Machine learning-based prediction of polaron-vacancy patterns on the TiO₂(110) surface

V. C. Birschitzky, I. Sokolović, M. Prezzi, K. Palotás, M. Setvín, U. Diebold, M. Reticcioli, C. Franchini

Faculty of Physics and Center for Computational Materials Science, University of Vienna, Vienna, Austria
Vienna Doctoral School in Physics, University of Vienna, 1090 Vienna, Austria
Institut für Angewandte Physik, TU Wien, 1040 Wien, Austria
Department of Physics and Astronomy, Alma Mater Studiorum - Università di Bologna, Bologna 40127, Italy

npj Comput. Mater. 10 (2024) 89

The multifaceted physics of oxides is shaped by their composition and the presence of defects, which are often accompanied by the formation of polarons. The simultaneous presence of polarons and defects, and their complex interactions, pose challenges for first-principles simulations and experimental techniques. In this study, we leverage machine learning and a first-principles database to analyze the distribution of surface oxygen vacancies (V_O) and induced small polarons on rutile TiO₂(110), effectively disentangling the interactions between polarons and defects. By combining neural-network supervised learning and simulated annealing, we elucidate the inhomogeneous VV_O distribution observed in scanning probe microscopy (SPM). Our approach allows us to understand and predict defective surface patterns at enhanced length scales, identifying the specific role of individual types of defects. Specifically, surface-polaron-stabilizing V_O-configurations are identified, which could have consequences for surface reactivity.

Corresponding author: Cesare Franchini. Reprints also available from Ulrike Diebold (diebold).

You can download a PDF file of this open-access article from npj Computational Materials or from the IAP/TU Wien web server.