A theory of the combustion of hydrocarbons

Abstract

A comprehensive theory of the combustion of hydrocarbons must describe in detail all the analytic and kinetic data. Up to the present no such theory has been proposed, for though the hydroxylation theory of Bone and his school has been very successful in the former field, it is inadequate in its original form to take account of the latter. In the present paper it is shown that a comparatively small modification, involving the introduction of the conception of chain propagation by free radicals, remedies the earlier deficiency and makes possible for the first time a detailed description of the widely varied phenomena of combustion. The work of Egerton, Hinshelwood, Haber, Semenoff and others has led to the recognition that both rapid and slow combustion are autocatalytic in character, the reaction being propagated through the gas from certain initial centres, so that from every centre started by the primary mechanism, a great many molecules of hydrocarbon are oxidized. The first attempt to interpret the chain character in terms of a concrete theory is embodied in the suggestion of Egerton who has extended the peroxidation theory of Callendar to include an energy-chain mechanism by way of which reactivity is handed on from the active products (peroxides) to new reactant molecules. This theory and its various modifications, however, is unsatisfactory in more than one kinetic aspect, the effect of inert gases in particular being the reverse of the deactivation to be expected for an “energy” chain. Moreover, the peroxidation theory is not reconcilable in all respects with the analytical data, for while the induction period has been interpreted as a period of peroxide building no evidence of any such peroxide formation at this stage has been established, nor is the induction period affected by the addition of any such bodies. There thus arises the need of some further attempt to bring the analytic and kinetic data into relation with one another and in the hypothesis developed below, which we shall call the “atomic chain hypothesis,” it is believed that we have a simple explana­tion which encompasses both the analytical and kinetic results in an adequate manner.