Stress Relaxation in Metal Halide Perovskites and use of Controlling Film Stress to Improve Stability

Challenges to upscaling metal halide perovskites (MHPs) include mechanical film stresses which accelerate degradation, dominate at the module scale, and can lead to delamination or fracture. In this work, we demonstrate open-air blade coating of single-step coated perovskite as a scalable method to control residual film stress after processing and introduce beneficial compression in the thin film with the use of polymer additives such as gellan gum and corn starch.<br/><br/>Given the strong link between residual stress and degradation, we assessed the stability of perovskite films with different residual stress states. Perovskite films on silicon and glass were stress engineered using additives to either be tensile or compressive and aged under either high heat or humidity conditions (85°C vacuum or 85% relative humidity). A control MAPbI₃ film spin coated without additives was also added as a reference under 85% relative humidity. The samples were periodically removed, and the stress was remeasured in ambient conditions (22°C and 40% RH). The highly tensile samples experienced stress relaxation quickly in the humid environment due to moisture induced degradation. This effect was not observed for samples with compressive stress, which also did not experience visible degradation during the 100-hour exposure. We performed XRD measurements on these films to verify degradation and the tensile sample had PbI₂ peaks after the aging experiment, whereas the compressive sample had lower intensity PbI₂ peaks and still had MAPbI₃ in the film. This confirms that perovskites are more structurally stable when exposed to environmental conditions while in compression. Samples were reannealed to ensure that trapped moisture was not the cause of the relaxation, and there was no change observed in the stress. Therefore, residual moisture is not the explanation for stress reduction.<br/>.<br/>A similar experiment was conducted for thermal stability at 85°C under vacuum to monitor the changes in stress under thermal aging. Once again, stress relaxation was observed for the highly tensile sample with gum. The improvement in stability was visually apparent since the tensile films turned completely yellow after being aged for 100 hours while the compressive one remained largely in the black photoactive phase. The samples with additives took longer to visibly degrade, which also points towards their improved stability. The tensile stress was therefore relaxing as the film degraded due to the strained bonds likely enabled the methylammonium cation to diffuse out, whereas a compressive stress showed better stability and unchanging stress under heat, similar to the effect observed with aging in high humidity.<br/>A thermal cycling experiment was also conducted for a tensile and compressive sample to monitor the perovskite films as the temperature changes between -40 to 85°C. The experiment ran for 200 cycles and the results supported what was observed under high heat and humidity conditions. The tensile sample relaxed by 180 MPa and turned completely yellow. However, the compressive sample was stable both in terms of its appearance and film stress. A spin coated MAPbI₃ film without additives (with tensile residual stress) was kept as the control sample. In all three aging conditions, films with residual tensile stress degraded significantly faster. The degradation also corresponded to stress relaxation in the perovskite film as the phase changed, marking the first time for which changes in stress have been directly mapped to stability. Stress relaxation has been observed in many other material systems (<i>i.e., </i>polymers and metals) when exposed to tensile stress. Perovskites therefore act similarly, and despite their brittle nature, share in this property.<br/><br/>This discovery of a new mechanism underpinning MHP degradation shows that film stress can be used as a parameter to screen MHP devices and modules for quality control before deployment as a design for reliability criterion.