THE INVESTIGATION OF CERTAIN TYPES OF SOUTH AFRICAN COALS, WITH SPECIAL REF- ERENCE TO THEIR HIGH NITROGEN CONTENT By VERNON BOSMAN A. B. University of the Cape of Good Hope, South Africa, 1917 M. A. University of Cape Town, South Africa, 1918 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN CHEMISTRY IN THE GRADUATE SCHOOL OF THE UNIVERSITY OF ILLINOIS, 1922 URBANA, ILLINOIS . THE GRADUATE SCHOOL Jtoy 10 - 192 - 2 . I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY- VERNON B OSMAN ENTITLED OP SOUTH AFRIC AN COALS, WITH SPECIAL REFERENCE TO THEIR HIGH NITROGEN CONTENT* BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE DEGREE OF by IaC AX ^ P In Charge of Thesis Head of Department *Required for doctor’s degree but not for master’s Digitized by the Internet Archive in 2015 https://archive.org/details/investigationofcOObosm ACKIIOY/LEDGgMSIIT The author wishes to thank Professor S. W. Parr tor his kind assistance and advice in compiling this thesis. TABLE OF CONTENTS I. Introduction 1 II. Historical The distribution of nitrogen in the distill of coal at higher temperatures ation 4. The constitution of coal 8. III. The coals studied Geography 17. Analytical data 18. | IV. The nitrogen in coals Theoretical considerations 25. The distribution of nitrogen at lower temperatures 27, Conclusions 28. The action of chemical reagents 53. Hydrolysing agents 53. The determination of NH 3 nitrogen 55. The action of selenium oxychloride 60 . The action of strong reducing agents 62. 66 V. General Summary < and Conclusions regarding the nitrogen in coal VI. Bibliography 70 VII. Vita. 75 ! ; ' . > INTRODUCTION The basis of this work is an investigation oi certain types of South African coals, planned with a view to obtaining a more fundamental knowledge of the form in which nitrogen exists in coals. Almost all the coals used may be looked upon as high nitrogen coals, five of them containing an average amount of over 2 per cent, as compared with an average of 1 to 1.5 per cent in American coals. In modern industry the greatest efforts for improve- ment have been in the direction of a thorough system of uti- lising by-products. The scientific status of an industry may very often be determined by the way in which it prevents wast9 in all directions. As an example of this kind the by-product coke industry is no exception. It fails, however, in one respect, namely to recover the maximum amount of nitrogen from the coal. In the destructive distillation of coal on a large scale, not more than 15 per cent of the nitrogen is obtained as ammonia, while 50 to 60 per cent remains behind in the coke. Attempts to recover this nitrogen from the coke have only succeeded in cases of special treatment, such as the action of steam in the Mond process, or the addition of lime to the coal. In the first of these processes 60 to 70 per cent of . ■ -J - . . . ■ * ■ . . - 3 - the nitrogen in the coal is obtained as ammonia, and the rest as free nitrogen gas arising from the dissociation of ammonia. But special treatment of this nature destroys the coke and acts detrimentally on other by-products. For this reason it is not always economical. We must, therefore, resort to other means' — to a process which would increase the yield of ammonia without affecting the remaining by-products, and, at the same time keep the total cost more or less constant. A fundamental study of the nitrogen in coal would add materially to any solution of our problem, and would bring us to a point where further investigation should not be difficult. At the same time a study of the nitrogen would probably throw some light on the constitution of coal-- a subject of great scientific interest and importance. Importance of the Problem. Fixed nitrogen in the form of ammonia or ammonium salts or combined with carbon has two main sources through on which it finds its way /to the markets of the world. One of these is from the atmosphere and the other from coal. Today it is uncertain which of these two sources is the more im- portant. Both have their advantages, but the disadvantages of the former — the heavy initial cost, and the uncertainty -3- and complexity of the process as a whole— encourage us to believe, that, for a long time to come, our chief source for ammonia will be from coal. In the United States alone the total reserves of nitrogen in coal are calculated at 31,000,000, 000 tons (13), Explosives and fertilisers— these two uses alone — the one largely for the destruction and the other for the maintai nance of life — have made this problem an important one. For the last twenty years or more, we have listened to warn- ings, that the world is in danger of starvation. However true these warnings may have been, and however well they may have been met by scientific progress, the danger still remains acute. Although there has been a sudden drop in the output of explo- sives during the last few years, the demand for fixed nitrogen is greater than it has ever been in normal times. This is due to the fact that during the wax, much of the nitrogen which should have been used for fertiliser was used for explosives, with the result that the soild suffered and the human fo^d supply decreased by as much as 40 per cent in some European countries. This deficiency has now to be met or the world will soon be underfed. The challenge forms one of the most important and far reaching problems with which the chemist has to deal. ■ . . . ■ ■ . . . . -4- Latest statistics (49) show that the total annual output of ammonium sulphate in the United States from by- product coke industries amounted to 935,000,000 pounds, due to the carbonization of 44,324,000 tons of coal. This quantity of ammonium sulphate is equivalent to 21.2 pounds per ton. Assuming an average of 1.5 per cent of the nitrogen in the coal used, approximately 15 per cent of the nitrogen in the coal was, therefore, recovered. Now if the efficiency of the process could be raised to 100 per cent, an amount of approxi- mately 6,230,000,000 pounds of ammonium sulphate would be obtained from the same amount of coal. Such a condition would lower the price appreciably and nitrogen fertilisers will be brought within general reach. One of the most pressing problems of modern agriculture would thus be solved. HISTORICAL A. On the Nitrogen Content of Coal and The Distribution on Distillation. As early as 1844 Swindells (l) believing that coke acted as a sort of catalyst in distillation processes, suggested a method of preparing ammonia on a large scale by passing nitrogen, nitric oxides and steam over red hot coke. -5- Some twenty years later Berthelot (2) tried, the preparation of hydrocyanic acid from acetylene and nitrogen in the same way. Seeing that in the distillation of coal quite 50 per cent of the nitrogen remains behind in a form which is decomposed by steam, it is likely that the respective yields obtained in these experiments resulted from this origin. Henin in 1382 (3) was further attracted by the subject and arrived at conclusions which led Mayer and Altmayer (4) to make a thorough investigation of the effect of steam on the yield of ammonia at different intervals of time between the temperatures of 600° and 900°. tBeyfound that a maximum yield of 62.7 per cent of ammonia was obtained at 800°. A little later James McLeod (5) made an extensive study of the behaviour and distribution of nitrogen in the distillation of coals. He concluded that at the temperature at which the gases are evolved, a part of the nitrogen, which is liberated entirely as free nitrogen, combines with hydrogen to form ammonia, a part with carbon and hydrogen to form cyanogen and pyridine and a part remains as free nitrogen. In distil- lation experiments in which 227 to 413 tons of coal were utilised, McLeod was able to recover as much as 17 per cent of the nitrogen as ammonia and states that this figure depends on certain factors, such as moisture in, and the physical condition of, the coal. Andrew Short (o) in a later article ‘ . . ~ i - 6 - summarises a whole series of results obtained in this way, as follows: — Distribution of nitrogen Foster Knoblauch McLeod Short as ammonia 14.5 12.14 17.1 15 . 16 as cyanogen 1,56 2 1.2 .43 in the coke 48.68 50 58 43.31 in the tar 5.9 2*98 in the gas 35.26 3C 19.5 37.12 In 1911 W older eck (7) applied these methods using peat instead of coal with a view to obtaining the nitrogen in the form of ammonia. In 1914 0, Simmersback (8) carried out a series of investigations regarding the relative amounts of ammonia and hydrocyanic acid at temperatures between 800° and 900° and showed the effect of steam on the respective yields. J. H?. Cobb in similar experiments found that by removing the ammonia immediately in a distillation process carried out in the lab- oratory, a yield of 22.5 per cent of the nitrogen could be obtained. He further treated the coke with steam and obtained 60 to 70 per cent of the nitrogen as ammonia— the conditions found in the Mond process. In these experiments the ammonia yield rapidly decreases as the temperature rises above 800° C, due to -7- dissociation of ammonia at these temperatures. Both Simmer s- back and Cobb conclude that the nitrogen comes off in the form of ammonia, rather than as nitrogen as McLeod suggests. In 1915 Terres (13) in attempting to determine directly the form in which the nitrogen exists in coal, sub- jected a number of organic compounds with nitrogen linkages to dry distillation. Such compounds as gly cocoll, asp ar agin, aloumin, animal glue, pyridin, azo benzol, hy dr azobenzol, phenyl isocyanate, nitrobenzol were taken. It was found that only those substances with amino and substituted amino groupings gave ammonia. He, therefore, concludes that the nitrogen has an "albuminous mother substance" for its origin. In 1320 Glund and Breuer (15) worked on a gas coal of nitrogen content equal to 1.86 per cent and subjected it to low temperature distillation in a revolving cylindrical retort. He found that 66 per cent of the nitrogen remained in the semi-coke and only 1.8 per cent of the nitrogen was liberated as ammonia, When the semi-coke was heated a further 16 per cent was liberated, thus bringing the total yield of ammonia to 17.8 per cent. In an unpublished thesis (l6a) by Chiles (1920) the author attempted to determine actually what changes occur in the nitrogen molecule during the coking of coal. Attempts were m^de to prepare synthetically the same compound as it - 8 - exists in coal by fusing chemically pure carbon obtained from sugar and protein substances. No definite conclusions were drawn, but results seemed to indicate the presence of nitrides, or direct combination of carbon and nitrogen. « In 1S21 Monkhouse and Cobb (16) determined the effect of hydrogen and nitrogen alone, and in the presence of steam, on coke residues prepared at 500°, 800°, and 1100° C, 34.24 per cent of the nitrogen of the coke was obtained with 500° C coke using hydrogen gas. This was the highest yield obtained. B. On the Constitution of Coal. Running almost parallel with these results is to be found a series of researches carried on with a view to obtaining a more fundamental idea of the constitution of coal. A determi- nation of the form of nitrogen in coal would be a step in this direction. For this reason these researches have been described, In this work three general methods have been uni- versally employed/ l) that of selective solvents (2) microscopic investigations and fe) distillation processes. The first work on the use of solvents seems to have been done by Dr Smythe (21) at Gottingen, and was published in a report to the commissioners of the 1851 exhibition. He used a Cologne coal and classified the following solvents in order of their extractive ability ' — benzene 3 per cent . . . ■ , . -9- chloroforin 1.8 per cent, ethyl alcohol 2.4 per cent, ether, petroleum ether and acetone — the latter three dissolving out a very small portion of the coal. In all these extracts , except that from alcohol, nitrogen is reported absent. In 1879 Guignet ( 1 7) tried the action of phenol on coal, and extracted 4 per cent of the coal. He then tried the action of nitric acid on the residue and extract. This work was followed up by Friswell (18), who compared the action of nitric acid on graphite and coal respectively. With coal he obtained a crystalline product which he believed to be a nitro compound, similar to nitrocellulose. From 90 grams of a bituminous coal, finely ground, a residue of 12.5 grams was obtained. About the same time Smith (19) in trying the action of benzene on coal, found that with a certain Japanese coal an extract as high as 10 per cent was obtained. Smith points out the unique nature of this coal, and goes a step further in a theory, in which he compares the origin of coal with that of petroleum. He suggests that the anthracite coal of Penn- sylvania was probably at one time of its formation a bituminous coal similar to the one of Japan on which he worked, but owing to pressure, temperature and other physical phenomena, had been deprived of its titumin which nowadays is being brought to the surface in the form of petroleum. Regarding the nitrogen - 10 - in the coal. Smith seems to think that a high nitrogen content should predict animal origin and substantiates his theory by comparing the nitrogen content of tar from wood and from bone respectively. Following these experiments work of Dr. Smythe (21) is again referred to, in which he attacked coal with dilute hydrochloric acid and potassium chlorate. Thirty to 35 per cent of the coal was then found to be soluole in alcohol and acetone. By extraction with benzene he was able to separate a number of substances containing chlorine and a high percentage of oxygen. Thus C^oH-gClsOlC and c 35 H jg 01 4. 0 20 were recognised. Anderson and Roberts (22) took up the work from here and extracted an El coal, which had been previously oxidised by atmosphere oxygen and by dilute HN0 3 , with potassium hydroxide. They were able by this means to extract a number of acid substances believed to be derivatives of humic acid. Finally they conclude that the nitrogen, or nitrogen containing substances, in the coal have absolutely nothing to do with the coking properties of the coal. They believe that the coking property of coal is due to the ease with which certain substances in the coal will volatilize or decompose, and that a considerable portion of the organic matter in coal consists of a complex compound comparatively rich in nitrogen, and containing sulphur as well. In addition to these, resinous material is always present to a small, but fairly constant - 11 - extent. They state that "the nitrogenous bodies obviously owe their origin to the proteid substances of the vegetable matter from which the coal was formed". Donat h and Margosches (24) added powdered permanganate to their alkali solvent and arrived at results which strengthened the view that when oxidised, coal yields acid substances, resembling humic acids. In 1901 Baker (23) tried the action of pyridin on coal. Donath (25) a little later carried the study further on a German coal. These results were followed by a series of experiments by Bedson (26) on gas coals. Twenty-four to 65 per cent of pyridin soluble material was obtained. He suggested that the pyridin extract might te related to the volatile matter in the coal, but this theory was afterwards disproved. He also applied different organic solvents to extracts and residue and obtained interesting results, though no definite conclusions. In 1911 Lewes (28) continued the work and arrived at interesting conclusions regarding the coking properties of the residue and extract. He explains the retention of the coking properties of some coals by assuming the presence of a resinic body not soluble in pyridin. He further noticed that the percentage of volatile matter of certain coals had increased after extraction and concluded that some of the pyridin was held back ty the coal to form a compound with the insoluble - 12 - part. Lewes points out that pyridin is, therefore, an unsuitable solvent, especially for nitrogen investigations. He further states that the resinous bodies in coal are of two kinds "the one easily oxidisable, soluble in pyridin and saponifiable by alkalies, and which on weathering is oxidised into humus bodies with the evolution of water and carbon dioxide, and the other non oxidisable, not saponified by alkalies, and forming with pyridin a compound insoluble in excess of the reagent; and this Class may be the hydro-carbons from decomposed resins, as the residue in which they are present yields rich liquid hydro- carbons, as tar and pitch, but not rich in gas." In 1911 Pictet (29) working at different times in conjunctions with Ramseyer, B©uvier, Labonchere, Combes and Kaiser, started on a series of researches which gradually de- veloped into a very interesting, as well as a useful piece of work. These investigators attached the problem from two angles. In their first series of experiments they extracted a "fat" coal of Montrambert (Loire) with benzene. The distillate was obtained in the form of a tar, which was fractionally distilled into a number of products, which the authors were able to identify. Amongst these compounds were hexahydrof luerene , a number of hydro-carbons, phenols and bases. The same coal was then subjected to vacuum distillation up to a temperature of 450° and at a pressure of 13 to 15 mm. The tar obtained was * • - -13- found to be similar to the tar obtained in the first case, and these in turn showed marked resemblances to the fractions obtained from petroleum. The authors finally conclude that the coal and petroleum have similar origins (41). These investigations were started with the object of finding out the form in which nitrogen exists in coal, but no conclusions on this point were drawn. About the same time that these researches were begun, Wheeler together with Burgess (30), Jones (35) and Clarke (33) set out on a thorough investigation on the solvent action of pyridin on coal and were able to arrive at conclusions on 'which a theory was based. They found that when coal was extracted with pyridin, the extract contained resinous decomposition products together with a substance of cellulosic origin. These two substances they were able to separate by means of chloroform. On destructive distillation of these three substances they found that the residue yielded chiefly hydrogen, while the pyridin soluble, chloroform insoluble part, yielded chiefly hydro-carbons. The chloroform soluble part of the extract resembled the residue in yielding chiefly hydrogen. They therefore concluded that coal was composed of two parts (l) the hydrogen yielding and (3) the paraffin yielding constituents. They further illustrated their results by analysing the gases given off at different temperatures in the destructive distillation of coal. . . '- A " * . ■ -14- In 1912 Hoffman and Fraser (31) published a paper on "The Constituents of Coal Soluble in Phenol". Illinois coal from Franklin County was taken and it was found that 10.8? per cent calculated on the moisture and ash free basis was soluble. By the use of sodium hydroxide and other organic solvents they were able to extract certain substances out of the coal which they believed to be pure. Complete analysis of these substances were given, but no constitutional formula established. This work was followed up by Parr and Hadley (34) in a very thorough investigation on the properties of the residue and extract, using phenol as solvent. These investigators believe that phenol is the best solvent to use as it undergoes decomposition only to a very slight degree and does not effect the coal as does pyridin. Both the residue and extract were analysed and examined in regard to their resistance to air and their coking properties. They find that the nitrogen divides itself more or less evenly between the residue and extract. Porter and Taylor (36), working on the volatile products of coal, carried out investigations on four coals, which unfor- tunately had been weathered. They arrived at results on the constitution of coal which were opposed to those of Wheeler and state that the "cellulosic" material in the coal is first d e- composed on exposure to heat. Microscopically some very brilliant researches have -15- been carried out on ccal, which throw a good deal of light on its origin and constitution. Amongst the foremost workers in this field may be mentioned Thiessen (59), White (58), Stopes and Wheeler (5?). In the last four or five years a great deal of inves- tigation has been done on the action of chemical reagents on coal which has produced interesting though not conclusive results. This work has mostly origined from German chemists, and unfortu- nately could not be obtained in the original articles by the writer. Thus Keller, Hilpert and Lepsius (40) treated a bituminous coal with acetic anlydride and zinc chloride in a sealed tube and then subjected the residue with nitric acid. The content of nitrogen increased from 1.8 per cent to 8 per cent. This substance they called "nitrc-coal". They found it insoluble in acetone, acetic acid and benzene. The extract resembled the residue when these were subjected to destructive distillation, and was soluble in alkalies. They failed to prepare sulphuric acid derivatives and suggested the absence of aromatic hydro- carbons as a result. Fisher and Groppel (41) about the same time tried the effect of pre-heating the coal and thereby increasing the extracts obtained by solvents. With Niggerman (42), Fisher tried the effect of ozone . , . . . , . . . -16- on coal, and found that the humic acid substances were rendered soluble- — the amount being inversely proportional to the coking properties of the coal. The ozonised product was dried at 110° and contained 3 per cent H, 50 per cent C, .8 per cent S and only traces of nitrogen--the original coal containing 5 per cent H, 84 per cent C, .9 per cent S. A little later Fisher and Tropsch (47) tried the same reaction in a non aqueous media, but could observe no changes in the reaction. The same authors also tried the effect of hydrogen iodide (47) on coal and found that coals of early origin were attacked sooner than those of later origin. The amount of extract obtained with chloroform after such treatment was increased from 3.7 per cent to 73 per cent in the case of a cannel coal. In 1921 F. Fischer and Schrader (48) published a paper on the origin and chemical structure of coal, which con- tained a number of new ideas and new theories, not the least of which was the fact that they question all previous theories on the subject. According to them coal originates from the lignin in the plant — the cellulose being decomposed by bacterial action in the early stages of peat formation, with the formation of C0 3 and H a 0. Lignin, they state, has an aromatic structure, with acetyl and methoxyl groups. Thus in the formation of peat the methoxyl groups increase, while the portion soluble in concentrated hydrochloric acid decreases. Finally the methoxyl . E ..-r :c ' v . o. A *' i--- •*>>« ft . . . . . -17- groups axe replaced by hydroxyl groups to form a compound identical with humic acid. Further splitting off of H S Q, C0 2 and CH 4 give rise to lignite. The authors quote experi- mental evidence in proof of their theory. Five months later a further article appeared in which the above theory received severe criticism at the hands of Klever, Forschner, Jonas and others. The question still remains unsettled. The Coals Studied. The coals which formed the basis of the study described in this paper came from South Africa. Fourteen samples from different mines were sent to this University by Mr. P. Wagener, Inspector of Mines for the Union of South Africa. These samples were taken according to standard methods and were received in good condition in well sealed tins. Unfortunately all but a few were packed in the finely ground condi tion — a condition in which weathering is most effective. No attempts were made to investi- gate the extent of weathering. The coal resources of the Union of South Africa were estimated in 1913 at 57,839 million tons. This is probably a considerable underestimate. Between 1913 and 1S20 the total output of the Union was increased by 45 per cent, it being nearly 12 million tons for 1920. . . . . . . . . • . . - 18 - Alraost all the coal in the Union comes from two states, the Transvaal and Natal — the latter coals being of a slightly higher quality. The position of these fields i3 shown in the map in(Figure I), according to the following list. 1. Utrecht Collieries — Natal 2. Cambria Collierie s— Dannhouse Natal 3. South African Northfield Collieries—Glencoe Natal 4. Navigation Collieries--Natal 5. Transvaal and Delegoa Bay--Witbank 6. Clydesdale Collieries 7. Cassel Coal Coy — Blackhill 8. Middelburg Steam Coal & Coke Coy 9. Coronation Collieries 10. Emyati Collieries — Langkjans — Natal 11. Uitspan Collieries — W itbank 12. Clydesdale Collieries 13. Tavistock Coal and Coke Co-r-W itbank 14. African Freehold Coal Lands Ltd. Vaalbank, Middelburg Experime ntal . In order to get a thorough idea of the nature of these coals complete analyses were made. The methods used were those described in "Gas and Fuel Analysis" by White, "Water and Fuel Analysis" by Parr and "Methods of Analysing Coal and Coke" by Stanton and Fildner. Total carbon was determined by the Parr"total carbon" apparatus. Oxygen and hydrogen were calculated from the Dulong formula. Heat values were made by the Parr Adiabatic Calorimeter. * ■ ' * . I , Coalfields. - 20 - The Kjeldahl Gunning method was employed in all nitrogen determinations. One gramme of coal is placed in a 500 cc Kjeldahl flask, and .5 grams of KHS0 4 , .5 grams of Hg. and 30 oc of concentrated sulphuric acid added. The mixture is digested for three to four hours and after removing the Hg with K 2 S is finally distilled in the presence of NaOH into standard H 2 S0 4 . In the case of coke it is advisable to take .5 grams of the material, digest for three hours then add a few crystals of KMn0 4 and finally digest for three hours longer. In each set of determinations blanks were made and subtracted from the total amount. Table 1 shows the proximate analyses, calculated to the moisture and ash free basis; Table 2 shows the ultimate analyses and Table 3 their classification. In order to get a better understanding of their classification, each coal was plotted graphically (Figure II) according to the classification of Parr and Vliet. In this classification the unit volatile matter is plotted against the unit B$.U value of the coal. Unit volatile matter is defined as the volatile matter on the pure coal substance and is calculated from the following formulas. unit volatile matter (pure coal) = 100- fixed carbon on pure coal basis fixed carbon (pure coal) = fixed carbon as determined 1-(1.0S A - 1/ 2S - 1/20S - M) where A=ash content. S=sulphur content M-moisture content . . . * . . . . TABiiS I SHOTOIG PROXIMATE ANALYSIS OF GOALS ~0T ra o H •H 02 tO £> r dd02tDCD'P>Hd04C'- 02 >s 02 CO rd 02 CD rdrd02rdCDrdrd02 CO W • • d P •H a} 03 <1 d lO ID CD o* 02 rH CO O rd 04 02 c- CO 02 d O o 04 tO o 00 03 C- 04 CO CO ID 00 02 d 08 © 3 • • • • • • • • • • • • • • M P o 00 d CD o C» CD CD D- co 04 to C*- to © © vH G? P © O CO 04 rd o 02 id d rH d I — 1 C- CD w • • • • • • • • • • • • • • •H i — i -P o. i — 1 lO d 04 02 d d 02 tO o CO 02 CO O O Cd ^4 > B CO CO 02 02 CO CD co CO CO 02 CO CO CO CO 1 — 1 00 tO d rd CO 04 04 CO d 02 02 rH to 04 co d 02 rd d tO 02 ID O 04 04 CO CO 3 • • • • • • • • • • 9 « • • m o CO O co rd o tO CO CO rH rd o <4 1 — I H rd rH rd id rH i — 1 i — 1 rH d o 00 CO tO CD o CO £- d CD CD O IP rd d O d rd o rd CO CD ip id CO 04 id 04 tO 02 © rQ • • • • • • • • • • • • • • M P 02 CO C4 CO 02 04 CO CD CO t- 04 CO 04 04 © © -h cd p © «H O tO to to to CD CD CD CD CD tO ID CD CD CD d p to o d 02 00 d CO rd to ao 02 CD •P 3 04 d d to O O CD ID o O 04 id O 04 CO to d • • • • • • • • • • • • •d rd -P • • CO 02 04 04 o CO CO d CD 04 04 04 O O Cd to 00 02 02 CO 02 co 02 02 02 02 02 02 02 l M > s 02 02 o to CD O to 04 00 CO CO tQ CD 02 CO 04 to 02 co i — 1 00 rH d 04 CO 04 tO tO O CO 3 • • • • • • • • • • • • • • 03 o D 04 02 rd o CO CO d rd rd o id rd i — 1 rd i — 1 rH i — 1 rH I — 1 rd d d 00 02 d CD CO rd 04 d to IP d to CO d o to 02 O rd CO 04 tO o 02 o o d d to © CO © rQ • • • • • • • • • • • • • • -P m M P o 02 CO C» id IP d o- £P 00 c- CO o cd co *h a 3 g >h«h o •H |d tO tO to tO CD ID CD CD iO to CD iD CD ID CO d o 04 rd CO CO 04 t<0 o ip tO CD tO O 13 rd -P 04 o rd 02 d 02 04 04 ID IP d d d rH P 4 O d • • • • • • • • • • • • • • PH t=> g CD ao co 02 CO CO 04 CP ip CO CD co ao 04 02 02 02 02 co 02 02 02 02 02 02 02 02 02 o CO 02 CO JP CO O CO d O OO rH CO CD 5> 0? 00 d CD d CO t o 04 04 co 02 00 d O lO W • • • • • • • • • • • • • • 02 rd i — 1 rH rd 02 rd H 1 — 1 rd rH 02 Cvj 02 i©q.nxnil H itaoq.'eaoq'B'l r*» -P rd rd O O -P 0} ■H rCj 1 ! *H rd © a$ ^1 s} © H-4 3 © P aj g o 3 c3 p? £> o CD d CD d cd id to O © CO 04 o rH 02 rd rH a M rd rH rd i )_ rd bQ'— rd 02 r^ rd O i — 1 i — i tO d i — 1 © CO rd O • o i — 1 O rd rd • H O rd — - O • O O O rH O cd o 1 • — * © o © o o 3 oS o r — 1 o o © rH •H rd PA! 3 rd bQ rd rH • 3 o © rH 'H 0} s O © S3 rd rd P d CD rd O O © N rC! -P o rd Gj rd d OH d 3 O O O £3 H • M o 3 rd ■P (J$ •H cd 3 rd i — I o -H O Q •H • O cd O © o od P 3 -p id cd -P d -h rd O © -P o d d 3 o d o o o o5 a3 t> *H CO 3 rd , K 4 rd «3 3 *H rd cd d 03 -P © cd to o SS5 • fciO-P 03 ^ © 3 © © S d d d> od 3 cd © 03 M © • i — 1 •h a} dO 03 cd c5 d o d -p 03 3 d •H 6 •H d CD : 1 rd > 3 rd >9 >9 35 03 rd © -H p cd -p p> >s P rd 44 • O rd P cd rd H ,d 3 -p 3 o S3 •H *H rd cd O o co o 3 ^3030 co O 3 O Eh og <{ o - 22 - Table 2 SHOWING ULTIMATE ANALYSIS OF COALS ‘ & O H Location «$ ,o Ultimate Analysis Calorific Value O P rQ S c3 Hi S. Ha C arb on n 2 0 3 Calories B.t, Utrecht Coll. Natal 1 1.58 4.45 70.70 2.12 7.72 6952 12334 Cambria Coll. N at al 2 1.67 4.90 75.93 2.17 5.66 7616 13707 S. A , Northf ield Coll. N at al 3 1.71 4.79 77.33 2.32 4.97 7726 13917 Navigation Coll. N at al 4 1.72 4.64 76.30 2.17 4.60 7606 13695 Transvaal & Dele go a Bay. Wit bank No. 8 5 1.24 4.81 69.33 1.68 8.20 6900 12421 Clydesdale Coll. Blackhill No. 2 6 1.26 4.22 70.50 1.62 8.61 6808 12255 Cassel Coal Coy Blackhill No. 1 7 1.29 4.46 71.98 1.68 8.18 7057 12702 Steam Coal & Coke Midaleburg 8 .87 4.16 67.45 1.53 8.07 6558 11805 Coronation Coll. No. 3 9 .88 4.80 68.18 1.67 9.27 6787 12119 Emyati Coll. N at al 10 1.24 4.54 77.47 2.11 5.40 7622 13720 Uitspan Coll. Wit bank No. 10 11 .84 4.23 68.43 1.61 8.43 6645 11960 Clyde sdale No, 2 12 1.40 4.33 69.86 1.60 8.61 6797 12232 Tavistock Coal Coy Witbank No. 6 13 1.39 4.30 71. 38 1.75 8.07 6925 12464 African Freehold Middle burg 14 .42 4.54 70 . 38 1.78 9.72 6885 12392 No. 11 * -Si- Table 3 SHOWING THE CLASSIFICATION Vol.Mat. Locality L ab ' y No. Vol.Mat. (air dried) Unit B.t .u. on Unit coal Cambria Coll. N at al 2 28.04 15351 29.80 Emyati Coll. N at al 10 23.70 15271 25.01 S, A.Northf ield Coll, Natal 5 23.10 15421 24.07 Navigation Coll. Natal 4 22.29 15508 2o . 50 Transvaal & Delegoa Bay. Witbank No. 8 5 33.41 14805 37,96 Cassel Coal Coy Blackball No. 1 7 29.98 14803 33.05 Steam Coal & Coke Middle burg 8 27.99 14767 32.65 Clydesdale Coll. Blackhill No. 2 6 28.28 14409 31.55 Coronation Coll. No. 3 9 27.56 14426 31.24 Clydesdale No. 2 12 28.46 14444 31.88 African Freehold Coal Coy. Middle- burg No. 11 14 29.16 14425 32.45 Utrecht Coll. Natal 1 25.93 14442 28.53 Uitspan Coll. Witbank No. 10 11 25.47 14550 29.13 Tavistock Coal Coy Witbank No. 6 13 28.43 14541 31.47 -25- The Nitrogen in the Coal. From this point attempts were made to determine in what form the nitrogen exists in coal. For this purpose two methods of attack were adopted (l) the distillation of coal in the presence of carbon dioxide (2) the use of chemical re- agents at lower temperatures. According to a recent thesis by Ho^bart (5?) carbon dioxide does not affect carbonization processes in any way. Theoretical In order to get some basis upon which to build, it is necessary to make use of certain theoretical considerations regarding the origin of coal — theories which, hov/ever, are now widely accepted by geologists and chemists. Thus David White (38) in a very extensive and thorough bulletin on "The Origin of Coal", and later Thiessen and White (3S) on the ’’Structure in Paleozic Bituminous Coals” claim that the decomposition towards the formation of coal, is exposed to two fundamental agencies Q) Biochemical or Bacterial and (2 )Dynamo- chemical . In the first of these stages the formation of peat, or its equivalent in rank, is brought about; in the latter the peat is exposed to the action of pressure and temperature, during which process a low grade lignite is formed. Water is expressed, thereby reducing the mass by as much as three-fourths, gases axe expelled, organic compounds are further broken dov/n and finally oxygen, nitrogen and hydrogen, together with varying amounts of carbon are liberated. . . - . . -26- Furthermore at this stage we assume that the nitrogen in coal originates from the protein at one time present in the plant from which the coal was formed. Coming from the plant it is certain that the nitrogen is of organic and not of inorganic origin. It is true that certain plants contain alkaloids, in which the nitrogen is bound differently in the molecule from protein, but it is not likely that these compounds play any part in coal formation, for such plants are not common and are small in bulk. All writers who have had anything to do with the nitrogen in coal have believed that protein is its true origin. dilute hydrochloric acid, readily decompose. This has been shown by Hlasiwetz and Haoerman (50), Kossel (51), Kutscher (52), Emil Fischer (5b) and others. Thus when proteins are hydrolysed in the laboratory we obtain chiefly as decomposition products the following groups of compounds: (1) Amino acids, which include monamino and di amino carboxylic acids, eg: Amino acetic, amino valeric, diamino propionic, diamino glutcefiiq. (2) Acid amides eg: acetamide. (b) Heterocyclic compounds eg: Histidine, having the for- and oxy prolin®. Now v/hen proteins are exposed to the natural agencies as are supposed to exist in the formation of coal, we would expect hydrolysis brought about by bacterial action, and as a result, the products of decomposition would be brought down in Proteins in the presence of hydrolysing agents such as mula . -27- the coal. There is, however, a possibility for chemical combination to take place with other organic compounds in the plant before complete hydrolysis has taken place, especially if the protein is protected in some physical way by the plant tissue. This possibility is dealt with in later experiments sing which determine the effect of hy dr oly /agents on the nitrogen molecule in coal. These theoretical considerations form the basis of the work which is to follow. Experimental. A. Distillation. In the first series of experiments a thorough investi- gation was made of the behaviour of the nitrogen when coal is distilled in an atmosphere of carbon dioxide, between the temperatures of 450° and 600° C. These temperatures were chosen because it was found that below 450° C the quantities of tar and ammonia which could be collected from 100 grams of coal wegg hardly sufficient to enable us to make any determinations. On the other hand at 600° C the gases could be removed before any appreciable decomposition of the ammonia or tar takes place. In the distillation of coal it is universally known that the respective yields of coke, tar and gas depend upon the conditions of the process. Such factors, for instance, as temperature, pressure, time and contact surfaces will alter the yields in appreciable amounts. Consequently all attempts were made to keep other factors, beside the temperature, constant in different experiments. For this purpose the apparatus was used as is shown in . . . . - 28 - the diagram (Fig. 3) 100 grams of coal in the finely ground ccnditionwer e placed in A for each run. By lowering F as much of the air is withdrawn from the apparatus as possible. Carbon dioxide is then passed through the apparatus through the tube C for one-half to three quarters of an hour. The heating is done electre^Kfcically and the temperature kept constant during each run by a series of resistances. As temperature and time are important factors in distillation processes, the rate of heating of the furnace is shown in (Fig. 4) . The tar collects in B, and the gases, after being filtered through glass wool in C, pass through the two flasks D, each containing about 100 cc of approximately normal sulphuric acid. Finally the gas is collected in E under reduced pressure obtained by lowering F to the same extent in each case. Four coals are heated in this way. In each case gases begin to come off as soon as the temperature begins to rise. This steady evolution of gas continues throughout the experiment and at about 400° C 5 a further decomposition in the coal is shown in the appearance of tar. The coal is subjected to each temperature for periods of five hours. The ammonia given off is collected very efficiently in the two conical flasks D of 150 cc capacity. The contents of the flasks are finally transferred to a measuring flask and made up to 200 cc. Aliquot parts are distilled in the presence of sodium hydroxide into standard acid and the amount of ammonia thus calculated. The tar and water given off are collected in B, A -29- separation of these products is brought about by the centrifuge and the respective quantities measured. Analyses of the gases given off in each case sho?»’ed small percentages of nitrogen. These amounts evidently arose, partly from the occluded gases which are always present in coals and partly from the decomposition of ammonia. The amounts were small and consequently neglected in the calculations. The results are shown in tables 4 to 11. In tables 12 to 15, the percentage distribution of nitrogen is given at each temper ature. In these cases the error, that i3 the per- centage of nitrogen unaccounted for, is given. In cases where the error is indicated by a positive sign, it appears that the sum total per cent of nitrogen in the coke, tar and ammonia is more than in the original coal. In other cases indicated by the minus sign, some of the nitrogen still remains unaccounted for. The discrepancy may be due to inconsistencies in the Kjeldahl- gunning method for determining nitrogen in coke. Some of these errors are greater than experimental error would warrant. Yet considering that the quantities of coal and coke are taken to the nearest gramme, on the whole the results agree remarkably well], and give us a very excellent idea of what happens to the nitrogen part of the molecule when coal is distilled between these temperatures. These results are explained in another way in the diagrams shown in Figures 5 to 13. Here we have endeavored to show the behaviour of each coal separately as it passes from one temperature to another. Thus Figures 5 to 8 show the respective amounts of volatile matter, tar, ammonia and gas of each coal at o NO SHOWING FURNACE RISE IN TEMPERATURE WITH TIME -33- Table 4 SHOWING ACTUAL QUANTITIES GIVEN OFF AT 450 ° c Experiment Number Amount of coal taken (gms) Coke Tax nh 3 Gas H a 0 3 A 100 88 3.23 .0085 6325 3.3 7 B 100 83 5 0070 6150 4.0 11 C 100 87 3.3 0081 5400 4.5 14 D 100 83 3.10 0010 5700 3.00 Table 5 SHOWING ACTUAL QUANTITIES GIVEN OFF AT 500° C Experiment Amount of Coke Tar nh 3 Gas H a 0 Number coal taken 3 E 100 87 3 . 90 .011 8900 3 ,5 7 F 100 81 5.6 0075 8400 5.8 11 G 100 86 3.53 0087 7200 5.8 14 H 100 79 6.00 0021 8000 7 -33- Table 6 SHOEING ACTUAL QUANTITIES GIVEN OFF AT 550° C periment Number Amount of coal taken (gms) Coke 3 I 100 64 7 J 100 7S 11 K 100 84 14 L 100 78 Tar WH 3 Gas H 8 0 3. S3 .054 13S00 3.5 5.7 .016 9950 o 5.80 .0088 8450 6 6.30 .016 9400 7 Table 7 SHOWING ACTUAL QUANTITIES GIVEN OFF AT 600° C Experiment Number Amount of coal taken (gms) Coke Tax nh 3 Gas H 2 0 5 M 100 82.5 4.10 056 15000 4.5 7 N 100 77 5.7 034 11900 6 11 0 100 61 3.82 023 10600 6 14 P 100 76 6.46 043 11650 7 -34- Table 8 SHOWING PERCENTAGE OF NITROGEN IN PRODUCTS GIVEN OFF AT 450 ° C Experiment Number Coal Coke Tar nh 3 Gas 3 A 2 « 0 '& 2.70 .85 0067 — f 7 B 1.68 2.03 .95 0058 1.5 11 C 1.61 1.85 .92 0067 2.8 14 D 1.78 2.07 .60 0008 1.9 Table 9 SHOWING PERCENTAGE OF NITROGEN IN PRODUCTS GIVEN OFF AT 500° C Experiment Number Coal Coke Tar NHg Gas 3 E 2.32 2.68 .95 .009 2.5 7 F 1.68 1.89 .99 0062 4 11 G 1.31 1.80 .97 0072 2.8 14 H 1.78 2.07 .92 0017 2.3 Table 12 -36- SHOWING PERCENTAGE DISTRIBUTION OF NITROGEN AT o 0 O 9 3 7 11 14 $ f distrib. jo jo distrib. jo jo distrib. jo jo distri Coke 2.38 28.60 1.68 93.94 1.61 97.75 1.72 98.85 Tax .027 1.12 .047 2.71 .03 1.82 .019 1.09 nh 3 0067 .27 .0058 .34 .0067 .41 0008 .05 Error +09 + .05 + .034 - .04 Table 15 SHOWING PERCENT AGE DISTRIBUT I ON OF NI TROGEN AT 500° C * jo distrib. 7 11 14 jo j> distrib. jo jo distrib. jo jo distr Coke 2.34 28.07 1.53 96.23 1.55 97.48 1.64 96.47 Tar .037 1.80 .055 3.46 .034 2.14 .055 0 . 23 nh 3 .009 .38 .0062 • CO .0072 .45 .0017 .10 Error + .07 1 • o CO .02 00 0 • 1 -37- Table 1 4 SHOWING PERCENTAGE DISTRIBUTION OF NITROGEN AT 550° C 3 7 11 14 %distrib. * ^distnb. fo distrib f-distr ib Coke 2.24 96.13 1.53 95.62 1.55 96.88 1.64 95.57 Tar .043 1.84 .057 3.55 .041 2.56 .062 3.61 NH S .045 1.93 .016 .9 .0073 .46 .014 .82 Error + .01 -.08 -.01 CD O • 0 Table 15 SHOWING PERCENTAGE DISTRIBUTION OF NITROGEN AT 600° C 3 7 11 14 i° fodistrib fo fodistr ib * fodistr ib. 1o fcdistrib Coke 2.23 96.12 1.47 94.53 1 .54 96.25 1 .61 94.16 Tar .045 1.94 .057 3.65 .042 2.63 .064 3.74 nh 3 .046 1.98 .028 1.79 .019 1.19 .035 2.04 Error 12 -.01 -.07 -58- ths different temperatures; Figures 9 to 10 the percentage of nitrogen in tar and ammonia at each temperature and Figures 11 to 13 the percentage distribution of nitrogen in coke, tar and ammoni a at the respective temperatures. A comparative study of these graphs is very useful and brings to light some very interesting facts regarding the consti- tution of coal. Thus in Figure 6 showing the quantities of tar given off, there is a decomposition of the coal at 450° (or slightly above) yielding a rapid evolution of tar. At 500° C most of the tar has been given off in all cases except that of No. 11, which continues gradually up to 550° C before a lag in the curve is observed. On the other hand in the case of ammonia (Fig. 7) the rapid evolution of ammonia only begins at 500° C, This again is true in all cases, except that of No. 11, when the rapid evolution only begins at 550° C. 'S’hus in all cases it is shown conclusively that the molecule containing the bulk of the nitrogen only starts decomposing after all the tar has been re- moved. Furthermore this decomposition follows directly after the expulsion of the tar. A small portion of the nitrogen, however, comes off with the tar and is closely associated with it. Furthermore it is seen that at 600° C these coals yield quantities of nitrogen as ammonia in direct proportion to the total quantity of original nitrogen in the coal. Thus in Figure 10 coal No. 3, containing the highest percentage of nitrogen liberates also the greatest amount of nitrogen as ammonia at 600° C. Coals No. 14, No. 7 and No. 11 follow in order. This, however, is not the case at lower temperatures. At 500° coal No. 11 liberates a greater amount of nitrogen than coal No. 7 . ■*' - . . . ■ • and coal No. 7 more than coal No. 14. This is further evidence that this part of the nitrogen is distinct from the main portion of the nitrogen molecule. The behaviour of the nitrogen in the tar shows the same character (Fig. 9). Here again we find that even at 600° C coals Nos. 11 and 3 and coals Nos. 7 and 14 approach one another very closely in their percentages of nitrogen. In all cases the percentages of nitrogen in the tar approach a certain limit which does not differ to any extent with different coals. Considering these results as a whole, though limited in scope and kind, we are led to believe that the nitrogen in coal exists in at least two forms— the one a less stable, and the other, which is by far the greater part, a very stable form. The less stable form comes over with the tar and is closely re- lated to it. There is no evidence to show that the tar and the nitrogen belong to distinct molecules. On the contrary the general behaviour of the coals as shown' by these graphs give strong indications that tne tar and the nitrogen originate from different parts of the same molecule. ^ Figures 11, 12, 13 show the percentage distribution of the nitrogen in the coke, tar and ammonia at different tempera- tures. These graphs throw some light on another property of the coal, namely its stability towards heat. Thus coal No. 14 distributes the greatest percentage of its nitrogen as ammonia and as tar at 600° C (Fig. 13). Coal No. 14, therefore, de- composes most rapidly after the temperature of decomposition is reached. Coal No. 3 on the other hand decomposes rapidly up to - 40 - 550° but stops decomposing, as far as nitrogen is concerned, at this temperature, indicating greater stability of the nitrogen molecule. The decrease of the nitrogen distribution in the coke (Fig. 11) illustrates the point in question in a still more interesting way. In all cases except No. 3 the decrease is continuous. In the case of No. 3 the decomposition stops at 550° C. This temperature of greater stability is important and seems to be distinct for each coal. Probably at this tempera- ture we get a decomposition of another kind leaving the nitrogen in a form which is not easily decomposed by heat. It is also possible, however, that the more stable form of nitrogen men- tioned above, itself exists in two forms — the one more stable than the other. It is further possible, and very likely, that we have here a case of polymerisation. The changes which take place follow one another in direct succession, which suggests such a state of affairs very strongly. Frans Fischer (4S) believes that the humic acid present in the coal polymerises during the course of its formation to form humin. It is not unlikely that this same condition exists in the resinous part of the coal and that the intricate nature of all coals is due almost entirely to the degree of polymerisation of the compounds present. Figure 8 shows the yield of gas given off at different temperatures. There are no sudden breaks in these curves. In fact the increase in gas is more or less gradual and consistent between these temperatures. It appears that the gas originates from a different part of the coal altogether, and for the sake of convenience we might look upon coal as being made up of two ■ . . . . . - . . . . “- 43 " Fig. 7 SHOWING QUANTITIES of AMMO NI A GIVEN OFF BETWEEN 4S0°C-680 o C 44- Fig. SHOWING QUANTITIES OF GAS GIVEN OFF BETWEEN J$60°- 660 °C * - 46 - SHQWING PERCENTAGE NITROGEN GIVEN OFF AS AMMONIA BETWEEN 460 °- 600 °C -48- Fig*. 13 SHOWING PERCENTAGE DISTRIBUTION OF NITROGEN IN TAR BETWEEN 4S0°-660°C -4 ?<*- Fig. 13 SHOWING PERCENTAGE DISTRIBUTION OF NITROGEN GIVEN OFF AS AMMONIA BETWEEN 4Q'0°- GOO °C -48V- porticns, (l) the gas producing and (2) the tar producing portion By the gas producing portion is meant that portion of the coal which supplies the more useful gases such as hydrogen on the action of heat and is therefore equivalent to what is often call- ed the "cellulosic" portion. The "tar producing' portion" is that part of the coal which has resulted from the final combination of those substances originating from the resins and proteins in the plant. This line of division is not intended to be absolutely marked, but rather one of degree. It is evident that the hydro- carbons in gas originate from the tar producing portion of the coal. But in the main this part of the coal is not gas produc- ing. The division is more exact than those previously suggested and is in agreement with the theory held by Taylor, Porter, Thiessen, Parr and others, that the cellulosic portion of coal is the first to decompose on heating, rather than that held by Wheeler and his co-workers, that the cellulosic or "hydrogen producing" part of the coal is the more stable to heat. The Effect of Steam on Coke Residues Following these results the effect of steam at 650°C and 750°C was tried on the coke residues obtained at 500°C and 600 °C. The apparatus used is shown in Figure 14. 2 gms of coke were placed in the U tube D, made of pyrex glass. In each side of the tube is placed a plug of asbestos wobl and finally small fireclay blocks about the size of a pea. Steam is generated in A, and at the same time tne temper— --49“ -ature of the furnace is raised until 850 °C is reached. The gases are then passed through a condenser and into E containing standard ^ sulphuric acid. 10 This process is continued for two and a half hours when the acid in E is renewed and the temperature raised to ?50°C. The contents of E obtained at 650°C are now distilled and the amount of ammonia which was evolved calculated. At ?50°6 the reaction is carried on for 3 l/2»4 hours until no more ammonia is evolved and the contents distilled as before. In this way an estimation is obtained of the exact amount of ammonia given off at these two temperatures in the presence of steam. By weighing the residue and analysing it for nitrogen we are able to calculate the distribution of nitrogen under these conditions. Tables 16,1? show the effect of steam on the coke residues at 650 °C and 750 °C and tables 18,19 the percentage distribution of nitrogen. The nitrogen which is listed as " unaccounted for nitrogen", in the tables is evidently free nitrogen arising from the dissociation of ammonia at this high temperature. The percentage of the nitrogen liberated from the coke is in all cases low. At these temperatures it seems as though steam has little effect in liberating the nitrogen In order to do this a higher temperature is evidently necessary so that the coke and steam may combine to form a oxides of carbon, thereby leaving the nitrogen free to combine with the hydrogen to form the ammonia. ■ . . . - ' V 1 : I :7 rig. i4 — ftyparcijius , showing passage oj steam over coke -52- Table 13 DISTRIBUTION WITH 5a ) °C COKE Nitrogen 3 E 7 F 11 G 14 H as Ammonia 14.18 18.52 18.67 16.57 remaining behind in coke resi- due Unaccounted 69.38 69.34 69.56 72.12 for 16.44 12.14 11.77 11.31 Table 19 DISTRIBUTION WITH 600 °C COKE Nitrogen 3 M 7 N 11 0 14 P as Ammonia 12.26 14.82 14.88 13.59 remaining in coke residue 70.01 69.34 70.01 71.61 Unaccounted for 17.73 15.84 15.11 14.80 - 53 - B, Chemical Reagents In the series of experiments just described it is evident that the nitrogen molecule constitutes the tar producing portion of the coal, and that the structure of this molecule, in one respect, constitutes the structure of coal. In the experiments which are to follow attempts are made to isolate the nitrogen mole- oule or break it up by the action of chemical reagents in such a way that it may be recognised in its original form in the coal. According to the theoretical considerations described above it is evident that, after hydrolysis of protein has taken place , several reactions may again take place amongst the products of decomposition. All of these products being acids, they may act as such on other organic compounds in the plant to form open chain derivatives, or ring compounds of a more stable nature. On the other hand they may remain in the coal as such in a highly poly- r meised form or they may decompose altogether and pass off in the form of Ng and ammonia in the secondary stages of coal formation. In order to get some idea of any such secondary changes which may have taken place, certain chemical reactions were tried on the coal, keeping the temperature in all cases as low' as pos- sible . In the first series of reactions the effect of hydrolysing agents was tried on the coal. Hydrolysing Agents In all these reactions preference was given to the method of examining the residue for nitrogen rather than the ex- . • . -54- tract, $n each case the percentage of nitrogen in the residue was determined and calculated to the ash and moisture free basis. Thus by comparing this figure with the original percentage of nitrogen in the coal, calculated on the moisture and ash free basis a satisfactory indication of the extent to which the nitrogen molecule was affected can be obtained. The following hydrolysing agents were tried. a. Two different strengths of hydrochloric acid were used, namely a 12 6 jo and a 40 fc solution. 50 gms of coal v/ere heated under a reflux condenser for 10 hours with an excess of these solutions. The residue was washed until the wash water is free from acid and the air dried and analysed, b. N. Zelinsky (54), preferred using fomic acid for hydroly- sing purposes, so that in these experiments a 25 fc solution of fomic acid was tried as well. c. Aquous Potassium Hydroxide v/as tried in three different strengths 1, 5 ;2,10 3,50 The residues were washed to • give no coloration with phenolphthaleine , air dried and analysed. The extracts were coloured brownish red, the intensity of the colour increasing with the strength of the solution of potassium hydroxide used. The extract, on treatment with HC1 gave a white gelatinous precipitate soluble in excess. The same precipitate was produced with all other common acids. In excess quantities of alcohol the precipitate was insoluble and was separated in this way from the extract. It was found to be aluminum oxide. The solution was finally evaporated to a small bulk and tested for nitrogen with negative results. a. A 10 fc alcoholic potash solution was used in a similar way « •X . . . -55- (1) at the ordinary pressure and (2) at 3 atmospheres pressure. Table 20 shows the results obtained in these reactions In all these cases a certain amount of resinous bodies are ex- tracted from the coal, together with quantities of alumina and iron The table clearly indicates that the main part of the nitrogen molecule remains unaffected. It is also evident that a very small portion of the nitrogen has been removed. If the reagent attacks only the ash of the coal, then the ash content would be lowered and the nitrogen content corres- pondingly raised, and the percentage of nitrogen on the moisture and ash free basis would not be altered. But if the volatile matter in addition is attacked so as to remove some of it, then the ash and nitrogen content would be raised and the percentage of nitrogen calculated on the moisture and asn free basis would be higher. In all cases a certain amount of the volatile matter is removed. It is, therefore, seen that all of the above reactions have had a small effect on the nitrogen content ofthe coal. An attempt was then made to determine quantitatively the amount of nitrogen that was removed in the above reactions. This nitrogen we would expect to be in the form of acid amides and amino acids, and in order to determine their amounts quantit- atively the following methods were employed. i 20 gms of coal are placed in a one litre round bottom flask and heated under reflux for 3-4 hours with 300CC of hydro- chloric acid (1.1). The mass is then filtered under suction and the residue washed with hot water. The solution is evaporated down to a small bulk, transferred to a measuring flask and made up to 100 CC.50C0 of this amount was k jfcldalilized to give the total 4 CO 3 -H CO 4 »-H O CO l£> MO lO lO in m in o eh a) cv2 CO o O o o o m , 4 4 • • • • • • • o 4 P CM rH CM CM cm CM cm • HO)© CM •H o o rH rH rH rH rH -H rH rH rH -4 i — i lO o O ■o CD to 00 01 O o o CM CM in •JO CM cd CO >H cu • • • • • • • • • ■r4 W rH CM CM CM CM lO CM CM 3 PI • • • « • • • P P CO CO CO 01 CO w CO • • >H : H O Eh Eh Eh Eh Eh Eh Eh CO CO 04 O 4 4 ,4 .4 .4 ,4 4 4 Eh Eh Eh 4 4 CD P O 00 00 00 00 O a m •H a> i — 1 rH rH EH EH O F4 PH P Ph 4 © Eh a a >s t>» O &a o •H •H 4 w W w M H • Eh Eh •H a P a P o O o a H o a 4 o p 4 rH rH <4 P 3 4 H p •H 4 © 4 4 •rH •H ■rH H •H tO rH g o O o CO CO CO O CO o to -P o o •H GO CO CO 4 CO cS ,4 4 Eh El a 4 & $ 3 O 4 O c8 3 © 4 4 Eh P p P o P o © eh & r’S O O o o 4 o O © H O CD EH w Ph PH Ph EH -4 PH m Eh ^ o cM 4 O H *:R 'Cfc. so a-* ^rH CO ^rH Q 4 o m in O O > 3 s in H O EH i H 4H CM rH SO H CM rH -57- amount of amido and amino nitrogen present. The remaining 50 C C of this amount were used to determine the amido nitrogen, by dis- tilling a known quantity in the presence of cream of magnesia. The amido nitrogen alone distills over as ammonia which is col- lected in standard acid. The difference in the two results gives the amount of amino nitrogen present. All fourteen samples of coal weretr eated in his way and the results tabulated (table 21) . It is seen that the amount of amido nitrogen is in almost all cases greater than the amount of amino nitrogen. This is what we would expect, if results obtained by Kelley (55) on the rate of ammonif ication of these forms of nitrogen can be applied in this case. Kelley added protein sub- stances to quartz sand, previously treated with soil infusion and studied the rate of ammonif i cat i on. He f ound, contrary to ex- pect at ion, that the basic nitrogen fraction was more completely ammonified than either the amido or non-basic fraction. Lathrop a little later (56) came to similar conclusions. The amounts of amido and amino nitrogen present in the coal are surprisingly low and in craer to be able to come to definite conclusions regarding this matter, an attempt v^as made to remove all KHg groupings in another way. When carbonyl chloride is passed over heated amines a carbamie chloride is formed, y;hich readily loses HC1 to form an isocyanate. RfiiV cecoce — » R.rsHCoft — > R.n-c-o *2 Hci. . . . » -58- Table 31 SHOWING PERCENTAGE OF NITROGEN PRESENT AS AMIDES AND AMINO ACIDS Coal Nitro gen as Nitrogen as Total "NH 2 " fo of Amides Amino Acids Nitrogen Total 1 .0078 .0048 .0122 .57 2 .010 .0076 .0176 <-• co o 3 .0336 .0026 .0362 1.56 4 .020 .0094 .0294 1.35 5 .0125 .0127 .0252 1,50 6 .0124 .0125 .0249 1.54 7 .0112 .0042 .0154 .92 8 .0155 .0069 .0224 1.46 9 .0140 .0082 .0222 1.33 10 .0309 .0030 .0339 1.60 11 .0175 .014 .0184 1.14 12 .020 .0045 .0265 1.53 13 .0186 .0071 .0257 1.47 14 .010 .007 .0170 .95 -59- The isocyanic esters axe volatile liquids with a powerful un- pleasant smell. If coal contains amino groupings then, when C0C1 is 2 passed over it at higher temperatures, we should expect the corr- esponding nitrogen compounds to volatilize. 100 gms of previously dried coal were placed in a long tube furnace, which was electrically heated. The temperature was controlled by means of a thermometer placed at each end. The C0Cl o was purified by passing it through cotton seed oil and finally over the coal at 250°C and 350°C. The results given in the case of two coals show conclusively that very little action took place. (Table 22) Table 22 SHOW IMG EFFECT OF CARBONYL CHLORIDE Treatment Expt. No. H^o Ash Ohg. % — - free basis h„ 0 & Ash Ng * free 250° 14* 1.36 11. ,37 190 2.18 2.05 350° 14 1.36 11. ,83 1.89 2.19 2.05 350° 2 3,. 1.45 7, ,11 2 • 35 2.57 2.55 We axe now in a better position to formulate our theorjf as to the changes which the protein molecule undergo during the different stages of coal formation. The monamino , diamino acids and acid amides which come into the mass as decomposition products of the protein molecule, are no longer present to any extent in the coal. Consequently these compounds, eitner combined with themselves or with other organic compounds to form more stable ring derivative s , or they are decomposed and pass off from the -60- mass in the form of ammonia or nitrogen , thus leaving the heter ocyclic compounds, which are present in proteins, to form the bulk of the nitrogen found in coal. It is an established fact that amino acids and acid amides will readily ammonify in the presence of bacteria in the soil. We axe, therefore, more entitled: to assume such a state of affairs in the case of coal than we are to assume sjtiGifcx a state of further combination. The conditions during coal formation are favorable for such decompositions. The fact that all coals from the same locality contain amounts of nitrogen which are more or less constant, further suggests a com- pound which is stable to the natural agencies to which it is ex- posed . The heterocyclic compounds are the only portions of the protein molecules that will satisfy this condition. If this is the case then in order to account for some of the properties of coal, we rust assume that these compounds have undergone polymerization to a great and unknown extent. i w Selenium Oxychloride According to researches done by Prof. Lenher of the University of Wisconsin, Selenium oxychloride is a very useful and active selective solvent. Its effect was therefore, tried on coal as an attempt to separate the stable nitrogen molecule. Five grammes of coal No. 3 and 25 CC of pure selenium oxychloride were placed in a conical flask and corked. The mixture is warmed and shaken continually for 1/4 hour, when the extraction is assumed to be complete. In order to get this solution to filter so as to separate the residue from the extract . ■ * . . -61- the mixture is diluted with 200 C C of benzene. Filtering under suction is thus rendered possible. The residue is now washed well with benzene to remove as far as possible any adhering SeOCl then with ether and finally dried at 105°C. 2 The extract when heated with water deposits red selenium. When benzene alone is mixed with selenium oxychloride in these proportions and water added to the mixture, the selen- ium oxychloride hydrolyses to selenic and hydrochloric acids with -out any deposit of the element. Thus when selenium oxychloride acts on coal a chemical reaction is assumed to take place and a compound is formed which readily deposits selenium. The extract was treated with excess of sodium hydroxide and steam distilled. Finally it was vacuum distilled but no nitrogen compounds were found. The residue was analysed for nitrogen and results calculated to the moisture and ash free basis. Finally for purposes of comparison the action of selenium oxychloride was tried on 500°C coke obtained in previous experiments (see table 9) . This coke had practically all its tar already removed . The reaction was significant. Hardly any result could be seen, except a coloration of the solution, showing that the selenium oxychloride reacted with the tar in the coal. The mass was easily filtered and the residue washed with benzene and ether and finally analysed. The results are given in table 23. Table 2b SHOWING ACTION OF SELENIUM OXYCHLORIDE Treatment Expt. H*Q Ash N; Orig. N s Coal & SeOCL: 32 HaQ.Ash free H a Q. Ash fr 6.23 1. 79 2.55 be -62- Table 33 SHOWING ACT Treatment Exot . h 2 Q Coal & SeOCl a 3' (residue) Coke & SeOCl 2 3” 2.07 (residue) orignial coke 3 ,,f .54 550 °C ( 3E) From these figures it appears that the nitrogen in the coke remains unaffected. In the case of the coal it would appear as though some of the nitrogen was removed. Selenium oxychloride on the other hand is extremely active dnd forms a compound with the coal, so that it is impossible to rid the residue of the reagent. This is seen very forcibly by the result obtained in a volatile matter determination on the residue. It was found matter that the volatile/on treatment with SeOCl 2 increased in the case of coal NO§ from 33.10 to 31.34 $>. Taking these figures into account the two results coincide showing* conclusively that selenium oxychloride will not attack the stable part of the nitroge:] molecule . The Action of Reducing Agent 3 Subsequent to these results the coal was subjected to the action of reducing agents in order to observe what effect hydrogenation would have on the nitrogen molecule. When pyrrol is reduced pyrroline is obtained. With strong reducing agents, such as red phosphorus and hydr iodic acid pyroline is further reduced to alkylarnine or still further to ON OF SELENIUM OXYCHLOrtlDE Ash 6.23 N a 1.79 HpO.Ash free 1.91 Orig. N h 2 0, A S.55 7 .94 2.65 2.94 2.96 9.14 2.68 3.96 2.96 lh fred * . 1 ■ ■ hydrocarbon and ammonia as is illustrated by the following form- ulas: & — > CH 3 (CHz) 3 NHz — » CHi(CHu) % CH 3 + flH 3 . If the nitrogen in coal is present in the form of pyrrol derivatives, as we have been led to believe, then strong reducing agents should have some effect on the nitrogen molecule. In order to try out this reaction propyl alcohol and sodium, and red phosphorus and hydriodic acid were used as reducing agents (a) Proovl Alcohol and Sodium 20 gms of coal were placed in a two litre round bottom flash containing a reflux condenser and 500 gms of propyl alcohol added. The mixture is warmed to boiling and small pieces of sodium added. The flask is shaken continually so that the contents are kept mixed all the time. When no more sodium reacts sufficient water is added to decompose any free sodium, At this stage the coal assumes the appearance of a soft, oily resinous mass, which floats on the surface of the liquid. The mixture is cooled, filtered washed and air dried. But while being air dried the coal seems to undergo oxidation again and is converted back into its original appearance. Table 24 shows its analysis on the moisture and ash free basis. Evidently very little nitrogen was removed in this case. (b) Hvdrlodio acid and Red phosphoru s The reaction with hydriodic acid and red phosphorus at higher temperatures is more difficult to carry out successfully owing to the enormous pressure set up during the reaction. A mercury container with a specially designed screw cap was used. -64- At 200°C the reaction was carried out successfully but at higher temperatures the container failed to withstand the pressure entirely and leaked through the cap. For this reason experiment ( 14C) could not be run satisfactorily, though a certain measure of success is evident from the amount of nitrogen that was re- moved in the process. 50 gms of coal were mixed with 10 gms of red phosphorus and 350 gms of hydrogen iodide of 50 strength. The mixture was heated for two hours (a) between 180-200°C and (b) 250°-280°C. On cooling and filtering the residue and extract in the latter case were treated as follows: ( 1) Treatment of residue; The residue was treated with iodine and water and boiled for a short time to remove any excess phosphorus. The phosphorus and iodine combine to form PI a which is decomp- osed by water, forming phosphoric acid which is soluble and hydrogen iodide which passes off as a gas. The excess of iodine is removed by 'washing with a strong solution of potassium,. Iodide and finally with 7/ater to remove the latter. Finally it is air dried and an aly sed, (Table 24) ( 2 ) Treatment of the Extract The extract was a green color. On diluting it with water a brilliant yellow precipitate settles out. This precipitate contained carbon, hydrogen, oxygen and iodine. On boiling with water it hydrolyses to a white insoluble sub- stance containing the first three elements mentioned above. As the precipitate contained no nitrogen it was not investigated -65- any further. On removing this precipitate, however, the solution v/ as steam distilled in the presence of sodium hydroxide. The steam distillate had a distinct ammoniacal smell and a ave a dense precipitate with Nessler' s reagent. No amimes could be detected •with benzene sulphonyl chloride. Table 24 (exp. 14C) shows that quite a large proportion of the nitrogen was removed in this process. Table 24 SHOWING EFFECT OF REDUCING AGEN TS Treatment Expt . H a 0 Ash n 2 N a f ree Orig N a free H 2 0 & Ash basis H a O & Ash basis Na & c 3 h ? oh 14 A 3. 29 10.52 1.69 1 . 98 2.05 HI & Phos. 180-200 14 B 1.44 14.10 1.67 1.98 2,05 HI & phos. 250-280 14 C 1.16 30.26 1,04 1.52 2.05 In these results we have further indiaations of the presence of pyrrol derivitives in coal. The fact that such a high temperature and pressure is necessary in the reaction further suggests polymensati on of these compounds. . . » - 66 -' SUMMARY OH THE ROEk Off iUTROGM IH OPAL The experimental data arrived at in this work indicates certain theories regarding the constitution of coal and the form in which the nitrogen exists. All past theories on coal take into account cellulosic and resinic decomposition products and state that coal is made up of these two portions alone. This work shows conclusively that the nitrogen, or the decomposition products of protein cannot be neglected. The nitrogen molecule is closely related to the resinic port- ion of coal , and forms different parts of tne same molecule from which the tar originates. Two main forms of nitrogen have been established in this work; l: RH2 form, present as amino acids and acid amides, and; £: a more stable form, probably in the form of a pyrrol ring which has undergone polymerisation. The BH2 form does not constitute more than 1*6% of the nitrogen in the coals studied. As amino acids and acid amides readily undergo ammonificati on in the presence of bacteria, it is quite probable that they decompose and pass off from the coal mass during the period of formation. The heterocyclic com- pounds, which are present in proteins in the form of pyrrol rings, on the other hand, remain behind and account for the more or less constant percentage of nitrogen in all coals from tne same locality. The pyrrol derivations undergo poly- merisation to an unknown degree in the presence of the organic acids in the vegetal matter, and combine with the products of ■ - 67 - decomposition arising from the resinic matter in plants. These products of decomposition are thus present in the form of long side chains attached to the polymerised pyrrol ring. The author would suggest that this compound be called the "tar producing portion" of the coal. When coal is heated tnese side chains split off and distil over in the form of tar. The HUE nitrogen comes off with the tar and quite probably is present in these side chains. This accounts for the small percentage of ammonia given off at low temperatures. When practically all of the tar has been split off, the polymerised pyrrol derivative starts decomposing in stages and liberates ammonia and hydrocarbons. relatively This accounts for the/ small percentage of hydrocarbons found in gas. In this case acetylene will evidently be formed. Part of this acetylene combines with hydrogen to form ethylene and methane, and a further portion polymerises to form benzene, if the conditions are favorable. Finally a stage is reached where the polymerised compound is unsaturated and at the same time unstaole. At this temperature the ring breaks and a com- pound remains in which nitrogen is probably combined directly with carbon. When this stage has been reached the foeiat way to obtain the nitrogen in the form of ammonia, is oy burning a- way the carbon in the presence of steam. On these lines the presence of pyrrol and pyridine in coal tar can be explained, since it is a distinct property of five membered heterocyclic compounds containing nitrogen to be converted into six membered ring compounds. Pyrrol derivht- ■ - 68 - “ ives furthermore readily polymerise in the presence of acids and these polymerised products decompose on heating to yield ammonia. The fact that polymerised pyrrol derivatives are un- stable at higher temperatures indicates that the degree and kind of polymerisation which takes place during the stages of coal formation is yet unknown. In regard to the actual behavior of the nitrogen in the coals studied, the following conclusions are drawn :- 1:- In distillation experiments carried out between 450 degrees and 600 degrees 0 the percentage of the nitrogen remaining behind in the coke is distinct for each coal and is independent of the total amount of nitrogen in the coal. As much as 98.85$ of the nitro- gen remains in the coal at 450 degrees C in one case. In this instance only .05$ of the nitrogen is liber- ated as ammonia. At 600 degrees G an average of 95$ of the nitrogen remains in the coal and not more than 2.0 4% of the nitrogen is liberated as ammonia. On further heating to 750 degrees G in the presence of steam an average of 70% of the nitrogen remains in the coke. 2:- Hydrolysing agents will affect the M2 nitrogen in the coal, but not the more stable form. 3:- In the coals studied the percentage of HH2-nitrogen varied from .57% to 1.60$ of the total nitrogen present. In almost all cases the greater part of this nitrogen was in the amido form. - 69 - 4:- Selenium oxychloride will react with the tar in coal and form a colloidal mass with it. It does not attack the stable form of nitrogen. 5:- Strong reducing agents will attack the nitrogen in coal to yield ammonia. Thus red phosphorus and Hydriodic acid at 250 degrees to 280 degrees reduces the percentage of nitrogen from 2.05% to 1.5%, calculated on the moisture and ash free basis. - 70 - BIBLIOGRAPHY A. On the nitrogen content of coal and the distribution on di stillation. 1. Swindells B. P. , June 1844. 2. Berthelot, M. Comptes Bendas 67--1141. 3. Henin, J. Gaslight (1892) p. 296. 4. Mayer, M. H. F. , Altmayer, 7. The formation of ammonia in the dry distillation of coal. J.S.C.I. 26. 125-6. (1907) 5. McLeod, J. Redistribution of nitrogen in the distill- ation of coals. J.S.C.I. 26. 135-6. (1907) 6. Short, A. Carbonization of Durham Coking coals and the distribution of nitrogen and sulphur. J.S.C.I. 23, 581-85. 7. Woltereck, H.C. Production of ammonia and the re- covery of nitrogen in heat. Comp. Rend. 152, 1245-7 (1911) 8. Summer sbach, 0. Formation of ammonia and cyanogen in coal distillation. Stahl und Risen 34, 1153-9. (1914) . 9. Cobib®, J." r . A problem of modern gas practice. J. Gas Lighting 126, 329-3. (1914) 10. Summersbach, 0 (a) Yield of nitrogen in coal. Coll. Guard 109, 1020 (1915) 11. ” (b) Formation of ammonia and cyan- ogen during carbonization. J. Gas Lighting 131, 246-7 (1915) . -71- 12. Terres, E. A study of the formation of the nitrogen in coal and coke. Chem. Ztg. 39 — 73 (1915) J. Gasbel 59 519 — 21. 13. Porter, H.C. Goal and coke by products as a source 'for fixed nitrogen. Chern. Met. Eng. 15, 470-5. (1916) 14. Mahler, P. Nitrogen content of oxidised coals. Comp. Rend. 165 634 (1917) 15. Gluud, W. , & Brener, P.K. Distribution of the nitrogen of coal on low temperature carbonization. Ges Abhandt Zur Remit) nis der Kohle 3. 227-37. (1920) . 16. Monkhouse & Cobb, J.W. Liberation of nitrogen from coal and coke as ammonia. Gas J. 156, 234-40. (1921 ) 16a. Chiles, H.C. The form of nitrogen in coking processes. University of Illinois. Thesis (unpublished ) 1920. B. On the Constitution of Coal. 17. Guignet , E. Constitution of Coal. Comp. Rend. 88, 590. j (1879) J.C.S. 36. 602. (1879) 16. Friswell , R.J. Notes on the action of dilute nitric acic on coal. Proc. Chem. Soc. (1892) 105-9. 19. Smith, A contribution to our knowledge of the sol- uble and resinic constituents of bituminous coal. J.S.C.I. (1891) 10 975. 20. Smythe , The proacinonate constituents of coal. Interim Report of Comm, of B. Assoc. (1894) p246. 21. The constituents of coal. Interim Report of Comm, of B. Assoc. (1896) p 340. ' - 72 - 22. Anderson, W. C. & Roberts, J. , Some Chemical properties of Scotch Coals. J.S.C.I. (1898) 17. 1013- 23. Baker, T. Solvent action of Pyfidirle on coals. J.S.C.I. (1901) 20 789. 24. Bona tn B. , & Margosches, B. An addition to the invest- igation of coal. Chem Ind. (1902) 226. 25. Bonath--Britnn Eossil Coals. Zeit £ Angew Chem 19, 657 ( 190o ) x iro 26. Bedson, P.P. Botes on the proscejmate constituents of coal. J.S.C.I. 27 147 (1908) 27. Boudouard, 0. humic substances from coal. Comp. Rend. 147 486 28. Lewes, V.B. Progressive age 29, 1030 (1911) 29. Pictet, A. & Ramseys L. Constituents of coal. Be. 44 2486-97 (1911) 30. Burgess, M. J. Sc Wheeler , R.V. The volatile constituents of coal. J.S.C.OI. 30 606 (1911) 31. Eraser & Hoffman The constituents of coal soluble in phenol. Tech paper 53. of Mines. (1912) 32. Pictet, A., h Ramseys L. Hydrocarbons from coal. Arch Sci. Phys. Hat. 34, 234-49 (1913) 33. Clarke, E.F. , & Wheeler, R.V. Volatile constituents of coal. J.C.S. 103 1904. (1913) 34. Parr, S. W. and Hadley, H. E. The analysis of coal with phenol as solvent. B 76 U. of I. Exp. Stn. 34a. Hager Coal and the Chemistry of its carbonization. J.S.C.I. (33) 389-392 (1914) 35. Jones & Wheeler The constituents of coal. J.C.S. 109 707 (1916) . - 73 - 36 . Porter, H.G. and Taylor, G.B. The primary volatile products of the carlo ni zat ion of coal. Tech. Pap. 140 B. of M. 57. Stopes, M. & Wheeler , R.V. The structure of coal. J. .G.I. 56. 176-8 (1917) 56. White, D. , Thiessen & Davis. The origin of coal. B. 58. B. of Mines (1913) 59. Thiessen. R.. Stricture in Paleozoic Bituminous Goals. B. 117 3. of M. (1920) 40. Hilpert, W. S. , Keller, E. , Lepsius,B. Action of Chemical reagents ©h coal. Ge© . Ahhandt Zur Kennfciris der Kohle (1917) 1 22-5. 41. Groppel, H. 8s Fischer, F. Extraction of previously heated coal. Gee. Ahhandt Zur Kenntnis der Konle I 88-77.(1917) J.S.G.I. 58. 400 A. 42. Fischer & Bigger man , T. Conversion of coal and similar substances into soluble products By ozone. Ges. Ahhandt Zur Kenntfflis der Kohle I. 50-42. (1917) 43. Fischer, F. Extraction of coal under pressure with solvents other than Benzene. Ges. Ahhandt Zur Kenntnis der Kohle 5, 2457 (1919) 44. Fischer, F. , & Gluud, ?/. Extraction of coal with dilute alkalies at nigh temperatures. Ges. Ahhandt Zur Kenntnis der Kohle 3, 243-5 (1919) 45. Pictet, A., Researches on coal. Ann. Chem (9) 10 249-330 (1918) ' - 74 . 46. Fischer, F. & Tropsch Ozonization of coal suspended in a now aquous media. Ges Ah hand t Zur Eenntnis der Kohle II 160 (1919) 47. Fischer, F. & Tropsch. Hydrogenation of various kinds of coal. Ges Ahhandt Zur Kenntnis der kohle I 154-59 (1919) , 48. Fischer, F. & Schrader. The origin and chemical struct- ure of coal. Brenstoff Ghem 2 37-45 (1921) " " 213-9 (1921) G. General. 49. Mineral Industry 1920. 50. Hlaswetz & Haberman. Ann. 169 150 (1873) 51. Eossel Zeitz Physiol. Ghem. 22 - 176 52. Eutscher ” tt TT 41 - 407 53. Emil Fischer n ii TT 43 - 151 54. Zelinsky, I. Ghem. Ztg. 36. 824 (1914) 55. Kelley Bui. 39 Hawaii Agric. Expt . Station (1915) 56. Lathrop Soil Science 1, 509 (1916) 57. Hohart , F.B. The effect of carbon dioxide in carboniz ation processes. Univ. of 111. Thesis (unpublished) (1921) 58. Yliet, E. B. , Trie classification of coal. Univ. of 111. Thesis 1918. ' — VITA The writer was horn in Gape Town, Union of ^outh Africa on January 21, 1897. He attended the South African College High School during the years 1909 — 1914, and entered the South African College, now the University of Capetown, in February 1915. While at this institution he was enrolled as a student in Liberal Arts and Science and graduated in Chemistry, Physics and Mathematics in December 1917. A year later he obtained the degree of master of Arts in Chemistry and finally spent a year as graduate assistant in that de- partment. In February 1920 he entered the Graduate School of the University of Illinois. I )