Hybrid organic / inorganic perovskites such as methylammonium lead tri-halides (MAPbX_3-nY_n: X, Y = halogen, n = 0-3) are materials of substantial interest as the light-harvester in photovoltaic devices. There is currently some debate about the effect of processing upon the structure and composition of the resulting material, including suggestion, in the case of the mixed halide MAPbI_3-nCl_n, that there is only one halogen (iodine) present in the resulting material. We have used thermal analysis, spectroscopy and ‘hyphenated’ techniques, which facilitate evolved gas analysis, to understand changes occurring during material processing. 

Samples of MAPbI_3-nCl_n were prepared by processing aliquots of 40 % precursor solution (in DMF) at different cure temperature / time combinations. TGA and DSC were used to monitor mass loss and heat flow isothermally (during the cure process) and on a temperature ramp (after the material had formed). The hyphenated techniques TGA-GCMS and DTA-FTIR were used to analyse any volatiles released during each process.  

TGA-FTIR and TGA-GCMS showed only solvent evolution during the cure step. Post-cure TGA-FTIR analysis of material prepared at 100 °C (15 minutes) and 30 °C (240 minutes) shows that the resulting material still contains residual solvent, which is released at increasing temperature; it is possible that a small amount of solvent becomes incorporated within the perovskite matrix, requiring bond cleavage (and consequently increased energy) for its release. 

Post-cure Differential Scanning Calorimetry (DSC) shows differences in thermodynamic properties. A sample cured at 100 °C for 80 minutes shows no features over the temperature range 20 – 120°C but a sample cured at 30 °C shows overlapping features around 80 °C; these are not present on subsequent scans. As solvent is demonstrably present in the processed sample this is not surprising; however, the number and nature of the features is interesting. They could be attributed to simple release of solvent or something more complex: if the incorporated solvent acted as a plasticizer, the material may have formed in an amorphous state; the thermodynamic features could therefore include a phase change.  

DSC also demonstrates that the material resulting from the MAPbI_3-nCl_n precursor is not simply MAPbI₃. It is well documented that MAPbI₃ undergoes a defined, reversible, tetragonal – cubic phase change around 55 °C, which is easy to replicate using the MAPbI₃ / DMF precursor; however, this phase change not present in materials produced from MAPbI_3-nCl_n / DMF; this shows that the material resulting from the mixed halide is not simply MAPbI₃.