Sustainable energy and chemical production largely hinges on the development of catalysts that efficiently transfer energy into molecules for storage (e.g. H2O oxidation or CO2 or N2 reduction into feedstoks for fuels or fertilizers). Doing so would open doors for renewable technologies to produce useful fuels and chemicals from O2, H2O, CO2, and N2.
A key hurdle in all of these processes is understanding how to drive multi-proton and electron transfers in a way so that products are made efficiently (i.e. with low overpotentials and high faradaic efficiencies). Having a quantum-level understanding of when and why good catalysts function and to what extent they can be improved would give guidance for rationally designing more sustainable processes.
We approach this problem using a variety of computational chemistry tools to enable predictions of potential energy surfaces to better understand reaction pathways and ensembles of mechanisms. Our group carries out computational investigations to determine the underlying thermodynamic energies and kinetics of reaction steps under their actual operating conditions. With this information we bridge a continuous understanding of homogeneous, heterogeneous, and biomimetic reaction mechanisms to aid the design of next-generation catalysts.
We computationally explore materials, dopants, and (electro)chemical processes on surfaces using quantum chemistry and alchemy.
We model and predict thermodynamics and kinetics of solvated processes using continuum, cluster-continuum, and explicit solvent models.
Our group uses quantum chemistry-based multiscale modeling to predict and study the atomic scale of materials and chemical reactions. With the combination of electronic structure and atomistic models as well as kinetic modeling, we can investigate fundamental reaction steps at different time- and length-scales that would otherwise be difficult or impossible to investigate with experiment.
Our modeling studies are entirely carried out in silico (on a computer) using quantum chemistry data. This makes our modeling predictions largely free from artificial biases that are present when using experimental inputs that may have unexpected uncertainties. Whether making computational predictions alone or in direct collaboration with experimentalists, our group provides deep perspective on the atomic-scale nature of chemical bonding in molecules and materials.
Our 'ground-up' multiscale modeling approach uses appropriate levels of quantum chemistry (QC) theory (typically on up to ca. 200 atoms) to model reaction energies, barrier heights, pKas, and standard redox potentials. Using data obtained from QC theory, we can also develop of analytic reactive force fields or machine learned methods, which are capable of modeling reaction dynamics on systems on the order of 100,000 atoms.