An Investigation of Computational Models for Surfactant Self-Assembly

Abstract

Surfactants are used in a wide variety of industrial and biological applications due to their self-assembling ability. Computational models can be used to understand the self-assembly process on an atomistic level. Creating a model that can quantitatively reproduce system properties has proven elusive. Since atomistic models can have difficulty reaching the necessary time and length scales for self-assembly, models have been investigated that are of different resolutions and predictive power. Molecular dynamics and Monte Carlo simulations were performed to determine the ability of these models to reproduce thermodynamic properties of surfactant self-assembly. A recent set of explicit solvent coarse grained (CG) chemistry-specific models are able to reproduce the equilibrium properties of critical micelle concentration (CMC) and aggregate size distribution. Non-ionic and zwitterionic surfactants were modeled using molecular dynamics. The CG models are able to achieve equilibrium in reduced computational time. Replica exchange can also provide a modest improvement in equilibration time. However, further model refinement is needed in order to match experimental values quantitatively and the models are unable to capture the temperature dependence of these properties due to the lack of solvent orientation.

Two additional studies with two different types of models were performed. Using Grand Canonical Monte Carlo and histogram reweighting, an implicit solvent model was developed with the ability to capture the temperature dependence of the CMC of ionic surfactants. The model can approximate the temperature dependence for two cationic surfactants without direct parameterization. The phase space of an atomistic model used to model sodium alkyl sulfates was also explored. The model is able to qualitatively account for the decrease in CMC with increasing salt concentration. The method of assessing the CMC of a model was improved by using an empirical equation which accounts for total surfactant concentration and counterion association. This method provided agreement in the CMC across multiple total surfactant concentrations. Also, the model is shown to qualitatively reproduce the effect of alkyl tail length and the minimum in the CMC with respect to temperature. Other atomistic models were tested and shown to provide higher CMC values, but all models underpredict the CMC. This indicates that there is a need for further development of atomistic models which are frequently used to parameterize coarser grained models. All of these studies can provide insight for developing a quantitative model for surfactant self-assembly.