Ionization potential and electron affinity via inverse surface-energy-weighted averaging in ZnO, NiO, and CH₃NH₃PbI₃

Accurate vacuum-referenced band edges are important for predicting contact barriers and interfacial transport, but values taken from a single surface can misrepresent real crystallites that expose many facets. We develop a workflow that explicitly calculates the ionization potential (IP) and electron affinity (EA) of each facet using slab models with dipole corrections. These facet-resolved quantities are then combined through an inverse–surface-energy weighted average to obtain the final band edges. We introduce three averaging protocols: CLES (Converged Low-Energy Set), the smallest set of low-γ facets for which the weighted average converges within 0.1 eV; PFS (Principal Facet Set), a symmetry-based average over seven low-index surfaces; and CFS (Comprehensive Facet Set), the average over all 25 facets studied. CLES converges with four facets for ZnO, eight for NiO, and twelve for CH₃NH₃PbI₃. PFS remains within ∼0.1–0.2 eV of experiment, and the resulting surface-weighted IP and EA support interface design in perovskite optoelectronics.