Ab initio dynamics in catalysis

Recycling waste carbon (e.g. CO₂, biomass) is a powerful strategy to address humanity’s longstanding challenges, including climate change and environmental degradation. Electrocatalysis is an efficient tool for converting waste carbon to value-added products and is an alternative approach for closing the carbon circle and achieving carbon neutrality. However, electrocatalysis faces many technical challenges, the most important of which is to have an atomic-scale understanding of the complex and highly dynamic electrochemical interfaces where chemical reactions take place. In this thesis, we applied ab initio methods to elucidate the water adsorption, solvation, thermal expansion, strain and dipole-field effect at the interfaces.
We begin by benchmarking water adsorption on metal surfaces with ab initio molecular dynamics (AIMD). The water adsorption energy, a central quantity to our goal of understanding water at interfaces and the solvation effect, is directly compared with temperature-programmed desorption (TPD) and other experimental results. Our results show that both RPBED3 and BEEF-vdW lead to appropriate water binding strengths, while PBED3 clearly overbinds near-surface water relative to experiments. Our study gives atomistic insight into the complex equilibrium of water adsorption and represents a guideline for future DFT-based simulations of the water/metal interface within molecular dynamics studies.
We then focus on the selectivity issue of glycerol electro-oxidation, which is an area of interest within biomass upcycling, because of the low overpotential and high feedstock availability. Inspired by our experimental collaborators, we study the limited selectivity toward lactic acid on Pt surface by DFT. We have formulated a theoretical descriptor that establishes a correlation between surface acidity and selectivity toward lactic acid. Furthermore, we investigate the solvation effect of glycerol and related intermediates via an AIMD approach, which is the key to understanding the selectivity to 2-electron products. A simple linear model was developed to estimate the solvation energy in glycerol oxidation.
Further, we explore the trends in strain effects from thermal expansion via ab initio phonon dynamics. We developed a thermal expansion strain parameter for transition metals which can quantitatively describe the strain effect induced by thermal expansion. The results offer a simple and easy method to correct adsorption energies at specific temperatures and assist in the strain engineering for catalysts.
Finally, we turn to studying CO₂ activation on the Cu surface. We develop a computational method to determine the energetic contribution and binding motifs of CO₂ adsorption at electrochemical interfaces. By splitting the energies into bending, interaction with surface, dipole field and solvation components, we find that the carbon and oxygen binding motif is the most feasible configuration. Employing ab initio electrochemistry methods, we also investigate the energy trend in CO₂ adsorption. These results help us understand the CO₂ activation and selectivity in CO₂ reduction.
In summary, we have used ab initio dynamics to explore the manifold aspects of catalysis. We investigate water adsorption, glycerol electro-oxidation, thermal expansion strain and CO₂ activation as case studies for the key factors at interfaces. The results and the methods we developed in this thesis will influence the theoretical development of electrochemical interfaces, industrial applications, and better catalyst design for upcycling of waste carbon.