Computer simulations of epitaxial growth

Publisher Website
Google Scholar
Abstract
Two popular computer simulation techniques for epitaxial crystal growth, based on the well-known Monte Carlo (MC) and molecular-dynamics (MD) methods, are examined by studying the same physical problem in a comparison study. The implementation of these methods for epitaxy is discussed. Our results display the sensitivity of the predictions from the calculations to the parameters in the methodology. An idealized concept of a growth pattern in the infinitely slow reversible limit is employed to interpret the present numerical simulation results. From the limiting growth patterns, it is found that the dynamical MD simulations lead to fuller growth in the epitaxial layers than the static treatment implicit in MC. Since multiparticle moves in MC also displayed partial growth, the degree of growth can be rationalized in terms of the kinetic role of diffusion being better described by the correlated dynamical treatment of atomic motions in MD.