Crystalline nanoporous solids such as zeolites are employed in a wide range of chemical processes due in part to their tunable micro-environments. Kinetics at internal zeolite acid sites can be modified by changing pore size, pore architecture, or polarity. Tunable environments impart shape-selectivity that preferentially stabilizes transition states but may also lead to mass transfer limitations or competitive adsorption leading to poor product selectivity. Density functional theory (DFT) describes the electronic states of reactive intermediates and transition states but cannot access the longer length scales necessary to quantify diffusion. Classical simulations can accurately simulate diffusion but require parameterized values that do not account for bond cleavage. Machine learning interatomic potentials have emerged as a technique to derive parameterized models from DFT data and our aim is to adapt these models to predict the entropy and diffusion of reactive intermediates in nanoporous catalysts.<br/>We used zeolite-catalyzed lactic acid dehydration to acrylic acid as a probe reaction to interrogate the relative importance of intrinsic kinetic barriers and mass transfer. We developed machine learning interatomic potentials to accelerate the quantification of hopping rate constants and adsorbate entropy under reaction conditions. We will also discuss how these techniques can be used to derive solvation models for liquid phase confined catalysts.

Leveraging machine learning interatomic potentials for coupled biomass reaction/diffusion in nanoporous catalysts

Where thousands of chemistry professionals meet to share ideas and advance scientific and technical knowledge. Spring 2023 abstracts, posters, and presentations will launch at the end of March.

ACS Spring 2023

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.