Structural and Catalytic Properties of Model Supported Nickel Catalysts

Abstract

The surface structure and catalytic properties of model silica-supported nickel catalysts have been investigated with infrared reflection-absorption spectroscopy (IRAS) and reaction kinetics of ethane hydrogenolysis and carbon monoxide methanation. Nickel particles in the range 15-80 %L were vacuum deposited onto silica thin films using established preparation procedures. Specific rates and activation energies for ethane hydrogenolysis and carbon monoxide methanation over the model catalysts are remarkbly similar to the corresponding values for both traditional silica-supported nickel catalysts and nickel single crystals. The turnover frequency for CO methanation is independent of particle size for the dimensions studied. The ethane hydrolysis rate, however, increases with particle size to a maximum at -25 A and thendecreases. This reactivity trend is found to parallel a change in the percentage of bridging CO on the nickel particles as determined by IRAS. An understanding of the relationship among surface structure, particle size, and chemical properties is of primary importance in the design of a cata1yst.l Typically, oriented single crystals have been used as model catalysts to investigate the influence of morphology on reactivity/selectivity. These surface science studies, under well-defined conditions, have provided insights into the catalytic mechanisms, active sites, and reacting species of heterogeneous catalysts at the molecular level.2 There are, however, distinct differences between the well-defined single crystals used in surface science studies and metal catalysts supported on high surface area oxides. It has been shown that the interaction between a dispersed modelcatalyst and its support can modify the activity of the catalyst depending on the reaction and the reaction condition^.^ These metal-support interactions have been investigated by depositing an oxide onto the surface of a single-crystal metal surface and monitoring the modifications to the reactivity of the metal? There are inherent problems with this approach to studying supported metal catalysts in that the metal maintains its bulklike characteristics. This is contrary to the effects observed for dispersed metal catalysts, where activity and selectivity often depend upon the size of the metal particle.5 A more favorable approach to the synthesis of a realistic model of a dispersed supported metal catalyst is to deposit a metal onto an oxide support material. The incorporation of surface science techniques in the study of these model supported metal catalysts has been limited due to the experimental difficulties related to sample mounting, sample heating/cooling, and surface charging of insulating oxides. Recently, we have circumvented these problems with an approach utilizing metal deposition onto a silicon dioxide thin film ( N 100 A) which, in turn, is supported on a molybdenum metal substrate.Gs These planar Sios thin film models have been shown to be good models for high surface area ~ilica.~ The Mo(ll0) substrate does not alter the chemical properties of the thin film, yet improves the thermal stability of the Si02 relative to Si02 grown on silic~n.~J~ The specific activity of supported nickel catalysts toward alkane hydrogenolysis generally depends markedly upon the metal particle size.sJ1 Specific activity is defined as the catalytic rate normalized to the number of exposed metal atoms and is therefore representative of the molecules produced per surface atom. Although the relationship between particle size and activity has been well documented, there is not a generally accepted expla- Abstract published in Aduonce ACS Abstracts, January 1, 1994.