Data-Guided Spacer Designing in Dion-Jacobson Phase Halide Perovskites: Boosting Optoelectronics through...

Dion-Jacobson (DJ) phase layered halide perovskites (LHPs) are promising candidates for next-generation optoelectronics, yet systematic optimization remains challenging due to the limited understanding of spacer cation-induced structure-property relationships. Here, we integrate advanced atomistic simulations with interpretable machine learning to uncover molecular design principles for lead iodide-based LHPs. Data-driven correlation analyses reveal that halogen-functionalized aromatic spacers enhance packing efficiency, thereby increasing lattice rigidity and suppressing detrimental electron-phonon coupling. Bromination stoichiometry in spacers emerges as a powerful handle to tune interspacer and spacer-inorganic noncovalent interactions, affording precise control over structural dynamics. Nonadiabatic molecular dynamics simulations demonstrate that such rigid architectures markedly suppress nonradiative recombination, extending carrier lifetimes through weakened instantaneous couplings and stabilized bandgaps. SHapley Additive exPlanations highlight decisive heterointerfacial descriptors, including layer spacing, Br-I distances, and axial Pb-I bonds, that govern bandgap fluctuations and transition probabilities. These insights establish halogen-mediated noncovalent interactions as a chemically intuitive design strategy for optimizing LHPs.