Despite the rapid progress in perovskite solar cells, their commercialization is still hindered by issues regarding long-term stability, which can be strongly affected by metal oxide-based charge extraction layers next to the perovskite material. With MoO3 being one of the most successful hole transport layers in organic photovoltaics, the disastrous results of its combination with perovskite films came as a surprise but was soon attributed to severe chemical instability at the MoO3/perovskite interface. To discover the atomistic origin of this instability, we combine density functional theory (DFT) calculations and X-ray photoelectron spectroscopy (XPS) measurements to investigate the interaction of MoO3 with the perovskite precursors MAI, MABr, FAI, and FABr. From DFT calculations we suggest a scenario that is based upon oxygen vacancies playing a key role in interface degradation reactions. Not only do these vacancies promote decomposition reactions of perovskite precursors, but they also constitute the reaction centers for redox reactions leading to oxidation of the halides and reduction of Mo. Specifically iodides are proposed to be reactive, while bromides do not significantly affect the oxide. XPS measurements reveal a severe reduction of Mo and a loss of the halide species when the oxide is interfaced with I-containing precursors, which is consistent with the proposed scenario. In line with the latter, experimentally observed effects are much less pronounced in case of Br-containing precursors. We further find that the reactivity of the MoO3 substrate can be moderated by reducing the number of oxygen vacancies through a UV/ozone treatment, though it cannot be fully eliminated.
Keywords: DFT; metal oxide; perovskite; photoelectron spectroscopy; stability.