Abstract

Recently developed machine learning methods hold great promise for simultaneously reducing the computational cost and increasing the accuracy in catalysis modeling, allowing us to capture more complexity, make our models more realistic and perhaps even obtain new physical insights [1]. In my talk I will focus on recent work aimed at predicting adsorption configurations and energies of complex molecules at the surfaces of transition metals and alloys using a graph-based Gaussian Process Regression ML model (WWL-GPR [2]). We apply the methodology to study CO2 hydrogenation (reverse water-gas shift) over single-atom alloy (SAA) catalysts, i.e., diluted bimetallic materials. SAAa have recently attracted considerable interest since adsorption at the sites offered by their surfaces can break the scaling relationships between adsorption energies and reaction barriers that limit conventional catalysts [3]. The accuracy and low cost of the applied methods allows us to consider a wide combinatorial space of elements of the periodic table, paving the way toward the design and nano-engineering of SAA catalysts.

[1] M. Andersen and K. Reuter, Acc. Chem. Res. 54, 2741 (2021).

[2] W. Xu, K. Reuter, and M. Andersen, Nat. Comp. Sci. 2, 443 (2022).

[3] RT. Hannagan et al., Chem. Rev. 120, 12044 (2020).