Detecting Stable Surface Adsorbates with Bayesian Optimization

Determining stable structures of organic molecules at inorganic surfaces requires both quantum mechanics and thorough exploration of the potential energy surface (PES). This is prohibitively expensive with density-functional theory (DFT). We combine DFT with artificial intelligence (AI) for global atomistic structure search of stable adsorbates. Bayesian Optimization Structure Search (BOSS) [1] is a new AI tool, which accelerates the structure search via a strategic sampling of the PES. BOSS minimizes the number of expensive DFT simulations to compute the complete PES. This allows a clear identification of the most stable minimum energy structures and the barriers between them.

We apply BOSS to study the adsorption of a camphor molecule on the Cu(111) surface as a function of molecular orientation and translations. We identify the optimal structure and 8 unique types of stable adsorbates, in which camphor chemisorbs via oxygen (global minimum) or physisorbs via hydrocarbons to the Cu(111) surface. This study demonstrates that the new cross-disciplinary tools, like BOSS, allow us to identify complex interface structures and properties, and tune the functionality of advanced materials.

[1] M. Todorović, M. Gutmann, J. Corander and P. Rinke, npj Comput. Mater. 2019, 5, 35.