The behavior of model nonionic amphiphilic molecules undergoing aggregation is studied using singlechain mean-field theory. The amphiphilic molecules are of the type HxTy where the H (head) monomers like the solvent molecules and the T (tail) monomers are solvophobic. In combination with the mass action model, the theory was used to study the critical micellar concentration, cmc, and the micellar size distribution as a function of head and tail lengths, architecture of the molecule, and temperature. The predictions of the theory are compared with the molecular dynamic results of Smit et al. (Langmuir 1993, 9, 9). Very good agreement is found between the theoretical predictions and the simulations. The theory is shown to predict quantitatively the two different free energy scales responsible for micellization and for micellar size distributions in model systems. The cmc is found to be only slightly dependent on the headgroup molecular architecture. However, the micellar size distribution is found to be quite different when comparing linear and branched headgroups. The structure of the micelle is found, in agreement with earlier theoretical predictions, to have a compact, almost solvent free, hydrophobic core and a wide interface region that includes the headgroups. The hydrophobic region is found to be more compact and larger for longer hydrophobic tails. The hydrophobic region is found to be more compact as the temperature decreases. The molecular organization in the micelles is not very sensitive to changes in the architecture of the headgroups.

Theoretical expressions are presented for the solvent configuration averaged force on a diatomic solute throughout the vapor–liquid density range. Analytical low density expansions and solvent configurational space averages are used to predict solvent induced changes in solute vibrational frequency. Purely classical Monte Carlo simulation results for a system representing bromine (Br2) dissolved in argon agree quantitatively with previous coupled quantum-classical results of Herman and Berne, up to liquid densities. It is found to be impossible to obtain a red gas to liquid shift (such as that typically observed experimentally) in any realistic diatomic system with only binary solvent atom–solute atom interaction potentials. However, redshifts are predicted when a three-atom potential, in which the solute–solvent interaction depends on solute bond length, is introduced.