MONTE-CARLO RESULTS FOR BINARY MULTI-YUKAWA MIXTURES - EVALUATION OF THE ACCURACY OF THE MEAN SPHERICAL APPROXIMATION FOR REALISTIC HARD-CORE POTENTIALS

Monte Carlo (MC) simulations are used to determine the properties of two different binary multi-Yukawa mixtures studied previously through a quasianalytic Yukawa-MSA (mean spherical approximation) algorithm [Arrieta, Jedrzejek, and Marsh, J. Chem. Phys. 95, 6806 (1991)]. These mixtures are composed of spherical hard-core molecules with multi-Yukawa interactions fitted (beyond the core diameters sigma-ij(LJ) to Lennard-Jones potentials, including thus a significant repulsion (negative slope) interval. The characteristics of the first mixture were chosen to produce a nearly ideal solution, while those of the second mixture (large size difference between components, weak unlike-particle attractions) favored nonideal behavior. For a variety of compositions, densities (in the liquid range), and temperatures, the following properties are determined: configurational energy, pressure, and chemical potentials. The latter were obtained through a new implementation of Widom's particle insertion method. This simple implementation allowed the calculation of chemical potentials at high densities, where the usual procedures tend to fail. An analysis of the standard deviations and of the internal consistency of the MC data was used to confirm the general reliability of the simulation results. The good general agreement found between MC and MSA leads to the conclusion that the Yukawa-MSA quasianalytic algorithm provides not only a convenient but an accurate description of dense fluid mixtures, both ideal and nonideal.