|Ph.D Student||Michael Monine|
|Subject||Pattern Formation on Catalytic Surfaces|
|Department||Department of Chemical Engineering||Supervisor||Professor Emeritus Pismen Leonid|
This thesis studies the interplay between processes occurring on a macroscale and nanoscale (atomic level, nm) in surface catalysis. The influence of an anisotropic atomic structure of catalytic surfaces on macroscale spatio-temporal phenomena is explored. The presented study is a theoretical work based on the development of mathematical models and computations using hybrid algorithms. The computation results are compared directly with experimental data. The first part of the present work considers reaction-diffusion fronts in a hydrogen oxidation reaction on Rh(111) surface. Experimentally, these fronts have been observed to be either isotropic or triangular depending on the conditions of preparation of the oxygen layer as well as on temperature and hydrogen pressure. It is proposed that adsorbed oxygen can penetrate into a subsurface layer of Rh bulk under high oxygen-exposure doses. The model is composed of two mean-field equations for oxygen and hydrogen coverages coupled to an equation for the angle-dependent speed of the oxygen phase transition front. The second part of this thesis is an extensive theoretical study of a carbon monoxide oxidation reaction on Pt(110) catalyst. A qualitative Monte Carlo model has been developed in order to describe dynamics of Pt(110) surface reconstruction under reactive conditions on atomic scale. Kinetics of a phenomenological time-dependent Ginzburg-Landau type is used. The developed model qualitatively reproduces such experimental results as a surface phase transition, surface roughening developing under the reaction conditions and changes in catalytic activity of the surface. Moreover, the model allows us to elucidate the nature of macroscale front propagation from a nanoscopic point of view. A one-dimensional extension of the phenomenological model reproduces macroscopic traveling waves on the CO diffusion scale. In a more realistic study of this reaction, a kinetic Monte Carlo (KMC) model has been proposed. Activation energies of the allowed atomic steps are estimated using available computational and experimental data. The KMC model correctly reproduces the Pt(110) surface reconstruction induced by CO adsorption and predicts the region in space of experimental parameters where faceting occurs. Under kinetic instability conditions, the simulated faceted pattern forms a periodic hill-and-valley structure with a periodicity comparable with experimental data. The simulations reproduce the development of faceting on a realistic time scale.