Baron Peters of the University of California, Santa Barbara, is supported by an award from the Chemical Theory, Models and Computational Methods program in the Chemistry Division for to develop simple and accurate theories to study chemical reactions in solution. Such reactions are central to many natural and industrial processes, including our own biochemistry. Current simulation approaches that treat the solvent as a polarizable continuum have severe limitations. Other, more sophisticated simulation approaches are able to predict rates but often do not give any insight into the underlying mechanisms of the reaction. The ability to predict rates without mechanistic understanding seems like an advantage, but these analyses do not help researchers graduate from large scale computation to simple, accurate, and understandable models. Simple theoretical models, such as those developed in this project; help scientists engineer processes involving complex reactions. Moreover, simple models can both predict rates and interpret rate data from experiments. The proposed work will have potential impact on chemistry, physics, biology, geology, materials science and engineering. Peters and his research group develop software, documentation, and YouTube tutorials to help others perform these analyses with LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator; Sandia National Laboratories), CHARMM (Chemistry at HARvard Macromolecular Mechanics), and other widely used software packages.
The simple but accurate theories that have been developed for gas phase reactions, surface reactions, and electron transfer reactions share one common characteristic: they are built around a single reaction coordinate which accurately summarizes the mechanism. In contrast, solvent reaction coordinates remain elusive. Peters and coworkers propose a new strategy for constructing the solvent coordinates using procedures that can be applied across broad families of reactions with similar electrostatic characteristics. The approach combines elements of continuum solvation approaches, variational transition state theory, energy gap coordinates from electron transfer theory, and a new polarization progress coordinate. The relative importance of these coordinates will be determined using transition path sampling and likelihood maximization approaches. Instead of directly generating rates, this computational analysis guides the construction of simple rate theories that can be used to predict trends across entire reaction families. These results are used to classify reactions in solution into mechanistic families and to develop practical rules-of-thumb for equilibrium and non-equilibrium solvation effects across the different reaction families.
