Ion Migration and Dopant Effects in the Gamma-CsPbI Perovskite Photovoltaic Material: Atomistic Insights through and Machine Learning Methods.

Inorganic halide perovskites such as CsPbI are attracting increasing attention for solar cell and optoelectronic applications. Ion migration is known to be an important factor in perovskite behavior, but the impact of cation dopants on iodide diffusion in the room-temperature orthorhombic γ-CsPbI is not fully understood, especially at the atomic level. Here, we investigate the effect on iodide migration of incorporating different cations (including Sn, Ba, and Cu) into γ-CsPbI, focusing on maintaining an inorganic phase rather than doping with molecular organic ions. Through a combination of and machine learning (ML) techniques, our results show that the simulated structure, band gap, and ion migration energies are in good agreement with experimental data. We find that partial Pb-site substitution does not have a major suppressing effect on iodide ion transport, which is important for guiding future doping work. An ML interatomic potential model was derived for large-scale simulations (∼80 ns) of the pristine and Sn-doped materials, which reveal iodide diffusion paths along the Pb-I octahedral edges with no correlated cation motion. Structural analysis indicates an ordered cation sublattice but disorder in the anion sublattice, indicative of high iodide ion mobility similar to fast-ion conductors.