The rate of reaction for oxidation of CO over (210) and (111) single-crystal surfaces of platinum has been studied as a function of reactant pressures (PO2,PCO) and sample temperature (T), both experimentally and by computer simulation. Experimental results on both surfaces show regions with a steady high rate of reaction followed by a nonsteady transition region and, at high CO pressures, a region with low reactivity caused by CO poisoning of the surface. At constant sample temperature, the transition region can be narrow and depends critically on the ratio of the gas phase concentration of reactants (PCO/PO2). The temperature dependences of the experimental data indicate that the critical ratio and the details for the occurrence of CO poisoning are strongly affected by surface processes such as adsorption, desorption, and diffusion ordering and reconstruction phenomena. A computer simulation model of the Langmuir–Hinshelwood surface reaction as developed by Ziff et al. was used for the simulation of the reaction under flow conditions. The initial fair agreement between this model and the experiment can be significantly improved if processes such as adsorption, desorption, and diffusion are taken into account in an extended simulation model which in turn provides an insight into the kinetics of adsorbate poisoning and the effect of adsorbate-induced processes on the reaction.

Influence of extended interactions on the surface dynamics in the catalytic CO oxidation on Pt single crystal substrateThe catalytic oxidation of CO by oxygen on a platinum ͑111͒ single-crystal surface in a gas-flow reactor follows the Langmuir-Hinshelwood reaction mechanism. It exhibits two macroscopic stable steady states ͑low reactivity: CO-covered surface; high reactivity: O-covered surface͒, as determined by mass spectrometry. Unlike other Pt and Pd surface orientations no temporal and spatiotemporal oscillations are formed. Accordingly, COϩO/Pt͑111͒ can be considered as one of the least complicated heterogeneous reaction systems. We measured both the macroscopic and mesoscopic reaction behavior by mass spectrometry and photoelectron emission microscopy ͑PEEM͒, respectively, and explored especially the region of the phase transition between low and high reactivity. We followed the rate-dependent width of an observed hysteresis in the reactivity and the kinetics of nucleation and growth of individual oxygen and CO islands using the PEEM technique. We were able to adjust conditions of the external control parameters which totally inhibited the motion of the reaction/diffusion front. By systematic variation of these conditions we could pinpoint a whole region of external control parameters in which the reaction/diffusion front does not move. Parallel model calculations suggest that the front is actually pinned by surface defects. In summary, our experiments and simulation reveal the existence of an ''experimental'' bistable region inside the ''computed'' bistable region of the reactivity diagram ͑S-shaped curve͒ leading to a novel dollar ͑$͒-shaped curve.

The steady and oscillatory regions of the CO oxidation reaction on the Pd(110) surface have been determined as a function of externally controlled parameters (flow rate, CO and oxygen partial pressures, temperature) over a wide range. At constant sample temperature and flow rate, the experiments yield a characteristic cross-shaped phase diagram separating regions of monostability, bistability, and oscillatory behavior. The existence of a cross-shaped phase diagram indicates the operation of a slow feedback process, which could be traced back to the (experimentally verified) formation and retarded removal of subsurface oxygen during the reaction. The diagram reflects one of the first well-defined oscillatory systems in heterogeneous catalysis and may provide a general basis for mechanistic studies and models of oscillatory surface reactions.

No abstract

No abstract