Ideal surface reactions catalytic mechanisms

Thus, the mechanism of catalytic processes near and far from the equilibrium of the reaction can differ. In general, linear models are valid only within a narrow range of (boundary) conditions near equilibrium. The rate constants, as functions of the concentration of the reactants and temperature, found near the equilibrium may be unsuitable for the description of the reaction far from equilibrium. The coverage of adsorbed species substantially affects the properties of a catalytic surface. The multiplicity of steady states, their stability, the ordering of adsorbed species, and catalyst surface reconstruction under the influence of adsorbed species also depend on the surface coverage. Non-linear phenomena at the atomic-molecular level strongly affect the rate and selectivity of a heterogeneous catalytic reaction. For the two-step sequence (eq.7.87) when step 1 is considered to be reversible and step 2 is in quasi-equilibria, it can be demonstrated for ideal surfaces that... 

Partial oxidations over complex mixed metal oxides are far from ideal for singlecrystal like studies of catalyst structure and reaction mechanisms, although several detailed (and by no means unreasonable) catalytic cycles have been postulated. Successful catalysts are believed to have surfaces that react selectively vith adsorbed organic reactants at positions where oxygen of only limited reactivity is present. This results in the desired partially oxidized products and a reduced catalytic site, exposing oxygen deficiencies. Such sites are reoxidized by oxygen from the bulk that is supplied by gas-phase O2 activated at remote sites. 

The presence of solution at a metal surface, as has been discussed, can significantly influence the pathways and energetics of a variety of catalytic reactions, especially electrocatalytic reactions that have the additional complexity of electrode potential. We describe here how the presence of a solution and an electrochemical potential influence the reaction pathways and the reaction mechanism for methanol dehydrogenation over ideal single-crystal surfaces. 

Another important difference in the poison formation reaction is observed when studying this reaction on Pt(lll) electrodes covered with different adatoms. On Pt(lll) electrodes covered with bismuth, the formation of CO ceased at relatively high coverages only when isolated Pt sites were found on the surface [Herrero et al., 1993]. For formic acid, the formation takes place only at defects thus, small bismuth coverages are able to stop poison formation [Herrero et al., 1993 Macia et al., 1999]. Thus, an ideal Pt(lll) electrode would form CO from methanol but not from formic acid. This important difference indicates that the mechanism proposed in (6.17) is not vahd. It should be noted that the most difhcult step in the oxidation mechanism of methanol is probably the addition of the oxygen atom required to yield CO2. In the case of formic acid, this step is not necessary, since the molecule has already two oxygen atoms. For that reason, the adatoms that enhance formic acid oxidation, such as bismuth or palladium, do not show any catalytic effect for methanol oxidation. 

Thus if the multiplicity of steady states for the catalyst surface manifesting itself in the multiplicity of steady-state catalytic reaction rates has been found experimentally and for its interpretation a three-step adsorption mechanism of type (4) and a hypothesis about the ideal adsorbed layer are used, the number of concrete admissible models is limited (there are four). It can be claimed that some types of adsorption mechanism have "feedbacks , but for the appearance of the multiplicity of steady states these "feedbacks must possess sufficient "strength . The analysis of these cases (mechanisms 4-7 in Table 2) shows that, to achieve multiplicity, the reaction conditions must "help the non-linear step. 

On the other hand, it is clear that the "ideal models cannot describe the behaviour of complex catalytic reactions in complete detail. In particular, we cannot quantitatively explain the values of the self-oscillation periods obtained by Orlik et al. Secondly, for example, we have failed to describe the critical effects obtained by Barelko et al. in terms of model (2)—(3) corresponding to the two-route mechanism with the parameters taken from ref. 49 or ref. 142. Our calculated reaction rates proved to be at least two orders of magnitude higher than the experimental values. Apparently our models must be considerably modified, primarily in the region of normal pressures. It is necessary to take into account the formation of unreactive oxygen forms that considerably decrease the rate of C02 generation, the dependence of the reaction parameters on the surface composition and catalyst volume and finally the diffusion of oxygen into the catalyst. 

---