نبذة مختصرة : International audience ; Special Quasi-random Structures (SQSs) are often used to model disorderedalloys in small simulation cells. Yet, SQS-based sampling yields chemical po-tentials (and other thermodynamic properties) that do not match the equi-librium values, for instance measured in atomic Monte Carlo simulations.This is due to the lack of chemical short-range order in random samples.In this paper, we present a probabilistic analysis of chemical potential cal-culations based on the distribution of substitution energies and the Widomtechnique. Performing the analysis by sampling either equilibrium configu-rations or SQSs, we show that they both yield different results, but that itis possible to correct the results from the random sampling in order to get aresult which is much closer to the equilibrium values. The correction is verysimple to apply and does not require additional total energy calculations.
Processing Request
No Comments.