Perovskite solar cells have entered a stage where market introduction is within reach. For successful large-scale commercialization, precise control over the quality of the perovskite material is crucial. To achieve this, the research community currently relies heavily on laboratory experience, engineering solutions, and other heuristic approaches. One of the central challenges remains the controlled crystallization of perovskite thin films. Various strategies, such as anti-solvent techniques and additive engineering, are widely used to increase process control. While these approaches have demonstrably influenced the morphology and quality of the resulting layers, the fundamental mechanisms governing the crystallization process are still the subject of vigorous debate. A frequently cited hypothesis is that crystallization additives mediate the heterogeneous nucleation, where solvate complexes are expected to form the colloidal seeds for crystallization.[1,2] However, direct and unambiguous insights linking the complex formation in precursor inks to the final perovskite structure remain elusive.

In this work, we present compelling evidence that the decisive influence of typical crystallization additives is not the nucleation phase. Instead, they promote coarsening grain growth by enhancing ion mobility across grain boundaries. Our conclusions are drawn from a comprehensive study that integrates ex-situ and in-situ characterization techniques, device performance analysis, phase-field simulations, and density functional theory (DFT) calculations. This multi-faceted approach allows us to propose a general mechanism that holds true across a variety of additives and perovskite compositions.

Starting from the precursor ink, we monitor the evolution of colloidal species using NMR, UV–vis spectroscopy, and electrochemical conductivity measurements. We then investigate the perovskite formation dynamics via in-situ GIWAXS, supported by phase-field simulations that model the crystal growth process. These analyses provide the first clear evidence that grain coarsening, rather than nucleation, is the dominant mechanism shaping the final film morphology. To identify the mechanism underlying the enhanced coarsening rate, we employ solid-state NMR, which reveals that additives significantly increase ion mobility across gain boundaries. Complementary techniques, including UPS, FTIR, and DFT calculations, reveal that this enhanced mobility likely arises from the interaction of additives with the grain boundaries. Upon annealing, weakly bound additives detach, opening up defect states, that act as ion channels. In some cases, additives can also serve as ion shuttles, further facilitating ion transport across grain boundaries.

We validate the generalizability of our proposed mechanism by applying it to various perovskite formulations and additive types. Moreover, we demonstrate that the same underlying principles can explain the effects of post-processing techniques, where an additive may be processed on top of an already crystallized perovskite film. Indeed surprising similarities exist to the thermal hot-pressing process, where ion mobility is increased through elevated temperatures.[3]

Our finding provides a decisive piece that complements the nucleation theory for perovskite thin film fabrication and which bridges the gap between the precursor phase and the final film. Thereby, we take a crucial step beyond heuristics and pave the way for revised additive and crystallization engineering.