Correlated random walks on two-sublattice systems. II. Monte Carlo simulations

Correlated random walks on two-sublattice systems, with a potential difference between the sublattices resulting in site-dependent hopping rates, are studied by Monte Carlo simulation techniques. Large effective samples, consisting of approximately a million particles each, are used. Equally important, the systems are monitored over long effective times, e.g., up to 1000 Monte Carlo steps per particle (MCS's/P) for the three-dimensional lattices of NaCl and CsCl types, and up to 2500 MCS's/P for their two-dimensional analogs. The data are analyzed to accurately estimate the labeled-particle diffusion rates. These high-precision Monte Carlo estimates are compared with the predictions of the theory described in the preceding paper. Qualitative agreement between the two sets of results obtains over a wide range of effective potential differences between the two sublattices. Quantitatively, however, the agreement is found to be reasonable only as long as the effective-potential differences, reflected through the intersublattice hopping rates ratio {σ}, are not too large, or, equivalently, as long as {σ} lies within the restricted range (Z-2)>{σ}>1/(Z-2). (Here Z is the coordination number for the relevant particle hops.)