Hybrid Monte Carlo method for off-lattice simulation of processes involving steps with widely varying rates

Abstract

A hybrid Monte Carlo simulation method consisting of kinetic and equilibrium moves is presented. The method is applicable to any system involving N different processes whose rates may vary by orders of magnitude. We show that the method easily permits the propagation of the system through macroscopic time scales by application to a study of chemical vapor deposition diamond-film growth. The application is an off-lattice simulation utilizing CH3 as the growth species on a diamond [111] surface, and incorporating a computationally expensive, many-body potential surface. We show that, by using our method, it is possible to simulate a continuous system of several thousand atoms, with no underlying grid and with a realistic potential for times on the order of milliseconds. The growth rate of the simulated surface is consistent with experimental growth rates, and the simulation sheds light on possible morphologies during the early stages of diamond film formation. © 1996 American Institute of Physics.