Metal halide perovskite solar cells (PSCs) emerged only two decades ago and are already approaching commercialization, holding significant promise for contributing to sustainable development goals. Laboratory‑scale PSCs now demonstrate performance comparable to well‑established photovoltaic technologies [1]; however, their widespread adoption is hindered primarily by limited operational stability and reduced performance when scaled to module‑level areas. In terms of stability, device degradation can originate from the intrinsic instability of the perovskite absorber as well as from chemical and electronic interactions with adjacent layers. Consequently, understanding interfacial chemistry is essential for elucidating how charge‑transport layers interact with perovskite absorbers.

State‑of‑the‑art PSCs commonly employ p‑i‑n architectures featuring molecular self‑assembled monolayers as the hole transport layer (HTL), followed by the perovskite absorber and a bilayer electron transport layer (ETL). This ETL typically consists of C60 combined with atomic‑layer‑deposited tin oxide (SnO_x) on top of the perovskite. The perovskite/C60 interface is known to be a significant source of non‑radiative recombination [2]. Additionally, several studies have shown that direct contact between perovskite and ALD‑grown SnO_x can lead to the formation of performance‑limiting interfaces [3]. Device performance is most often assessed only after full device fabrication, which provides limited insight into how each subsequent deposition step affects the underlying layers or alters interfacial chemistry. X‑ray photoelectron spectroscopy (XPS) is a powerful tool for probing chemical environments, and by using multiple photon energies available at synchrotron facilities, different depths within the solar‑cell stack can be investigated. Therefore, it makes XPS particularly well‑suited for studying interfacial chemical interactions in PSCs involving C60, SnO_x, and perovskite.

In this work, we performed in‑situ SnO_x growth on various perovskite/C60 structures while simultaneously conducting ambient‑pressure XPS measurements at MAX IV synchrotron facility. Our results reveal that the initial ALD cycles of SnO_x induce perovskite‑composition‑dependent ion migration through the C60 layer, which in turn affects both SnO_x stoichiometry and growth‑per‑cycle behaviour. This ion migration is primarily triggered by chemical interactions between the Sn precursor and Br‑containing perovskite surfaces, with the Sn precursor first permeating through the C60 film. The direct interaction between SnO_x precursors and perovskites of different compositions was further investigated using soft and hard XPS on samples where SnO_x was deposited directly on FAPbI₃ and FAPbBr₃. The measurements clearly show degradation of the FA⁺ cation and formation of PbI₂ in FAPbI₃, as well as the formation of Sn–Br bonds in FAPbBr₃ [4].

Although the presence of a C60 interlayer can partially mitigate these detrimental reactions affecting both performance and stability, it does not completely suppress them. These findings therefore highlight the need for incorporating dedicated ion‑blocking layers between C60 and perovskite, as well as between C60 and SnO_x, as a crucial strategy for enhancing the long‑term stability of PSCs.