomposition at a temperature not much above 
 500° C . 
 
 2. Decomposition should be complete in 5 or 6 minutes. 
 
 3. No complicating side reactions should occur. 
 
 4. The substance should not melt, for if it did, it 
 would run out of the grid. 
 
 5. The products of decomposition should preferably be 
 solids insoluble in water, etc., etc. 
 
 Ba(Cl 0^)3 diluted wi th Ba SO 4 was finally decided upon. The 
 reaction 
 
 2 Ba (Cl 0 3)0 Ba ( CIO 4 ) g-hBaClg + BOg 
 proceeding rather sharply at 425 C # with a liberation of 92 calories 
 per gram of Ba(C 103 ) 2 . A few experiments were made but the work 
 
 a 
 
 had to be discontinued on account of the corrosive action, at 450C , 
 of the oxygen and a small amount of chlorine which also formed. Ow- 
 ing to these circumstances, the results were not sufficiently reli- 
 able. However, such indications as were obtained, seemed to lead 
 to the conclusion, that the accuracy was probably not better than 
 +20$ or 25$. This agrees fairly well with the estimation of the ac- 
 curacy, made in SECTION 2, on purely hypothetical grounds. 
 
 TABLE III, below, gives the exothermic values of the five 
 
34 . 
 
 coals studied in this investigation. The quantities recorded are 
 means of 4 determinations, with the average deviation from the mean 
 in the extreme right hand column. 
 
 TABLE III 
 
 EXOTHERMIC HEAT VALUES FOR FIVE COALS 
 
 SAMPLE NAME EXOTHERMIC™ HEAT IN a7~d7 _ 
 
 NO. CALORIES PER GRAM IN 
 
 — i 
 
 AIR-DRY COAL MOISTURE 
 FREE COAL 
 
 
 1 
 
 Frankl in 
 
 26 
 
 
 27 
 
 
 22 
 
 
 
 3 
 
 Vermillion 
 
 39 
 
 
 41 
 
 
 8. 
 
 5 
 
 
 4 
 
 Saline 
 
 13 
 
 
 13 
 
 
 16 
 
 
 
 5 
 
 #1 Weathered 
 
 55 
 
 
 57 
 
 
 6. 
 
 3 
 
 
 6 
 
 #3 Weathered 
 
 58 
 
 
 62 
 
 
 13 
 
 
 
 In 
 
 the case of SAMPLE #4, the 
 
 material 
 
 was only raised 
 
 to 
 
 500 
 
 0 
 
 and 
 
 the heating was 
 
 stopped, i. 
 
 e . the 2 
 
 minute per 
 
 i od , 
 
 (SEE 
 
 SECTION 4 
 
 , infra) , had to 
 
 be 
 
 di spens 
 
 ed with b 
 
 ecause of 
 
 the d 
 
 if f i cul by 
 
 exp 
 
 srienced in maintaining at 
 
 500° a 
 
 constant 
 
 potential 
 
 drop 
 
 across 
 
 the 
 
 grid. 
 
 Thi s , however , 
 
 should not 
 
 material 
 
 ly affect 
 
 the r 
 
 e sul t , 
 
 since very little decomposition -- as indicated "by gas evolution -- 
 occurred at 500° as compared with the amount of decomposition up to 
 500° C. 
 
 DISCUSSION 
 
 It is evident from the results that the exothermic heat in- 
 creases wi th oxygen content, "both with the oxygen normally present 
 and with the oxygen taken up on weathering f especially when the 
 coals are compared on the moisture free basis, which is quite ap- 
 
35 
 
 propriate, considering that the material must necessarily he free of 
 its moisture at the more or less elevated decomposition temperatures. 
 For the three coals studied, the exothermic heat, on the moisture 
 free basis, is given roughly by the expression 
 
 12.5 + 28 (n - 8.75) 
 
 where n is the per cent of oxygen on the moisture free basis and the 
 other terms are empirical constants. The oxygen taken up on weather- 
 ing obviously yields much greater quantities of exothermic heat than 
 the oxygen of composition, the ratios for 1$ of oxygen being: for 
 
 Franklin County coal, 12 to 3, for Vermillion County coal ? 8 to 4 
 calories. 
 
 It is interesting to note that though the Franklin and Ver- 
 
 million County coals 
 
 are rather far apart in 
 
 exo thermi ci ty , the 
 
 weathered 
 
 coals corresponding to them 
 
 come very much closer to 
 
 being the 
 
 same . 
 
 
 
 
 The 
 
 ratio of exothermic heat to 
 
 per cent oxygen on the "unit 
 
 coal” basis is interesting, at least 
 
 for the 
 
 oxygenated coals. 
 
 
 
 TABLE IV 
 
 
 
 RATIO OF 
 
 EXOTHERMIC HEAT 
 
 TO PERCENT OXYGEN 
 
 
 
 "UNIT COAL” 
 
 BASIS 
 
 
 SAMPLE 
 
 NO. 
 
 NAME 
 
 EXOTHERMIC 
 HEAT IN 
 CALORIES 
 
 ITo 
 
 CALS. 
 
 $ 0 
 
 1 
 
 Franklin 
 
 31 
 
 10.3 
 
 3.01 
 
 3 
 
 Vermillion 
 
 46 
 
 1C. 8 
 
 4.26 
 
 4 
 
 Saline 
 
 15 
 
 10.1 
 
 1.48 
 
 5 
 
 #1 , We athered 63 
 
 13.2 
 
 4.77 
 
 6 
 
 # 3, Weathered 68 
 
 13.9 
 
 4.88 
 

 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 ■ 
 
 
 
 
 
 
 : 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 " 
 
 
 
 
 
 
 
 
 
 . 
 
 . 
 
 t 
 
 
 
 
 
 
 * 
 
 
 
 
 
 
 
 
 
 
36 
 
 Here, again, the ratios for #1 and #3 are rather far apart, 
 whereas for #5 and #6, the corresponding oxygenated coals, the 
 ratios are remarkably close. 
 
 Heretofore, it has been customary to compare exothermic heats 
 as per cents of the calorific values of the coals. It is difficult 
 to see what importance such a ratio can have. It seems that it 
 would be more instructive to know what part of the heat actually 
 required to coke the coal is furnished by the exothermic reaction. 
 
 By the "heat actually required to coke the coal" is meant the sum 
 of the various quantities used up in vaporizing the liquid products 
 of carbonization and the combined heat capacities of all the pro- 
 ducts. Such a definition is not purely arbitrary. It has a real 
 significance in the low- temperature carbonization process; for, 
 the coal is brought up to a suitable reaction temperature, i.e., 
 a temperature at which a vigorously exothermic stage sets in and 
 from then on, only such a quantity of heat is fed in through the 
 walls of the retort as is necessary to compensate for the loss due 
 to radiation, conduction and heat carried out by volatile products. 
 The actual energy of carbonization, as stated in SECTION 1, is fur- 
 nished primarily by the exothermic heat of the reaction. 
 
 Fortunately, the data taken in connection with SAMPLES 3 and 5, 
 the Vermillion coals, make it possible to estimate the ratio of the 
 exothermic heat to the heat actually consumed in carbonizing the 
 coal. At first hand, it may appear that the difference between 
 the heat input in an actual run and the heat input in a correspond- 
 ing water equivalent determination should represent the heat ac- 
 tually required to carbonize the coal. This would be true if no 
 heat were lost by radiation and conduction. This loss, which must 
 
37 
 
 "be subtracted from the difference referred to above, may be estima- 
 ted from the temperature rise, indicated on thermometer 6 , at 
 
 the time when the heating period was ended* In the case of SAMPLES 
 3 and 4, this observation was made, in addition to the observa- 
 tions mentioned in SECTION 4* Now the ratio of the temperature 
 rise, up to the end of the heating period, to the total rise is 
 a measure of the heat lost by radiation and conduction, the amount 
 by which the difference in the heat inputs for coke and coal must 
 be diminished to arrive at the quantity of heat actually required 
 to decompose the coal. TABLE Y gives all the necessary data and 
 calculations. In this case, too, the values are averages of sever- 
 al determinations. 
 
 TABLE V 
 
 SHOWING RATIO OF EXOTHERMIC HEAT TO THE HEAT 
 
 
 
 "ACTUALLY 
 
 REQUIRED TO 
 
 DECOMPOSE THE 
 
 COAL. " 
 
 
 
 SAM- 
 
 PLE 
 
 NO. 
 
 AYERAGEHKEAT 
 INPUT IN 
 CALORIES 
 
 DIFFERENCE 
 
 TEMPERATURE 
 RISE ° F 
 
 HEAT IN CALS. 
 PER GRAM 
 
 RATIO 
 
 
 COKE 
 
 COAL 
 
 TOTAL PER 
 GRAM 
 
 TOTAL AT END 
 OF HEAT 
 PERIOD 
 
 OF DE- 
 . COMPO- 
 SITION 
 
 EXO- 
 
 THER- 
 
 MIC 
 
 
 3 
 
 31180 
 
 36290 
 
 5110 265 
 
 14.37 6.50 
 
 113 
 
 39 
 
 35$ 
 
 5 
 
 34550 
 
 37580 
 
 3030 130 
 
 15.26 5.64 
 
 82 
 
 55 
 
 67/o 
 
 
 In the case of 
 
 #3, the heat 
 
 "actually re 
 
 qui red 
 
 to decompose 
 
 the 
 
 coal" 
 
 also includes a correction for the 
 
 heat of 
 
 vapori zation 
 
 of its moisture, i.e. the moisture lost at 105 , for that cannot 
 in any sense be considered a product of decomposition. 
 
 The samples listed above are, of course, relatively strongly 
 exothermic and less exothermic coals would no doubt show a smaller 
 

ratio. 
 
 35 . 
 
 Still, the figures do show that the exothermic heat is not 
 an inconsiderable factor in the energy required to decompose the 
 coal . 
 
 Perhaps the most significant point to he noted is that the 
 exothermic values here reported are very much lower than those that 
 are to he found in the literature. Taking the highest value ob- 
 tained, 39 calories (SAMPLE #3) and increasing it by 50$, to cor- 
 rect for a possible though highly improbable error, the exothermic 
 heat is still only 0.9$ of the calorific value, whereas the values 
 in the literature range from 2 to 7$, as stated in SECTION #1. The 
 most likely explanation for this difference is that the results 
 
 o 
 
 here given represent exothermic heat up to 500 C. only, while the 
 high results mentioned were obtained in high temperature carboniza- 
 tions. Evidence for the existence of marked exothermic reactions 
 
 above 500 is not lacking. The heating curves furnished by the 
 
 1 
 
 work of Vliet (unpublished report) and Hollings and Cobb give 
 unmistakable proof that considerable evolution of heat occurs be- 
 tween 600 and 800° C. 
 
39 
 
 SECTION 7 
 SUMMARY 
 
 1. A new method has been developed for the study of the 
 thermal behavior of coals during carbonization. 
 
 2. The method affords semi- quanti tat ive results with an 
 accuracy probably not better than +25 
 
 3. The exothermi city of high oxygen coals has been de- 
 monstrated by this new method. 
 
 4. The exothermic heat of carbonization of three Illi- 
 nois coals and two weathered Illinois coals has been 
 measured. 
 
 5. It has been confirmed that exothermi ci ty increases with 
 oxygen content. 
 
 6. It has been established, for the first time, that the 
 oxygen absorbed during weathering contributes from 2 to 4 
 times as much heat, per unit of oxygen, as the oxygen 
 originally in the coal. 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
SECTION 8 
 
 40. 
 
 -a- 
 
 I 1 
 
 2 . 
 
 3. 
 
 4. 
 
 5. 
 
 6 . 
 
 7 . 
 
 8 . 
 
 9 . 
 10 . 
 
 11 . 
 
 12 . 
 
 13. 
 
 14. 
 
 15. 
 16 . 
 17 . 
 
 BIBLIOGRAPHY 
 
 J. Gas lighting 126 917 (1914), 131 290 (1914) 
 J. Chem. Soc. 107 T 1106-15 (1915) 
 
 Gas World 60 872-8 (1914) 
 
 Weyman, J. Soc. Chem. Ind. 39 _ #12, 168 T (1920) 
 
 Heuser and Skiolde brand , Z. Angew, Chem. 32. I 41 (1919) 
 
 J. Soc. Chem. Ind. 38 2I5A (1919) 
 
 "Fuel Production and Utilization," (pp. 258-259), H. S. Taylor 
 
 Balliere Tyndall and Cox, London, 1920. 
 
 Chem. Ahst . 13 71 (1919), British Patent 119, 040 (1918) 
 
 Bulletin #24 (1908) Illinois Eng. Expt. Station 
 Files of Prof. S. W. Parr, University of Illinois 
 Euchene , Trans, Int. Gas Congress, Paris, 1900 
 
 Barnum , Amer. Gas Light J. (1906) 576 
 
 Mahler, Comptes Rendus (1891) 863 
 
 Constam and Schlapfer, Journ. fur Gasbel. 741-7, 774-9 (1906) 
 Constam and Kolhe , Journ. fur Gasbel. 770-780 (1909). 
 
 Poole, quoted in Journal fur Gasbel. (1906) 776 
 
 Amer. Gas Light. J. 68 125 (1898) 
 
 Constam and Kolbe , J. fur Gasbel. (1908) 669-73, 693-99. 
 
 Frankenfeld, Gas World (1914) 36 
 
 Klason, Zeit. fur Angew. Chem. (1910) 1256. 
 
 Klason, J. fur Prakt. Chem. 90 442 (1914) 
 
41 
 
 SECTION 9 
 APPENDIX 
 
 In compiling ths table which follows, an attempt was made to 
 recalculate the data more or less to the same basis throughout 
 but in many cases that was not possible owing to the failure of 
 the authors to state explicitly what the figures were or how they 
 were arrived at. So that, although the "heat losses," (SEE SEC- 
 TION 2, for a definition of this term), are not strictly compar- 
 able, they do serve at least to show the general order of magni- 
 tude . 
 
 It would, of course, be interesting to compare ultimate com- 
 positions, especially oxygen contents, with the respective ex- 
 othermic heats but the ultimate analyses are not available in a 
 sufficient number of cases. It is, however, true that exother- 
 micity increases with oxygen content. 
 
 A number of values for wood is included for purposes of com- 
 parison . 
 
 The first figure in the table was found by a calculation 
 based on statements made in Lewes', "Carbonization of Coal" 
 page 90, Benn Bros., Ltd., London 1918. 
 

 . 
 
 ■ 
 
 f 
 
 ■ 
 
 
 . 
 
 
 
COMPILATION OF EXOTHERMIC HBAT VALUES FROM THE LITERATURE ^2. 
 
 Name of Coal 
 
 Fixed Car- 
 bon % 
 
 Ash % 
 
 h 2 o % 
 
 Heat of Com- 
 bustion Cals, 
 per Gram 
 
 "Heat Loss" 
 % of Cal- 
 orific 
 Value 
 
 Index 
 No. to 
 Biblio- 
 graphy 
 
 
 
 • • • • 
 
 • • • • 
 
 • • • • 
 
 0.9 
 
 8 
 
 
 
 • • • • 
 
 • • • ♦ 
 
 • • • • 
 
 6.6 
 
 9 
 
 Commentry 
 
 
 • • • * 
 
 • • • # 
 
 # • • # 
 
 3.5 
 
 10 
 
 Belgian Anthracite 
 
 86.75 
 
 3.75 
 
 1,23 
 
 8217 
 
 4.8 
 
 11 
 
 Ruhrmage rkohle 
 
 81.92 
 
 4.18 
 
 1.19 
 
 8124 
 
 5.67 
 
 11 
 
 Huh re sskohle 
 
 81.26 
 
 3.96 
 
 0.97 
 
 8202 
 
 4.93 
 
 11 
 
 Ruhrfe t tkohle I 
 
 79.56 
 
 2.64 
 
 0.77 
 
 8427 
 
 5.10 
 
 11 
 
 Ruhrfe t tkohle II 
 
 74.63 
 
 3.17 
 
 1.02 
 
 8279 
 
 3.01 
 
 11 
 
 Ruhrgassflammkohle 
 
 62.72 
 
 6.42 
 
 1.57 
 
 7686 
 
 5.76 
 
 11 
 
 Bright Coal Not- 
 tinghamshire 
 
 49.23 
 
 3.06 
 
 9.60 
 
 7004 
 
 4.6 
 
 12 
 
 Trencherbone Coal 
 Lancashire 
 
 57.64 
 
 1.42 
 
 2.58 
 
 8066 
 
 3.8 
 
 12 
 
 Kinneil Coal 
 
 57.61 
 
 4.11 
 
 7.50 
 
 7811 
 
 4.1 
 
 12 
 
 Low Main Seam Durham 
 
 61.74 
 
 1.91 
 
 2.36 
 
 8250 
 
 4.5 
 
 12 
 
 Hutton Seam Durham 
 
 65.60 
 
 0.94 
 
 1.61 
 
 8594 
 
 3.3 
 
 12 
 
 Barnsely Coal York- 
 shire 
 
 59.09 
 
 8.65 
 
 0.94 
 
 7276 
 
 2.6 
 
 12 
 
 Durham Ballarat Seam 
 
 70.71 
 
 2.95 
 
 2.81 
 
 8321 
 
 3.9 
 
 12 
 
 Nixons Navigation 
 Coal, Wales 
 
 77.20 
 
 1.79 
 
 0.98 
 
 8486 
 
 3.0 
 
 12 
 
 Best Hard Coal Not- 
 tinghamshire 
 
 53.84 
 
 3.36 
 
 7.50 
 
 7208 
 
 4.2 
 
 12 
 
 
 
 • • • • 
 
 • • • • 
 
 • # • • 
 
 3.06 
 
 13 
 
 • •••••• ••• 
 
 • • • • • 
 
 • • • • 
 
 • • • • 
 
 • • • • 
 
 3.1 
 
 13 
 
 Gardanne 
 
 39.8 
 
 10.79 
 
 10.55 
 
 5673 
 
 7.20 
 
 14 
 
 Chapelle sous Dun 
 
 42.2 
 
 14.39 
 
 10.43 
 
 5876 
 
 4.4 
 
 14 
 
 Pyranaenkohle 
 
 52.4 
 
 4.75 
 
 3.88 
 
 7156 
 
 6.68 
 
 14 
 
43 
 
 Name of Coal Fixed Car- Ash % H£0% Heat of Com- "Heat Loss" Index 
 
 
 bon % 
 
 
 
 bustion Gals. 
 
 % of Cal- 
 
 No. to 
 
 
 
 
 
 per Gram 
 
 orific 
 
 Biblio 
 
 
 
 
 
 
 Value 
 
 Graphy 
 
 Bruay 
 
 58.1 
 
 3.96 
 
 2.36 
 
 7663 
 
 6.0 
 
 14 
 
 Blanze 
 
 51.7 
 
 13.70 
 
 2.93 
 
 6793 
 
 4.8 
 
 14 
 
 St* Etienne 
 
 64.9 
 
 5.50 
 
 0.94 
 
 8085 
 
 3.3 
 
 14 
 
 Lens 
 
 67.3 
 
 4.66 
 
 0.98 
 
 8138 
 
 3.7 
 
 14 
 
 Ronchamp 
 
 63.8 
 
 14.05 
 
 0.87 
 
 7369 
 
 4.5 
 
 14 
 
 Grand' Combe 
 
 69.3 
 
 10.53 
 
 0.74 
 
 7531 
 
 2.6 
 
 14 
 
 Meurchin 
 
 84.3 
 
 2.93 
 
 1.08 
 
 8326 
 
 2.2 
 
 14 
 
 Epinac 
 
 77.8 
 
 11.05 
 
 0.97 
 
 7505 
 
 4.4 
 
 14 
 
 Ostricourt 
 
 85.0 
 
 4.71 
 
 1.10 
 
 8115 
 
 3.2 
 
 14 
 
 La Mure 
 
 89.2 
 
 3.41 
 
 3.48 
 
 7627 
 
 3.6 
 
 14 
 
 
 
 • • • • 
 
 • • • • 
 
 • • • • 
 
 3.4 
 
 15 
 
 Wood 
 
 
 • • • • 
 
 0 0 0 0 
 
 • • • • 
 
 6.3 
 
 16 
 
 tf 
 
 
 • • • • 
 
 • • • • 
 
 • • • • 
 
 4.6 
 
 16 
 
 ft 
 
 
 • • • • 
 
 • • • • 
 
 • • • • 
 
 5.9 
 
 16 
 
 ff 
 
 
 • • • • 
 
 • • • • 
 
 • • • • 
 
 6.6 
 
 16 
 
 Purified Cellulose 
 
 
 • • • • 
 
 • • • • 
 
 # • • • 
 
 3.8 
 
 16 
 
 ft ff 
 
 
 • • • • 
 
 • • « • 
 
 • • • • 
 
 5.4 
 
 16 
 
 tl M 
 
 
 • • • • 
 
 • • • • 
 
 • • • • 
 
 5.3 
 
 16 
 
 tf tf 
 
 
 • • • • 
 
 ♦ • • • 
 
 • • • • 
 
 6.4 
 
 16 
 
 Wood 
 
 
 0 0 0 0 
 
 • • • • 
 
 0 0 0 0 
 
 3.5 
 
 17 
 
 ff 
 
 
 # * ♦ • 
 
 • « • • 
 
 0 0 0 0 
 
 2.0 
 
 17 
 
 ft 
 
 
 • • # • 
 
 • • • • 
 
 0 0 0 0 
 
 -1.0 
 
 17 
 
 ft 
 
 
 0000 
 
 • • • • 
 
 0 0 0 0 
 
 10-15 
 
 4