Sudipta Seth, Eduard A. Podshivaylov, Jun Li, Marina Gerhard, Alexander Kiligaridis, Pavel A. Frantsuzov, Ivan G. Scheblykin

Photoluminescence (PL) blinking is a common phenomenon in nanostructured semiconductors associated with charge trapping and defect dynamics. PL blinking kinetics exhibit very broadly distributed timescales. The traditionally employed analysis of probability distribution of ON and OFF events suffers from ambiguities in their determination in complex PL traces making its suitability questionable. Here, the statistically correct power spectral density (PSD) estimation method applicable for fluctuations of any complexity is employed. PSDs of the blinking traces of submicrometer MAPbI3 crystals at high frequencies follow power law with excitation power density dependent parameters. However, at frequencies less than 0.3 Hz, the majority of the PSDs saturate revealing the presence of a maximal characteristic timescale of blinking in the range of 0.5–10 s independently of the excitation power density. Super-resolution optical microscopy shows the characteristic timescale to be an inherent material property independent of polycrystallinity. Thus, for the first time the maximum timescale of the multiscale blinking behavior of nanoparticles is observed demonstrating that the power law statistics are not universal for semiconductors. It is proposed that the viscoelasticity of metal-halide perovskites can limit the maximum timescale for the PL fluctuations by limiting the memory of preceded deformations/re-arrangements of the crystal lattice.