The role of adsorbate-layer nonuniformities in catalytic reactor design: multiscale simulations for CO oxidation on Pt

Abstract

Monte Carlo (MC) simulations under ultra high vacuum (UHV) conditions and multiscale models at atmospheric pressure are compared to the corresponding continuum-based models in order to study the effect of species spatial inhomogeneities on reactor design. These models are applied to the catalytic oxidation of CO on Pt. Under UHV conditions, there are significant deviations in model responses between mean field (MF) and MC models in the absence of surface diffusion, due to adjacency requirements and lateral interactions. The disparity in time scales caused by surface diffusion is also addressed. Sensitivity and partial equilibrium analyses are introduced to delineate when surface diffusion should be included in MC or multiscale simulations. It is shown that including sufficiently fast Fickian surface diffusion of CO* in MC simulations does not always results in the MF limit, especially near extinction and for fuel lean mixtures. In the presence of repulsive interactions, it is shown that thermodynamically driven ordering of CO* affects the ignition temperature. Finally, it is found that in multiscale reactors, continuum-based models more accurately predict design performance, compared to UHV conditions, due to homogenization induced by desorption and readsorption of chemical species.