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

Abstract

Controlling the properties of organic/inorganic materials requires detailed knowledge of their molecular adsorption geometries. This is often unattainable, even with current state-of-the-art tools. Visualizing the structure of complex non-planar adsorbates with atomic force microscopy (AFM) is challenging, and identifying it computationally is intractable with conventional structure search. In a fresh approach, we propose to integrate cross-disciplinary tools for a robust and automated identification of 3D adsorbate configurations. We employ Bayesian optimization with first-principles simulations for accurate and unbiased structure inference of multiple adsorbates. The corresponding AFM simulations then allow us to fingerprint adsorbate structures appearing in AFM experimental images. In the instance of bulky (1S)-camphor adsorbed on the Cu(111) surface, we found three matching AFM image contrasts, which allowed us to correlate experimental image features to distinct cases of molecular adsorption.

Protective Coating Interfaces for Perovskite Solar Cell Materials: A First-Principles Study

“…In analogy to [133,177,178], the structural search with the BOSS code and DFT, follows a four-step approach as shown in Figure 3.9a. In all the structural models, I kept the distance between the coating and the perovskite along the z-axis constant at ∼ 4 Å based on preliminary investigations with a 3-dimensional search space.…”

Protective Coating Interfaces for perovskite Solar Cell Materials: A first Principles Study

“…BOSS has already shown its power in problems such as conformer search for organic molecules, 53,54 organic molecule adsorption at semiconductor surfaces, 53 and film growth of organic adsorbates on metallic surfaces. [55][56][57] In this work, we employ BOSS to tackle the interface problem between two inorganic materials (coating and HP) in step I, which we believe can efficiently narrow down the search space for further geometry relaxation in step II. In our BOSS search, state-of-the-art single-point DFT calculations are performed for different structures, and BOSS correlates the structures with an energy landscape through active learning with a BO algorithm.…”