The paper draws ancntion to the importance of the part played by peroxidic bodies or active oxygen in the weathering of coal; comparative investigations have been made with both coal and carbon, which are shown to behave similarly.
The initial reaction appears to involve chemisorption, which, in the-presence of water, is followed by the formation of a coalor carbon-oxygen-water complex. The normal inherent wafer content of coals is a function of their internal surface; this is a minimum with coals containing c. 89 per cent carbon. The peroxygen content of coals shows a similar variation w i ~ carbon content.
The peroxygen complex is markedly unstable above approximately 70' c., a temperature long known to be critical in regard to spontaneous combustion of coal. Below it the complex is formed easily and remains relatively stable, linle oxygen entering the coal substance ; also the amount of ulmification remains very small. The small quantity of gaseous oxidation products consists chiefly of carbon monoxide with a less amount of CO,. Above 7oDc. the complex starts to break down, although with carbon it does not appear to be completely destroyed until 350" c. is reached. Oxidntion above 70"c. is also marked by a big increase in the oxygen content of a coal, its general ulmification and in the yield of gaseous oxidation products ; also carbon dioxide then preponderates over the monoxide. These observations were substantiated by means of gravimctric balances for the coal-oxygen reaction at 50" and p" c. The pressure changes at these two tCmperaturCS were also different ; whereas a gradual fall in pressure was recorded a t 50" c., at goo c. a temporary rise in pressure occurred. The remarkable difference in the behaviour above and lielow this critical temperature is common to both coal and carbon, and in view of the ability of carbon to form a carbon-oxygen-watcr complex, it is evident that with both matcrials the overall oxidation is controlled by the rate of breakdown of the complex. Of paramount importance is the necessity for the presence of water; indeed, an explanation is now forthcoming of the classical experiments of C. J. Baker and H. B.
Baker in 1887-88, which demonstrated that dry carbon cannot be readily oxidized even at red heat.
Bituminous coal absorbs oxygen slowly at ordinary tcmperatures to the detriment of its swelling and caking powers and ultimately also of its calorific value. With large heaps of coal the hazard of spontaneous ignition arises under conditions of limited accessibility of air to the interior of the pile. It is known that coals of lower rank are more liable than others to spontaneous ignition; also that while-on account of slow oxidation-the temperature inside a large coal pile may rise higher than that of the surroundings without this hazard becoming serious, the risk of ignition becomes definitely more acute if the temperature is allowed to rise above a critical range of 70-80ยฐ.1
The object of the investigation described in this paper was to gain an insight into the chemical processes involved in the prc-ignition stages of the oxidation process. It was known from the work of Francis? that weathered coals possess oxidizing properties, a fact in keeping with the suggestion, first made by Erdmann and Stolzenberg3 that the formation of organic peroxides may play some part in their spontaneous ignition. Lamb and Elder' showed that aerated charcoal-also liable to spontaneous ignition-markedly catalyses the air oxidation of ferrous sulphate solution, and showed that traces of peroxidic material could be extracted therefrom. More recently Yohc and his collaborators6* a have traced the growth and decay of the oxidizing power of bituminous coals exposed to oxygen. These authors used the reaction with titanous chloride as an indication of oxidizing power and were able to show that fresh seam coal (unexposed to air) was devoid of oxidizing action.
These facts suggested a similarity between the mechanism of oxidation of coal and that of the oxidative processes undergone by oils, rubber, foodstuffs, in all of which the formation of endothermic complexes (possibly hydroperoxidcs) plays an * Paper read before the Newcastle Section on 7/5/47