(2hp) Multi-Scale DFT/MD Computational Approaches to Condensed Phase and Electrocatalytic Reactions.

During my Ph.D. training, I have developed and implemented a multi-scale model that combines density functional theory (DFT) with classical molecular dynamics (MD) for modeling reactions in complex condensed phases. This method development was motivated by the general struggle of state-of-the-art computational tools to balance computational cost and accuracy. For example, ab initio molecular dynamics accurately simulate atomistic interactions but fail to adequately sample slow dynamics, when implicit solvation models offer reduced computational cost but lack description of solvent structuring.

As a chemical engineer, I applied this newly developed multi-scale DFT/MD approach to study the thermodynamics and kinetics of industrially relevant condensed-phase reaction systems. The ultimate goal is not only to provide a simple computational tool to make predictions and assist in exploring the vast design space of catalysts and solvents, but also to advance fundamental understanding of the physical and chemical phenomena within these reaction systems at the atomistic level. For example, a key subject of study for the DFT/MD model is electrocatalysis, which has numerous applications in electrical capacitors, CO₂ conversion, and fuel cells. I aim to predict, from the DFT/MD model, the thermodynamics and kinetics of various electrochemical charge transfer processes at different metal electrodes, solvent compositions, and supporting ions, which altogether can guide in rationally designing and optimizing a particular electrocatalytic system. At the same time, this model allows me to probe for fundamental properties of the electrode/electrolyte interface such as the double layer polarizability and capacitance. Assessing these simulation-measured interfacial properties require knowledge of both physical thermodynamic theories (classical Gouy-Chapman-Stern theory and its modern rendition) and electrochemical experiments (cyclic voltammetry or impedance spectroscopy). Finally, these fundamental properties of the electrochemical interface can enable theory-supported exploration of the electrocatalytic design space.

In this poster presentation, I will walk through the development of the DFT/MD model in 3 reaction systems that are progressively more complex. First, the thermodynamics of hydrogen bonding in an aqueous acidic solution is explored, where the DFT/MD approach successfully modelled the relative thermodynamic stability between the two well-known H₃O⁺ and H₅O₂⁺ motifs. Second, I apply the DFT/MD model to study the kinetics of acid-catalyzed alcohol dehydration reactions in mixed solvent systems. I calculate the activation barriers of 2 dehydration reactions (t-butanol and fructose) in 3 different water/organic co-solvents (water with DMSO, acetonitrile or γ-GVL), and show excellent agreement with barrier values inferred from experimentally measured rate constants. Finally, I use DFT/MD to tackle the complex electrical double layer (EDL) at a metal electrode and aqueous electrolyte interface and its effect on charge transfer processes. In this study, extensive method development for the classical MD model is done to measure key physical properties of the EDL such as the dielectric polarizability and the capacitance. Through assessing these properties with classical double-layer theories and experimental impedance measurements, we are informed about both the validity and limitation of a pure classical model of the EDL. The classical EDL model is then combined with DFT description of ionic charge-transfer adsorption to gain better atomistic understanding for the complex effects of EDL on electrochemical processes. The multi-scale DFT/MD model has potential to be further applied to even more complex condensed-phase systems such as electrochemistry with polyelectrolytes or biological reactions with enzymes.

Following the completion of my Ph.D. this academic year, I am eager to find a post-doctoral position that allows me to continue developing my electronic structure and atomistic modeling capabilities. As I move closer to an independent academic career, I’m also eager to explore further modeling opportunities in catalysis subfields such as zeolites, metal-organic-framework, enzymes, or polymer catalysts.