Detecting stable adsorbates of (1S)-camphor on Cu(111) with Bayesian optimization

Abstract

Identifying the atomic structure of organic–inorganic interfaces is challenging with current research tools. Interpreting the structure of complex molecular adsorbates from microscopy images can be difficult, and using atomistic simulations to find the most stable structures is limited to partial exploration of the potential energy surface due to the high-dimensional phase space. In this study, we present the recently developed Bayesian Optimization Structure Search (BOSS) method as an efficient solution for identifying the structure of non-planar adsorbates. We apply BOSS with density-functional theory simulations to detect the stable adsorbate structures of (1S)-camphor on the Cu(111) surface. We identify the optimal structure among eight 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 new cross-disciplinary tools, such as BOSS, facilitate the description of complex surface structures and their properties, and ultimately allow us to tune the functionality of advanced materials.

First-principles calculations of hybrid inorganic–organic interfaces: from state-of-the-art to best practice

“…Generally, Bayesian Optimization techniques work well when there are only few degrees of freedom, but the computational effort becomes too large for larger systems. 344 Explained in a very simplified way, Bayesian Optimization relies on calculating a few selected geometries and constructing a (conditional) probability distribution for the energy of other geometries, i.e. effectively interpolating between them.…”

“…In contrast to the other optimization methods, it can, therefore, provide a general (at least qualitative) overview of the potential energy surface, including the positions and locations of the minima and saddle points. We presently recommend to employ Bayesian Optimization following the strategy described by the Rinke group: 349 Keep the internal structure of the adsorbate (bond lengths, angles, etc. ) and the substrate fixed, only moving and rotating the molecule as a whole across the surface.…”

Integrating Bayesian Inference with Scanning Probe Experiments for Robust Identification of Surface Adsorbate Configurations

“…In a preparatory study, [ 36 ] we applied BOSS with DFT to identify the stable adsorbate structures of (1 S )‐camphor on the Cu(111) surface. We identified eight stable structures with varying molecular orientations, adsorption sites, and energy barriers.…”

“…We identified eight stable structures with varying molecular orientations, adsorption sites, and energy barriers. Based on their adsorption properties, the structures were classified into two categories, A and B (Ox and Hy in previous study [ 36 ] ). Class A structures, in which (1 S )‐camphor chemisorbs to Cu(111) via oxygen (O), are the most stable and have the highest energy barriers of molecular rotation and diffusion.…”

Integrating Bayesian Inference with Scanning Probe Experiments for Robust Identification of Surface Adsorbate Configurations

“…Stable structures of (1S)-camphor on Cu(111) were identified with BOSS in the minima of the 6D AES [36]. The AES was defined with respect to the position and orientation of the molecule using 3 translational and 3 rotational degrees of freedom.…”