Additives are used to control nucleation in many natural and industrial environments. However, the mechanisms by which additives inhibit or accelerate solute precipitate nucleation are not well understood. We propose an equation that predicts changes in nucleation barriers based on the adsorption properties and concentrations of trace additives. The equation shows that nucleant efficacy depends on the product of an adsorption equilibrium constant and the reduction in interfacial tension. Moreover, the two factors that determine the potency of additives are related to each other, suggesting that assays of just one property might facilitate additive design. We test the design equation for a Potts lattice gas model with surfactant-like additives in addition to solutes and solvents.

Nucleation kinetics, induction times, and metastable zone widths are often modeled in quiescent, quasi-steady, and/or spatially uniform concentration fields. However, freezing an aqueous solution can concentrate the solute and effectively increase the supersaturation. During the freezing process, a boundary layer of giant supersaturation develops ahead of the moving ice front. We develop stochastic models of nucleation in the boundary layer when the growing ice perfectly excludes the solute for a one-dimensional system. The models make three simplifying assumptions: quasi-stationary nucleation kinetics, nuclei that are small compared to the boundary layer thickness, and a constant solvent crystallization growth velocity. Whether heterogeneous on the ice surface, or homogeneous in the boundary layer, the models suggest that nucleation is dramatically accelerated by the growing ice. For methane hydrates, which form at conditions similar to that of ice, induction times for hydrate nucleation can be reduced by as much as 10 105 times because of the moving supersaturation zone.