Fluids of strongly interacting dipoles: Monte Carlo sampling using Tsallis statistics

Monte Carlo simulations of dipolar hard-sphere fluids in a generalized ensemble based on the physical realisation of Tsallis probabilities are reported. The new algorithm leads to a significantly better sampling than the standard Monte Carlo method. At very low densities, when the dipolar system is characterized by the presence of polymeric chains of head-to-tail aligned dipoles, simulations with q>1 may exhibit a better convergence to equilibrium. The improved sampling can be explained by the use in the generalized ensemble of an effective dipolar potential with a q-dependent anisotropy.