Surface properties of metal halide perovskites from density functional theory and neural network potentials

Surface properties of halide perovskites play important roles in the performance of perovskite-based devices as they can influence the stability, defect formation, and charge carrier transport properties. In this study, we reveal the structural and energetic properties of surfaces and high-index facets of both Sn-based and Pb-based perovskites. These properties are calculated combining density functional theory and neural network potentials. Neural network potentials are trained using the low-index surface data points [1] and applied to predict the surface energies of high-index facets. We will discuss the surface phase diagrams and reconstructions of the surfaces and facets considering different stoichiometries and evaluate their stabilities under different thermodynamic conditions. We will also compare the surface property of Sn-based and Pb-based and discuss how surfaces affect defect formation.

[1] K. Li, C. Khamdang, and M. Wang “Surface phase diagram of CsSnI₃ from first-principles calculations”, Phys. Rev. Mater., 8, 093401 (2024).