Research

Modeling Chemical Reactions in Condensed Phases

Studying chemical reactions in condensed phases—particularly liquids—is vital across many scientific domains. However, these complex processes can sometimes be difficult to characterize experimentally, highlighting the need for robust and predictive theoretical frameworks capable of capturing chemical reactivity in condensed phases on nanosecond timescales. Achieving computationally efficient yet predictive simulations necessitates accurate modeling of intermolecular interactions at the gas-phase clusters, which can subsequently be translated into reliable condensed-phase descriptions. To address this challenge, our research combines high-level quantum chemical methods with advanced machine-learning architectures. This synergistic approach captures subtle intermolecular interactions across phases and enables efficient, predictive simulations of condensed-phase processes. By coupling these machine-learned potentials with enhanced sampling techniques, we systematically explore reaction pathways and free-energy landscapes, enabling realistic and accurate modeling of chemical reactions in condensed environments under diverse conditions.

Computational Modeling of Chemical Reactivity in Materials

Chemical reactions at material interfaces and heterogeneous surfaces play a vital role in advancing industrial processes and addressing global challenges in energy, environmental sustainability, and chemical manufacturing. Accurate computational modeling of these reactions not only reveals performance metrics of existing materials but also informs the rational design of next-generation catalysts with optimized efficiencies and tailored functionalities. Although quantum-mechanical methods are essential for capturing the bond rearrangements that occur during catalytic reactions, a full quantum-chemical treatment of extended materials systems remains computationally prohibitive. Our research focuses on developing methodologies that partition the system into a reactive center and an extended environment, allowing each to be described using computational approaches of appropriate accuracy and efficiency. When combined with proper sampling techniques, this strategy enables a detailed investigation of catalytic reactivity in materials—offering mechanistic insights into intermolecular interactions, reaction thermodynamics, and structure–property relationships.