Optimal Monte Carlo sampling for simulation of classical fluids

We have carried out Monte Carlo simulations to assess the efficiency in sampling the configuration space, using the conventional single-particle move and some novel multi-particle move schemes. We have studied model systems of one-component hard spheres and a system of highly asymmetric charged hard spheres. Using the mean square displacement as a measure of the extent of sampling of the configuration space, we find evidence that the optimal sampling corresponds to acceptance ratios or probabilities much smaller than the widely used value of 50%. In all the cases investigated, the multi-particle move schemes show no advantage over the single-particle move scheme. However, for some extreme conditions, like for example, higher charge asymmetries in the asymmetric electrolytic systems, multi-particle moves could indeed become advantageous as have been demonstrated recently.