Search

The experiments which will be described in this and succeeding papers of the series are a continuation of the previous researches of Bone and Wheeler upon “The Combination of Hydrogen and Oxygen in Contact with Hot Surfaces,” which were published in 1906. They were begun at Manchester University twenty years ago, continued at the University of Leeds, and are now being completed at the Imperial College of Science and Technology, London. The delay in publishing the results has not been disadvantageous, because it has allowed of our testing some recent theoretical developments concerning catalytic combustion, and more particularly that due to Langmuir. These, it will be remembered, formed the subject of a special discussion held in London on September 20, 1921, under the auspices of the Faraday Society, in which a number of the leading investigators throughout the world participated. The previous work of Bone and Wheeler ( loc . cit .), in which the actions of a number of widely different hot surfaces upon the combination of hydrogen and oxygen at pressures between 500 and 50 mm. were systematically studied, had disclosed the following outstanding facts:— (1) That the catalyzing power of a new surface at a given temperature usually increases up to a certain steady “normal” state, after which the rate of steam formation is always directly proportional to the pressure, provided that the two gases (hydrogen and oxygen) are present in their combining ratios, and that the product (steam) is rapidly removed from the sphere of action.

In connection with the researches upon catalytic combustion conducted in my laboratories for some years past, a great deal of attention has been paid to the experimental investigation of the important question of whether or no the presence of moisture has any specific influence upon the catalytic combustion of carbonic oxide. The present paper embodies the principal results of our investigations up to date. They were begun in 1908 at the University of Leeds, in collaboration with the two Gas Research Fellows—Mr. A. Forshaw, M. Sc., and Dr. H. Hartley—as well as with Mr. A. Appleyard, B. Sc., and have been completed at the Imperial College of Science and Technology, London, with the collaboration of the late W. A. Haward, M. Sc., Mr. S. Robson, B. Sc., A. Whitaker, B. Sc., and Prof. D. S. Chamberlin, of Lehigh University, U. S. A., who recently spent a "sabbatical year" working with us. The difficult nature of the experimental work involved operations demanding the utmost care, great attention to details, and unusual skill on the part of my collaborators. The surfaces experimented with were porous porcelain, the oxides of copper and nickel, gold and silver, And, in view of the importance of the results in connection with the theoretical aspects of the subject, many independent repetitions of the experiments were made at various times to ensure their complete confirmation. In the case of a very porous surface, such as fireclay or porous porcelain, it was ultimately found necessary to extend a given experiment uninterruptedly day and night over three months, in order to ensure complete dryness of the system. Indeed, not until the later stages of the research was it possible to view correctly and reconcile all the results.

After investigating the action of gold in causing the union of carbon monoxide and oxygen at temperatures between 275° and 325° C., as already described in Part I of this series,* we turned our attention to nickel, copper and their oxides, as examples of surfaces which certainly are not chemically indifferent to the reacting gases. For, as is well known, freshly reduced nickel forms a carbonyl with carbon monoxide at quite moderate temperatures (30° to 80° C.), and copper (as we have found) is by no means indifferent to the gas at somewhat higher temperatures. Again, whilst copper very readily combines with oxygen at moderate temperatures, nickel is more resistent to oxidation, but, like iron, will burn if sufficiently heated in oxygen. Nickel, therefore, has a greater affinity for carbon monoxide, but a less affinity for oxygen, than has copper ; also, their oxides are all readily reducible by carbon monoxide at moderate temperatures. Altogether, the properties of these two metals and their oxides are such as would seem to be highly conducive to “catalytic” activity towards mixtures of carbon monoxide and oxygen, if, in any circumstances, such activity is essentially chemical. Hence our choice of them for a detailed experimental study, the results of which are summarized in this paper. The apparatus and experimental method employed were in all essential respects the same as those already detailed in Part I hereof, namely, the “circulation method” as introduced by Bone and Wheeler, with the inclusion of a suitable arrangement for continuously removing the reaction product (CO 2 ) from the system by means of a solution of purified barium hydroxide. Also, the plan of recording the experiments in their chronological sequence will again be adopted, using the same symbols as before.

An account was given in a paper entitled “Proknocks and Hydrocarbon Combustion” (Ubbelohde, Drinkwater and Egerton 1935) of some experiments made to trace the source of the nitrogen peroxide which had been found by sampling the products from the cylinder of a petrol engine at various stages of the stroke. Those experiments indicated that it was not simply a matter of nitric oxide formed by the flame giving rise to the nitrogen peroxide, for different results were obtained using different exhaust-valve surfaces. Nevertheless it seemed probable that the flame should be mainly responsible for the formation of nitric oxide, and so further experiments have been made. In order to make progress it was essential to determine the amount of nitric oxide as well as the amount of nitrogen peroxide, and analytical methods had to be devised to do this. The first part of this note deals with the methods of determining small quantities (of the order of 10 -4 mol. fraction) of total nitrogen oxides and of nitrogen peroxide, and the second part with the results of analyses of the gases sampled from the cylinder of internal combustion engines by the methods described by Egerton, Smith and Ubbelohde (1935) and by Drinkwater and Egerton for the C. I. engine in a paper shortly to be published.

In a former communication to the Society, Bone and Wheeler dealt very fully with the action of a porous porcelain surface in promoting the slow com­bination of hydrogen and oxygen at 430° C. It was shown (inter alia) (i) that previous exposure of the surface to hydrogen at such temperature stimulates its catalysing power towards normal electrolytic gas in a very remarkable degree, and (ii) that when excess of one or other of the reacting gases is present, the rate of combination is always proportional to the partial pressure of the hydrogen. In Sections A and B of the present communication, similar experiments will be described upon the behaviour of a fireclay surface at 500° C. towards mixtures of carbonic oxide and oxygen, showing (i) that previous exposure of carbonic oxide has much the same stimulating effect upon its catalysing power as was previously observed by Bone and Wheeler in regard to a porcelain surface and hydrogen, and (ii) that in most other respects the surface catalysed the com­bination of CO-oxygen mixtures in much the same way as does one of porcelain that of hydrogen and oxygen mixtures, except that in the former case the catalys­ing power is feebler than in the latter. In Sections C and D will be described comparative experiments upon the relative speeds of the catalytic combustion of carbonic oxide, hydrogen and methane, respectively, in contact with a fireclay surface at 500° C.

A study of the electrical condition of hot surfaces in contact with gases has revealed the fact in several important respects a striking parallel exists between such electrical condition and the catalytic properties of the surfaces. Thus, it has been found that, whereas at comparatively low temperatures the specific effect of a surface upon the potential it acquires in contact with a gas is hightly pronounced, with increasing temperature such specificity tends to disappear until, at a sufficiently high temperature, the value of the surface potential is almost wholly determined by the nature of the gas and is nearly independent of the nature of the surface itself. This may be compared with the fact, established by Bone and his co-workers, that the differences between the catalytic powers of various surfaces, which are often considerable at low temperatures, diminish as the temperature is raised, until they practically disappear. Furthermore, it is now well recognized that the catalytic properties of a surface are intimately connected with its structure, and a close parallelism has been observed between the surface potential of a metal and changes in its superficial structure brought about by heat treatment. Finally, it has been found that the electric current which can be drawn from a heated metallic surface in contact with a gas increases rapidly with increasing temperature, an observation which has its parallel in the well-known rapid increase in the catalytic activity of such a surface with increasing temperature. The present investigation was undertaken with the object of determining the relationship, if any, between the role played by moisture in promoting the heterogeneous catalytic combustion of carbonic oxide with oxygen and the structure of the surface as revealed by its catalytic activity and electrical condition.

In 1905 Bone and Wheeler, in their memoir upon "The Combination of Hydrogen and Oxygen in Contact with Hot Surface," drew attention to the possible connection between the electrical conditions of hot surfaces and their power of rendering "condensed" or "occluded" gases chemically active. Shortly afterwards Bone and Makower, in a series of unpublished experiments, observed that a gold surface becomes negatively electrified during the catalytic combination of hydrogen and oxygen at 300° or there abouts. The recent inventions, by Bone and his co-workers, of processes of "flameless incandescent surface combustion," whereby the constituents of an explosive gaseous mixture, in correct combinig proportions, are caused to combine at a highly accelerated rate, and without development of flame, in contact with (or within the pores or interstices of) an incandescent solid, have again called attention to the important rôle played by hot surfaces in accelerating combustion, and to the great economic advantages derivable thereform. The phenomenon is now known to depend upon a peculiar "activation" of the combustible gas, and probably of the oxygen also, by association with the surface. At relatively low temperatures the intensity of such action varies with the chemical characters and physical condition of the surface, but it rapidly increases with the temperature until at bright incandescence there is little to choose between the catalysing power of chemically different surfaces, porosity and refractoriness being now the dominant factors.

In connection with investigations now in progress in these laboratories, on the cathodic and heterogeneous catalytic combustion of gases and on the electrical condition of metals, the need has arisen for a knowledge of the nature of the structure of certain metal films, whether obtained by cathodic sputtering or by thermal evaporation and subsequent condensation. It has previously been found by spectrographic means that metal sputtered into the cathode zone of a cold-cathode high-tension arc appears therein, at least in part, in the atomic state; but the results do not enable it to be inferred that aggregates of the atoms are not likewise present. A further important question arising out of the results of our investigations required for its solution the preparation of oxygen-free metal surfaces and their subsequent examination in the absence of oxygen.