Thermal stability of lead-halide perovskites in perovskite solar cells

• Main Author: Tay, Darrell Jun Jie
• Other Authors: Nripan Mathews
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2025
• Subjects: Engineering; Perovskite; Stability; Thermal
• Online Access: https://hdl.handle.net/10356/182581
• Institution: Nanyang Technological University

Of the various third-generation solar cell technologies, lead-halide perovskite-based solar cells have achieved power efficiencies rivalling that of traditional silicon-based solar cells. This has spurred attempts to commercialise this new technology. However, thermal stability remains an unresolved issue for lead-halide perovskites, which impede their use in commercial applications. Elevated temperatures cause the perovskite structure to break down, and the material loses the property to absorb light efficiently. Thus, methods to stabilise the perovskite structure are necessary for eventual commercialisation. Compositional variation offers one pathway to attaining thermally stable perovskite films, although the current literature is shrouded in contradictions. One example is the effect of varying the FA+/MA+ ratio and Cs+/FA+ ratio in mixed FA-MA perovskites and Cs-FA perovskites respectively. Another example is the lack of empirical knowledge, such as that of all-inorganic caesium lead iodide. In addition, various claims, some untested and some contradictory, have been put forth in an attempt to explain the thermal stability of these perovskites, such as morphological issues, or defect densities. The composition of the perovskite is the most fundamental component of the perovskite solar cell to design and alter. Therefore, it is of utmost importance to look into the thermal stability trends of perovskites with these A-site cations, as well as to assess these claims on the other factors, to ensure that the intrinsic thermal stability of the perovskite film is as high as possible, before consideration of its use in devices. To achieve this end, the first work investigates the thermal stability trend of FA+/MA+ perovskites, along with the other claims that are often blamed for instigating the thermal breakdown of the perovskite film. FA-MA perovskites were found to be more thermally unstable than pure FAPbI3 perovskite, owing to the volatility of the MA+ cation. Other factors such as the morphology of the films, and defect densities in the film, do not correlate with the thermal stability trend in FA-MA perovskites. In addition, vertical segregation of the MA+ cation to the top of the perovskite film was observed, but it does not appear to affect the stability trend. This was shown via coating FAI above a layer of MAPbI3. When compared with pure MAPbI3, ; its thermal stability was similar to pure MAPbI3 filmsthe difference in thermal stability was negligible. The presence of a layer above the film does not alter this trend too, highlighting the fact that the trend is preserved even in devices. Finally, it was shown that FAPbI3 films with MACl additive not removed from the interior of the film also triggers thermal breakdown, therefore showing the importance of eliminating MA+ from the perovskite film. As FA+ is also volatile and susceptible to breakdown, an investigation into CsPbI3 was made – as numerous claims tout the thermal stability of CsPbI3. However, it was shown that thermally-induced phase degradation was not taken into account, which would render a device useless. Therefore, thermal stability has to be re-defined to take the thermally-induced phase degradation into account. The degradation was found to speed up with increasing temperatures, highlighting that the perovskite phase of CsPbI3 is only kinetically-stable at ambient temperatures. DMAI-doped CsPbI3 films (at 100 mol%) were also tested, and shown to undergo thermally-induced phase degradation as well, albeit at a much slower rate than pure CsPbI3. The stabilisation effect disappears however when the concentration of DMAI introduced into the CsPbI3 film is in great excess (150 mol%). The DMA+ was found to be part of the perovskite structure, contrary to the claims made by several works in literature. DMA+ then leaves the perovskite, forming CsPbI3, which then degrades, and therefore the DMA+ prolongs the longevity of DMAI-doped CsPbI3 films. The effect of the excess I on the system was also tested, by deliberately doping CsPbI3 films with KI instead. The KI was found to accelerate the degradation process, although the evidence suggests that the KI did not actually dissociate. Therefore, it points to the possibility that the excess I- in the DMAI-doped film may not be in free iodide form, but rather as part of a compound, likely DMAPbI¬3, which was found to remain in the film after it degrades. The low temperature onset for thermally-induced phase degradation of CsPbI3 suggests that mixing FA+ and Cs+ would produce films with highest thermal stabilities. Therefore, Cs-FA perovskites were investigated as an alternative, and it was shown that Cs0.2FA0.8PbI3 appeared to have the highest thermal stability of the Cs-FA perovskites. Morphologies were ruled out as a factor once again. Angle-dependent grazing-incidence XRD was done on these films, and it was shown that the top surface has a smaller lattice parameter than the bottom surface. This indicates that Cs+ is segregated to the top of the film, much like in the case of the MA+ in FA-MA perovskites. It is posited that the higher amount of Cs+ on the surface reduces the surface area from which FA+ can leave the perovskite, thus slowing down degradation, and it is likely that Cs0.2FA0.8PbI3 has more Cs+ on the surface, thereby preventing the escape of FA+ better than pure FAPbI3, although degradation rates for both perovskites appear similar in the long run. This therefore spurs on additional future works which tackle the thermal stabilities of these films by preventing the loss of FA+ in order to maximise thermal stability of the perovskite films.