Wide-bandgap (WBG) perovskites are key absorbers for perovskite/silicon tandem solar cells, where bandgap tuning is commonly achieved through the incorporation of bromide and cesium. However,the influence of precursor salt selection on the formation pathway and long-term stability of the absorber remains poorly understood. In particular, different precursor combinations may yield comparable final absorbers while introducing substantial differences in the chemistry of the inorganic template, ultimately resulting in markedly different operational stabilities.

Here, we investigate how the choice of cesium and bromide precursors affects the formation and degradation behaviour of two-step processed WBG perovskites. A series of absorbers with bandgaps above ~1.6 eV was prepared using different combinations of CsI, CsBr, FABr and PbBr₂. Device performance and photoluminescence measurements identified two representative high-performing systems, using either CsI or CsBr as the cesium source and FABr as the bromide source, which exhibit markedly different operational stability across many devices (Figure 1, TOC).

Despite displaying nearly identical X-ray diffraction fingerprints, the two absorbers show distinct degradation behaviour during operation at 75 °C. Ex-situ and in-situ GIWAXS measurements reveal that the precursor salt selection substantially modifies the structure of the inorganic template, influencing the formation of PbI₂ complexes and polytype intermediates during crystallization. While the annealed absorbers converge toward similar final crystalline structures, their formation pathways differ significantly up to the final annealing step.

To identify the origin of the stability differences, complementary characterization techniques including TOF-SIMS depth profiling, in-situ photoluminescence under illumination, and structural analysis upon ageing were employed. While no evidence of significant intrinsic bulk compositional inhomogeneity or light-induced halide segregation was observed, structural changes upon thermal ageing suggest that the two absorbers follow distinct degradation pathways.

These results demonstrate that precursor selection influences not only the nominal absorber composition, but also the intermediate chemistry governing crystallization and degradation. Understanding and controlling these precursor-driven processes provides an important route toward more efficient and stable WBG perovskite solar cells.