CDS&E: Development and application of cluster-integral methods for dispersions and complex solutions

With this award, the the Chemical Theory, Models and Computational Methods Program in the Chemistry Division is supporting David Kofke and Andrew Schultz of the University at Buffalo/SUNY to develop a hierarchal theoretical and computational approach to studying chemical solutions. Chemical solutions are ubiquitous in natural and man-made systems, and their behavior is critical to many applications in technology, including healthcare, energy storage, environmental preservation, and food and consumer products. One approach to the description of chemical solutions is via a hierarchy of solute interactions, considering first the solute molecules individually, then taken in pairs, then groups of three, four, and so on. Each level in this hierarchy introduces new information about the mixture, such that one can infer the full behavior by proper analysis of these interactions. The quantities needed for this formulation can in principle be computed from detailed molecular considerations, which is desirable because it leads to greater predictive and descriptive capability. However, the means to perform these calculations for practical purposes have not been developed, so the full potential of the overall approach is relatively untested. Developing and benchmarking this approach is the focus of this research projects. The project integrates training of graduate students with outreach activities and development of learning materials and computer-based tools that permit others to benefit from the developments in this project. The osmotic virial equation forms the theoretical framework the work. The research objectives of are: (1) to develop robust and accurate methods for calculating osmotic virial coefficients up to at least 5th order, from a molecular model, while formulating approximants that extend their reach; (2) to examine behavior of mixtures of hard convex particles, focusing in particular on entropy-driven aggregation and fluid-fluid phase instability. High-throughput screening using these cluster-integral methods is anticipated to yield data for at least 5000 species pairs; (3) to investigate association of proteins in solution, considering in particular the effect of crowder molecules on this behavior.