Development and Application of Physically-Motivated, First-Principles Force Fields for Complex Chemical Systems

In this dissertation, we present a methodology for generating transferable force fields entirely based on quantum mechanical calculations, and show how these force fields can be applied to accurately compute the kinetic and thermodynamic properties of a wide range of complex chemical systems. Our approach is novel in that it provides a well-defined algorithm for generating physically-meaningful, transferable force field parameters that describe all types of two- and three-body intermolecular interactions necessary to accurately predict the properties of most (non-reactive) chemical systems. This methodology relies heavily on the framework of symmetry-adapted perturbation theory (SAPT, described in Chapter 1) to accurately compute the intermolecular interactions between molecules in terms of the constituent exchange repulsion, electrostatic, induction, and dispersion physical interactions. In order to obtain transferable force field parameters, monomer properties are used to describe all asymptotic intermolecular interactions at long range (Chapter 2). Our basic methodology for generating force field parameters is outlined in Chapter 3, and the accuracy and transferability of this approach is validated by computing second-virial coefficients for a wide range of small, organic molecules. In Chapter 4, we show that three-body dispersion and exchange interactions make non-trivial contributions to the pressure-dependent properties of bulk organic liquids, and by incorporating these interactions in a systematic and transferable fashion, we are able to accurately predict the bulk properties of a variety of neat-organic liquids entirely from first-principles. In Chapters 5-8, we employ our force fields to study metal organic frameworks, which are a class of complex nano-porous materials that have recently received a great deal of interest for their potential use in gas adsorption and separation problems, with great industrial relevance. In Chapter 9, we demonstrate that our force fields can be used to accurately model ionic liquids, systems which have recently been highly studied due to their important industrial applications as solvents and electrolytes. Finally, we discuss potential future improvements and applications of our force field developmental methodology.