Physics-informed automated surface reconstructing via low-energy electron diffraction based on Bayesian optimization

Low-energy electron diffraction (LEED) is a cornerstone technique for determining surface atomic structures[heldStructureDeterminationLowenergy2025], yet the quantitative analysis of electron diffraction intensity as a function of incident electron energy – that is, LEED-$\mathit{\text{I(V)}}$ analysis – remains a complex inverse problem. In this work, we tackle quantitative LEED-$\mathit{\text{I(V)}}$ analysis based on physics-informed Bayesian optimization (BO). By embedding multiple scattering LEED forward models directly into a trust-region BO loop, our approach simultaneously optimizes both structural and experimental parameters, adaptively adjusting trust regions for efficient exploration of complex non-convex parameter spaces without manual intervention. The robustness and scalability of the approach are demonstrated using the Ag(100)-(1$\text{time}$1) and Fe\textsubscript{2}O\textsubscript{3}(1$\bar{1}$02)-(1$\text{time}$1) surfaces as examples. Our work establishes a general framework for solving inverse problems in various characterization techniques, unlocking a physics-informed efficient, reproducible, and autonomous paradigm.