Solvation effects have a decisive influence on mechanisms and rates of chemical reactions in solution. Therefore, their accurate modeling is of central importance in computational quantum mechanical investigations of chemical reactions, especially for catalytic reactions in solution like hydrogen production, CO2 conversion, biomass conversion, electrocatalysis, and processes in fuel cells.
The goal of the project is to develop and apply a novel accurate, yet efficient computational model of solvent effects which reflects the atomistic structure of the solvent. This challenging problem will be tackled with an interdisciplinary approach, combining expertise from quantum chemistry, computational materials science, and computer science of the project partners TU München (Scientific Computing, Catalysis Research Center) and Northwestern University (Chemistry Department). A hybrid approach will be implemented that combines a quantum-mechanical description of the solute with a direct statistical microscopic description of a molecular solvent. The structure of the solvent will be determined via the Born-Green-Yvon (BGY) hierarchy of integral equations for solvent (pair) distribution densities. Structural properties of the solute, essential for modeling reactions in solution in the ground and excited states, will be determined in this way. The high-performance computing demands of the method will be tackled by efficient numerical and scalable parallel algorithms.
This new model of solvation will supersede state-of-the-art continuum solvation models and will be more cost effective than models that rely on dynamic simulations with quantum mechanical forces. The capabilities of this novel computational strategy will be showcased for catalytic processes at metal nanoparticles in solution, including photocatalytic production of hydrogen from water and electrochemistry, as well as on actinide complexation chemistry in solutions of finite ionic strength.