Partitioning Molecular And Surface Environments: Practice And Approximations
The physics of complex environments, though poorly understood, plays a critical role in a wide range of systems from the biological to the technological. Despite the importance of these systems, they are poorly understood. The large required simulation size of complex environments currently makes them unsuitable for direct ab initio calculations, yet inexpensive methods such as molecular modeling can be too approximate to describe the relevant chemical interactions. One resolution to this problem is to accurately describe specific interactions, and to inexpensively approximate interactions which are less relevant. The objective of this thesis is to determine to what extent one can partition systems into a molecule or surface treated at one level of theory, and a surrounding fluid or other molecular environment treated with another theory. I explore varying levels of theory, using different theories depending on the properties of the system and type of problem, and I address conflicts that arise from these choices when working to calculate properties of real systems. In the first project of the thesis, I strategically explore very different levels of theory for the solvent and solute to subvert the typical relationship found in computational work between cost and accuracy. In this thesis chapter, I combine a simple continuum description of a fluid with highly accurate quantum Monte Carlo calculations, to find solvation free energies of molecules. I find that this approach successfully preserves the accuracy of quantum Monte Carlo, while creating a framework for future calculations with more complicated fluid models. In the next two chapters, I investigate the continuum between expensive calculations of realistic systems, and inexpensive idealized systems. I examine the properties of lithium-sulfur battery systems and defects in pentacene thin films, using established polarizable continuum methods and density-functional theory, exploring the extent to which experimental realities alter the relevant observables. I am able to produce idealized voltammograms that capture key aspects of the battery experiments, and I also identify an experimentally observed defect in the pentacene. In the last chapter, I investigate the creation of a whole crystal from the sum of its parts, as I develop approximations of the dielectric matrix in molecular crystals. I successfully apply this to two crystals, creating dielectric band structures from molecular calculations that very closely resemble those from the entire crystal.
ab initio electronic structure; solvation
Marohn, John A.
Hennig, Richard G.; Hoffmann, Roald; Arias, Tomas A.
Chemistry and Chemical Biology
Ph.D. of Chemistry and Chemical Biology
Doctor of Philosophy
dissertation or thesis