Papers - Classification - Reactivity of Anthracite with Carbon Dioxide (With Discussion)

It is well known that different forms of carbon show differences in chemical behavior under apparently identical conditions. It is fairly well known that these differences persist, although to a lesser degree, in a single group of carbons, such as the anthracites of Pennsylvania. It is the purpose of this paper to show the extent of these variations for Pennsylvania anthracites from actual quantitative measurements of the amount of carbon dioxide reduced under the standardized test conditions employed. As the amount of carbon dioxide reduced to the monoxide (under comparative conditions termed, for convenience, the "reactivity") is important in combustion, gas production and metallurgy, it is hoped that this paper will be a useful contribution to the store of knowledge on these subjects. In carrying out the investigation, considerable time and thought were spent in obtaining what seemed to be the best possible design of apparatus. The next step consisted in making a number of trial runs on the same sample, varying temperature, particle diameter, etc. and then outlining a standardized procedure to be followed in testing all samples. Fifty anthracites were tested. As a matter of interest, some 10 miscellaneous samples were also tested. The data show initial reaction temperatures (abbreviated I.R.T.), reaction velocity constants at 900" and 950" C., some effects of added substances, and a few additional tests designed to throw some light on the causes of differences in reactivity. The sample of charcoal, as was anticipated, proved to be easily the most reactive of any sample tested. The anthracites as a group came second. The five samples of gas and by-product cokes, obtained from five different coke manufacturers, came third, and a single sample of Rhode Island anthracite ranked fourth, with respect to reactivity as above defined. Although the five coke samples do not necessarily represent the average market coke, and five samples constitute a rather small number of samples to serve as a basis for comparisons, the fact that none