Predicting Surfactant Phase Behavior from Molecularly Informed Field Theories

Surfactants are amphiphilic molecules that play an important role in numerous industrial applications ranging from detergents to oil-recovery to synthesis of advanced materials. In aqueous solution, surfactants can self-assemble into a variety of supramolecular arrays, such as spherical micelles or regular packings of liquid crystal-like phases. The morphologies of these self-assembled mesostructures are major determinants for material functionality and are controllable via a multivariate design space (temperature, composition, chemistry, and solution conditions, etc.). To better understand the effect of different parameters on the resulting self-assembled morphology, numerous phase diagrams have been experimentally constructed for different surfactant chemistries. However, these phase diagrams can be extremely complex, requiring intensive labor and large amounts of material in order to construct precisely. Thus, to accelerate design of surfactant based materials, improved methods for predictive computational modeling are attractive for efficiently sweeping the expansive design space. Modern computational chemistry techniques, such as all-atom molecular dynamics and coarse-grained methods, are limited either by time and length scale disparities or predictive power in studying surfactant mesophases. To overcome these limitations, we propose to employ a novel multiscale methodology that uses small-scale atomistic simulations to parameterize statistical field theory models via bottom-up relative entropy coarse-graining. The resulting molecularly-informed field theory can then be simulated with the mean-field approximation to efficiently determine the stable surfactant mesophases. In this project, we will demonstrate how we apply this multiscale methodology to predict surfactant solution phase behavior for several, well-studied surfactants de novo and obtain qualitative agreement with experimentally observed phase behavior.