TIME-DEPENDENT MONTE-CARLO STUDIES OF DIFFUSION WITH SURFACE TRAPS

Surface adsorbate diffusion is examined using a numerical algorithm which incorporates a kinetic treatment in conjunction with a time-dependent Monte Carlo formalism. The method is based upon a probabilistic description of adparticle jump events. The diffusion rate is determined by the energetics of interactions among adparticles, the surface substrate atoms, and surface defects. The rare event problem associated with other theoretical treatments of diffusion is overcome by our highly efficient algorithms. Consequently, we are able to observe events, including adsorbate overlayer ordering and island formation, which occur over time scales which are longer by orders of magnitude than those for simple adsorbate diffusion. Our initial investigations of perfect surfaces indicated that a variety of diffusion mechanisms were operative depending on the adparticle interactions and other controlling parameters in the system. We have now examined the effects of surface traps on adsorbate diffusion dynamics. We find that, at low attractive adparticle interactions, the surface traps become titrated after which diffusion proceeds as it might otherwise. At higher adparticle interactions, diffusion is found to be strongly affected by the surface traps, even beyond the range of direct influence of the trap sites. In fact, we find that traps may act as nucleation sites for both cluster formation and long range island formation. We conclude that no comprehensive analysis of surface diffusion can fail to address the effects of imperfections such as these on the diffusion rate and mechanism.