(582h) Site Geometry Driven Selective Hydrogenation on Intermetallic Catalysts
AIChE Annual Meeting
2022
2022 Annual Meeting
Catalysis and Reaction Engineering Division
Data Science & Machine Learning Approaches to Catalysis III: Applications of Machine Learning to Heterogeneous Catalysis: From Porous Materials to Cluster Catalysis
Thursday, November 17, 2022 - 10:06am to 10:24am
Discovery of catalysts that are stable, active and selectively perform specific desired reactions has been a major challenge for decades. The effect of active site ensembles on catalyst selectivity has gained attention in recent years with intermetallics becoming more popular as catalysts. However, a quantifiable understanding of this phenomenon is absent. Improvement in computational resources has enabled high-throughput calculations of adsorption energies across thousands of surfaces from a wide range of materials. In this work, adsorption energies have been predicted for all species involved in the acetylene semi-hydrogenation reaction using ML models trained on initial DFT data. The impact of site geometries identified using a graph network based tool on adsorption energies and by extension, catalyst selectivity has been investigated. The correlations between the size and shape of the active site ensemble and the adsorption energies have been presented with the objective to determine the most suitable intermetallic catalysts for selective hydrogenation. Available ML models have been compared for adsorption energy prediction to optimize catalyst discovery for general reaction problems.