Chemical potentials in ionic systems from Monte Carlo simulations with distance-biased test particle insertions

In this paper, we investigate sampling techniques for determination of chemical potentials of ionic systems based on distance-biased insertions of ion pairs. After a brief review of previous work in the area of calculation of chemical potentials and phase equilibria for ionic systems, we present conventional Monte Carlo calculations to illustrate that the standard test particle insertion or grand canonical algorithms become highly inefficient at the low temperatures for which phase coexistence occurs. A possible reason for the failure of conventional algorithms is their inability to adequately sample short interparticle distances that are dominant in the low density, low temperature region. We propose a distance-biased insertion method for determination of chemical potentials that preferentially samples short distances. Our biased procedure gives results in good agreement with standard insertions for the range of conditions for which standard insertions are valid, but does not improve the statistical uncertainty of the results or the range of temperatures and densities that can be studied.