(190d) Advancing Kinetic Study of Catalytic Reaction: Hybrid Modeling Approach for Predicting Effective Activation Energy Barrier

Catalytic reactions play a pivotal role in numerous industrial applications and environmental remediation efforts, driving progress in sustainable energy development. Central to understanding these reactions is the determination of activation energy barriers, which govern the rates and pathways of diverse chemical reactions at the catalyst surface. Specifically, understanding the variability of these barriers can offer valuable insights into which reaction steps are favorable or unfavorable within numerous reaction mechanisms. Traditionally, activation energy barriers in catalytic reactions are calculated using density functional theory (DFT), a method critical for interpreting experimental results and identifying activation energy barriers. However, the static approach of DFT fails to capture the dynamic nature of catalytic reactions, leading to inaccuracies due to its reliance on assumptions and the lack of experimental validation. Hence, this study addresses these limitations in capturing the dynamic nature of catalytic reactions, particularly in determining effective activation energy barriers which cannot be validated with experimental results. Specifically, these barriers consider the dynamic and latent chemical mechanisms within catalytic systems, thus providing a more representative and precise understanding of catalytic processes. It is to be noted that as the effective activation energy barriers are derived by incorporating the underlying dynamics of the catalyst surface, they are essentially different from the values obtained from nominal DFT calculations done under static conditions. Recognizing the gap between theoretical predictions and experimental validation, we propose a novel hybrid modeling technique that integrates kinetic Monte Carlo (kMC) simulations into a neural network model. This framework uses real-time kinetic data to learn the values of effective activation energy barriers for different catalytic reactions. This approach not only aims to bridge theoretical knowledge with experimental outcomes but also enhances the adaptability and accuracy of kMC simulations by focusing on effective activation energy barriers.