Multi-layer methods for quantum chemistry in the condensed phase: Combining density functional theory, molecular mechanics, and continuum solvation models.

摘要

We discuss the development and application of a number of theoretical physical models focused on improving our understanding of quantum chemical phenomena in condensed phase environments, especially aqueous solutions. The large number of atoms and molecules present in such systems precludes the application of the most advanced and accurate quantum chemistry theories available due to their exponential growth of required computational power with respect to the number of electrons in a system. As a feasible alternative, we opt to take a "multi-layer" approach, wherein the full chemical system is partitioned into different layers treated with varying levels of approximation, circumventing the exponential scaling computational cost. How this partitioning is performed and applied appropriately is the principal emphasis of this work.;Our main chemical system of interest is aqueous DNA and its excited electronic states. We examine applications of mixing quantum mechanics and classical molecular mechanics models, a multi-layer approach known as "QM/MM," to simulate the electronic absorption spectrum of aqueous uracil as computed with Time-Dependent Density Functional Theory (TDDFT). We encounter a major issue of spurious charge-transfer (CT) states in TDDFT even at small uracil--water clusters. Applying Long-Range Corrected TDDFT (LRC-TDDFT), however, alleviates this issue and allows us to investigate the absorption spectrum of aqueous DNA systems of up to as much as 8 nucleobases, providing some important clues to the initial dynamics of aqueous DNA excited by ultraviolet light and its possible ensuing damage. Then, to overcome certain computational limitations in modeling solvent by QM/MM alone, we turn to the methodology of polarizable continuum models (PCMs), which can be added on top of the QM/MM multi-layer approach as an "implicit" solvent model (in the sense that the average solvent charge density is approximated as a dielectric medium). We find that several extant PCM techniques are prone to numerical instabilities and discontinuous potential energy surfaces, and we propose ways to overcome such. Furthermore, we develop insights into the theory of PCM that yield an entirely new PCM for modeling the electrostatic effects of salty solutions. The culmination of our efforts is a cutting-edge QM/MM/PCM multi-layer approach for modeling quantum chemistry in the condensed phase.