0000000000746452

AUTHOR

S. C. Ying

Dynamical mean field theory: an efficient method to study surface diffusion coefficients

Abstract We test the accuracy of the dynamical mean field theory (DMF) developed recently for the collective and tracer diffusion coefficients D C and D T , respectively, by Monte Carlo simulations of two very strongly interacting model systems. The deviation of the DMF results from the true hydrodynamic diffusion coefficients is a measure of memory effects, which are not fully accounted for in DMF. In the cases studied here, DMF predicts the behavior of both D C and D T accurately, while the memory effects are found to be most pronounced at low temperatures, and at high coverages and stronger interactions. Nevertheless, the computational cost of DMF is just a fraction of what is needed for…

research product

Non-equilibrium surface diffusion in the O/W(110) system

In this Letter, we present results of an extensive Monte Carlo study of the O/W(110) system under non-equilibrium conditions. We study the mean square displacements and long wavelength density fluctuations of adatoms. From these quantities, we define effective and time-dependent values for the collective and tracer diffusion mobilities. These mobilities reduce to the usual diffusion constants when equilibrium is reached. We discuss our results in view of existing experimental measurements of effective diffusion barriers, and the difficulties associated with interpreting non-equilibrium data.

research product

Comment on “Finite-size scaling behavior of the tracer surface diffusion coefficient near a second-order phase transition” by F. Nieto et al.

research product

Comment on “Surface diffusion near the points corresponding to continuous phase transitions” [J. Chem. Phys. 109, 3197 (1998)]

It is well known that unlike static equilibrium properties, kinetic quantities in Monte Carlo simulations are very sensitive to the details of the algorithm used for the microscopic transition rates. This is particularly true near the critical region where fluctuations are pronounced. We demonstrate that when diffusion of oxygen adatoms near the order–disorder transition of a lattice-gas model of the O/W(110) model system is studied, the transition rates must be chosen carefully. In particular, we show that the choice by Uebing and Zhdanov [J. Chem. Phys. 109, 3197 (1998)] is inappropriate for the study of critical effects in diffusion.

research product