Premium
A Monte Carlo model utilizing local chemical potentials for simulating segregation and diffusion. Part 2–implementation
Author(s) -
Joubert H. D.,
Swart H. C.,
Terblans J. J.
Publication year - 2005
Publication title -
surface and interface analysis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.52
H-Index - 90
eISSN - 1096-9918
pISSN - 0142-2421
DOI - 10.1002/sia.2118
Subject(s) - monte carlo method , kinetic monte carlo , statistical physics , diffusion , scaling , computation , dynamic monte carlo method , monte carlo molecular modeling , thermodynamics , materials science , physics , mathematics , algorithm , markov chain monte carlo , statistics , geometry
Abstract The Monte Carlo model described in Part 1 was implemented in a software package. Comparisons were made between the new model and the existing, modified Darken model. The first comparison focused on converting the jumps per atom to time. This was accomplished by fitting the Fick model on the kinetic part of the segregation profile. From the fit, a scaling parameter was found with which the Monte Carlo steps were converted to time. The scaled Monte Carlo segregation profile was compared to the modified Darken model, and a good correlation between the profiles was found. The influence of various segregation energies was also studied by computing segregation profiles for different segregation energies. The equilibrium concentrations obtained from the different segregation profiles follow the values predicted by the Langmuir–McLean equation. A parallel calculation scheme was successfully implemented that led to a decrease in the effective calculation time. The surface concentrations of the parallel computations were added together and an average segregation profile was obtained. This resulted in a reduction of the statistical noise on the segregation profile. Copyright © 2005 John Wiley & Sons, Ltd.