Multiscale prediction of self-assembly and phase behavior in multicomponent surfactant mixtures

Surfactant self-assembly in soft matter formulations is a complex process that is subject to a multivariate design space. Improved methods for predictive computational modeling are attractive for efficiently sweeping the expansive design space to aid in formulation design. Modern molecular simulation techniques, such as all-atom molecular dynamics and coarse-grained methods, are limited in studying surfactant self-assembly either by time and length scale disparities or by predictive power. To overcome these limitations, we employ a novel multiscale methodology that utilizes small-scale all-atom molecular dynamics simulations to parameterize statistical field-theoretic models via bottom-up coarse-graining, forgoing the need for experimental input. The resulting molecularly informed field theory can then be sampled via mean-field calculations for rigorous determination of self-assembly behavior. In this work, we develop a molecularly informed field theory to demonstrate state-of-the-art first-principles predictions of the binary and ternary phase diagrams for cationic quaternary-ammonium surfactants in water or in water/oil mixtures over a range of chemistries, temperatures, and compositions. For the binary phase diagrams, we are able to predict the stability of all experimentally observed surfactant phases and a majority of the correct ordering of phase transitions completely de novo. Furthermore, we showcase how the methodology enables access to the additive design space by predicting salt-driven morphological transitions and aggregation numbers in self-assembly. For the ternary phase diagrams, we map out regions of homogenous phase, liquid crystal phases, and emulsion phases. Finally, we extend our methodology to demonstrate facile calculation of mesophase phase coexistence and macrophase separation in ternary blends of surfactant, water, and oil. We believe this multiscale methodology has the potential to be integrated into high-throughput screening workflows to inexpensively predict phase diagrams of novel surfactant formulations.