To consider measurement artifects is important when dealing with "soft" materials.

http://link.springer.com/article/10.1007%2Fs12200-016-0573-8

A systematic ambient Raman spectroscopy characterization study of MAPbI₃thin films is conducted by varying the excitation laser wavelength, integrationtime, and filter characteristics. The comprehensive results, which are obtainedusing available characterization resources, indicate that Raman scattering under laser excitation under ambient conditions from MAPbI₃perovskite is insignificant compared to that from the precursor phases and/or egradation products (MAI, PbI₂, PbO_x, etc.). This does not exclude the possibility of the existence of the weak Raman scatteringin MAPbI₃ perovskite. It is found that the use of typicallyavailable laser excitation sources (532 nm and 633 nm) easily damages the MAPbI₃thin films, and resulting Raman spectra contain spectral artifacts and bandsfrom Raman scattering by primary (PbI₂, MAI) and secondary (leadoxides) degradation products rather than the MAPbI₃ perovskiteitself. This becomes a challenging issue in the ambient Raman characterizationof MAPbI₃ perovskite, and it is likely to be the source of thediscrepancies found in the Raman spectra of MAPbI₃ perovskitereported in the literature. This challenge may be overcome by using the lesscommon infrared (>830 nmwavelength) excitation lasers. The lessons learnt from this Raman spectroscopystudies of MAPbI₃ perovskite can be extended to otherlaser-sensitive light-absorbing compounds such as CsSnI₃, HC(NH₂)₂PbI₃,etc.