The catalytic oxidation of ethylene on palladium films By C. K emball and W. R. Patterson Department of Chemistry, The Queen’s University of Belfast (Communicated by Sir Eric Bideal,F.R.S.—Received 30 April 1962) The oxidation of ethylene has been studied on evaporated films of palladium in the tem perature range 50 to 140 °C, and products were analyzed by means of a mass spectrometer. The main reaction was complete oxidation to carbon dioxide and water, but trace amounts of acetic anhydride and acetic acid were formed by a side reaction. Kinetic studies showed that the reaction was zero order with respect to the pressure of oxygen, first order in ethylene at low pressures but zero order at high pressures, and progressively poisoned by the two minor products. The activation energy associated with the initial rate of reaction was 14 kcal/mole.
220 C. Kemball and W. R. Patterson Experimental The reaction was studied in a static system on evaporated films of palladium and a small proportion of the reacting gases was allowed to bleed off through a fine capillary leak from the reaction vessel to the ion source of a Metropolitan-Vickers MS 2 mass spectrometer. The essential features of the apparatus have been described previously (Kemball 1951, 1952) and it permitted analyses of the gas mixtures to be made every few minutes.
Catalytic oxidation of ethylene on palladium films 221 section of the mass spectrometer and not included directly in the reaction mixture. Peak heights of all masses multiplied by the appropriate factor, based on the decreased sensititivity of the mass spectrometer for argon, remained virtually constant over periods of several hours. The amount of carbon dioxide formed in the ion source when a mixture containing oxygen was admitted to the spectrometer rose to a constant value after about 40 min. Consequently, each gas mixture was maintained at 0 °C for such a period before placing the furnace around the vessel to start the reaction. This allowed sufficient time for the background contribution to the peak corresponding to mass 44 to reach a stable value which could be sub tracted from subsequent measurements of that peak.
Catalytic oxidation of ethylene on palladium films 223 the films showing that substantial sintering had occurred either on the admission of the gases or at any rate before reaction started. The absolute initial rate of reaction, expressed as molecules of ethylene reacting per second per square centi metre of apparent surface, was given by r0 = l022'3±°'5exp (- 14 300 ± 600 for the standard gas mixture.
224 C. Kemball and W. R. Patterson behaviour occurred to the same extent in the various experiments with different gas mixtures as in the normal runs, showing that the kinetic dependence of the rate of production of poison was identical with that of the main oxidation to carbon dioxide and water.
Catalytic oxidation of ethylene on palladium films 225 The addition of substantial amounts of carbon dioxide and water to the initial mixture had no influence on the initial rate of reaction, or the course of the reaction, showing that neither of the main products were acting as poisons. The initial rate was decreased at 77 °C by factors of 3 and 15 when 0*5 mm of acetic anhydride and acetic acid respectively were included. Further evidence that acetic acid was the stronger poison was obtained from the course of the latter reaction (figure 5) which, after a short initial decrease, subsequently followed a first-order equation in dicating that the poisoning normally associated with the progress of the reaction was negligible compared with the influence of the added acetic acid.
226 C. Kemball and W. R. Patterson It is known that acetaldehyde is a primary product of the oxidation of ethylene on silver and that it can be converted into acetic anhydride by liquid-phase oxidation in the presence of metal catalysts (Twigg 1954). Consequently, a 10: 1 oxygen: acetaldehyde mixture was reacted on a palladium film at 77 °C. Small amounts of carbon dioxide were formed initially, but the main reaction was the rapid production of acetic anhydride in almost quantitative yield, together with some acetic acid which increased slowly due to the hydrolysis of the anhydride. The sequence of reactions is obvious from the heights of the various peaks shown in figure 6. A similar result was obtained at 130 °C and only very slow oxidation of the products was observed. Other experiments confirmed that no oxidation of acetic anhydride and acetic acid took place at 77 °C and only very slowly at 130 °C on palladium.
Catalytic oxidation ofe thylene on palladium films 227 between the percentage x of ethylene converted to carbon dioxide at time t and the constant a includes a temperature-dependent factor related to the heat of adsorp tion of the poison. Satisfactory agreement with the experimental results was found but only with n = 3 and the derived values of a at a series of temperatures gave 17kcal/mole for the heat of adsorption of the poison. The indication that three sites are required before oxidation occurs appeared to be unusual and so we prefer to regard equation (1) as an empirical rather than a fundamental means of describing the course of the reactions.
228 C. Kemball and W. R. Patterson gases. The constants ax and a2 contain terms related to the heats of adsorption of the two poisoning species and A — allc1 and B = a2k2. As the integration of (4) is complex, it was rewritten as 100 —# 1 Ax .