Development and evaluation of force fields for studying the formation of metal halide perovskites from solution

Metal halide perovskites have rapidly evolved into an efficient, low-cost alternative to conventional photovoltaic materials. Devices with power conversion efficiencies over 20% have been made within only a decade of their first application. However, further development is necessary before perovskite solar cells can be widely adopted by industry. For instance, they cannot currently be printed at scale. Even though a wide variety of specific fabrication methods exist to make hybrid perovskites with desired properties and composition, a molecular-level understanding of how they form from solution and degrade is still lacking. Here, we expand the AMOEBA polarisable force field so that it can be used to simulate a wide range of lead halide perovskites. We show that even with minimal parameter optimisation, this new polarisable force field offers better transferability between solution and solid state than existing non-polarisable force fields. For example, it yields accurate hydration free energies (within 5%) for typical ionic species while showing the correct ordering of stability for the different crystal polymorphs of CsPbI3 and MAPbI3. We use this new force field to study the coordination and thermodynamics of MAPbI3 precursors in DMF and DMSO, two solvents commonly used to synthesise metal halide perovskites. This provides insights that help rationalise experimental conditions favourable to the formation of MAPbI3, including concentration and temperature effects. We then consider the formation of MAPbI3 via one-step and two-step methods, and analyse the mechanism of crystal growth during an unbiased ~2.5 µs trajectory. Overall, we find that our newly parameterised force field shows promise as a transferable and reliable force field for studying crystal formation, degradation and other interfacial phenomena involving these ionic compounds