TIME-DEPENDENT MONTE-CARLO STUDIES OF SURFACE-DIFFUSION

Surface adsorbate diffusion is examined using a numerical algorithm which incorporates a kinetic treatment in conjunction with a time-dependent Monte Carlo formalism. Nearest- and next-nearest-neighbor adparticle interactions are included. The method is based on a probabilistic description of adparticle jump events; and the diffusion rate is determined by the energetics of adparticle interactions on the lattice. In addition, the rare event problem associated with other theoretical treatments of diffusion is overcome by our highly efficient algorithm. Consequently, we are able to observe events, including ordering and island formation, which occur on time scales which are longer by orders of magnitude than those for simple adsorbate diffusion. Our initial investigations indicate that a variety of diffusion mechanisms may be operative depending on the adparticle interactions in the system. With nearest-neighbor interactions, our systems achieve a random walk limit at long times. We have also observed ordering and island formation, as well as a change in diffusion mechanism, as next-nearest-neighbor attractive energies are increased.