Among the various deposition techniques available for perovskite thin-film fabrication, thermal evaporation stands out for its ability to produce uniform and high quality films with precise control over thickness and composition. However, it is known that this method comes with its challenges, such as the decomposition of organic components and limited control over morphology. While most studies on thermal evaporation focus on pure iodide systems, like MAPbI₃ and FAPbI₃, the challenges related to mix-halide perovskites, which are crucial for bandgap tuning and device stability, remain elusive. These challenges are further compounded by the critical role of substrates, which play a critical role in determining the morphology, crystallinity, and optoelectronic properties of the resulting films.

For this purpose, we systematically investigated the influence of various substrates, such as PTAA, NiOx, PEDOT:PSS and the self-assembled monolayer (SAM) MeO-2PACz, on the formation of thermally evaporated FAPbI₁Br₂ perovskite films with thicknesses ranging from 3 nm to 200 nm. Characterization techniques including X-ray photoelectron spectroscopy (XPS), ultraviolet photoelectron spectroscopy (UPS) and  scanning electron microscopy (SEM) were utilized to analyze the surface properties and the morphology of the evaporated films. Bulk properties, like crystal structures and optical absorption characteristics of the films, are investigated using X-ray diffraction (XRD) and UV-vis spectroscopy.

Our results reveal distinct substrate-dependent effects on the formation and composition of resulting perovskite thin films. For instance, on PEDOT:PSS  and NiOx we initially only observe FAPbI₃ formation for low coverages, and only after the deposition of ~45 nm the bromide becomes incorporated and the desired FAPbI₁Br₂ forms. This delayed incorporation is attributed to the formation of volatile bromine species, such as HBr or Br₂, during co-evaporation, triggered by the interaction of the precursors with the substrate surface at the interface. In contrast, on PTAA substrates, bromide is incorporated right away, forming the intended FAPbI₁Br₂ composition, which indicates the absence of reactions leading to volatile bromine species. Maybe the most surprising results were obtained for the SAM substrates, which are commonly and successfully employed in solution-processed perovskite devices. Here, no bromide incorporation was observed for any layer thickness up to 200 nm and overall the perovskite formation was hindered. This highlights the significant challenges SAM layers may present in thermal evaporation and that the choice of substrate can influence the film growth not only at the interface but also across device-relevant thicknesses.

In conclusion, this study provides valuable insight into the critical role of the chosen substrate in the thermal evaporation of FAPbI₁Br₂ perovskite thin-film, with the challenges of achieving consistent halide incorporation being particularly evident through the formation of volatile bromide species. Understanding the relationship between substrate properties and perovskite film characteristics is crucial for optimizing device performance and advancing the development of efficient and stable perovskite solar cells.