Monte Carlo simulations of amphiphilic systems.

Results are presented from NVT Monte Carlo simulations of a three-dimensional lattice model of a binary mixture of solvent and surfactant chains in which free self assembly is allowed. It is demonstrated that the model exhibits a critical micelle concentration together with cluster size distributions consistent with experiment and theory (minimum and maximum in the distribution within the micellar region). The weight average aggregation number, N[w], increases linearly with the square root of the concentration of micellised surfactant as predicted theoretically. The dilute solution excess chemical potential (beta[0n] --- beta[01]) is determined from the cluster size distribution. It is found to be a monotonically decreasing function of n with different functional forms for small and large clusters. A single analytical expression is found to describe the cluster size distribution and the X[1] versus X[a] curve on the concentration range from 0 to 5 vol. % . It is necessary to introduce an activity coefficient to accurately describe the behaviour of the model for amphiphile concentrations greater than 5 vol. % . The dependence of this expression on temperature and molecular interaction parameters is determined. Results are also presented for the simulation of longer chains. Following investigation of three of the models free parameters, regions of phase space in which 'spongy' structures and vesicles --- either spherical or tubular --- are successfully identified. A preliminary phase diagram is established by considering the variation in the cavity size distribution function. These results are discussed in relation with experimental data and existing phenomenological studies. An extension of the Configurational-bias Monte Carlo (CBMC) based on a self-avoiding walk using sites selected from subsets of sites known a priori available for the regrowth is also established. Its theoretical fondation is laid out together with a partial assessment of its efficiency relative to both the classic reptation and CBMC for the simulation of chains lying on a lattice of high coordination number. A methodology for the simulation of polyoxyethylene oxide (POE) surfactants using a cubic lattice of coordination number c = 26 is briefly discussed.