Publication date: 21st July 2025
Tin-based halide perovskites promise lead-free optoelectronic devices, but surface degradation remains the critical barrier to commercialization. Here, we combine density functional theory calculations with thermodynamic modeling to reveal the atomic-scale mechanisms governing stability in the 2D hybrid perovskite 4FPSI. By systematically investigating point defects including vacancies and interstitials in both in-plane and out-of-plane configurations, we uncover how defect spatial distribution controls surface reactivity and device performance.
Our calculations identify two competing degradation pathways. Sn(II) oxidation to Sn(IV) via O₂ exposure dominates under ambient conditions, while I⁻/I₃⁻ redox processes prevail in solution environments. Surface stoichiometry critically determines defect energetics. Sn-poor terminations generate deep acceptor V_Sn traps within the bandgap, whereas Sn-rich surfaces promote interstitial formation that disrupts the SnI₆ octahedral framework. Organic cation coverage provides crucial protection: complete 4F-PEA⁺ layers create substantial water diffusion barriers, but partial coverage exposes reactive Sn sites with significantly lower activation energies.
Most remarkably, we discover intrinsic self-healing through surface reconstruction and interstitial migration. This finding explains experimentally observed performance recovery after rest periods and fundamentally reframes our understanding of tin perovskite stability. The dynamic equilibrium positions Sn(IV) as both a degradation product and a healing intermediate. Our mechanistic insights reveal that surface engineering, not bulk composition, holds the key to stable lead-free photovoltaics. These results provide clear design principles for defect passivation and interface optimization in next-generation solar cells.
This work was supported by the OHPERA project (HORIZON-EIC-2021-PATHFINDEROPEN-01) funded by the European Union.
The Barcelona Supercomputing Center (BSC-RES) is acknowledged for providing computational resources.