Unlike conventional semiconductor materials, metal halide perovskites (MHP) possess soft and ionic crystal structures leading to several unique features like facile ion migration, self-healing, elasticity, and memory. Within this dynamic system, external stimuli like high photon doses, electron beams, electrical bias, and mechanical stress induce structural changes and alter associated optoelectronic properties. Therefore, it is crucial to investigate the structure-photophysics relationship in these materials, especially in operational devices like solar cells, where traditional methods such as scanning electron microscopy fall short due to the layered structure. Moreover, electron and x-ray-based analytical techniques are often invasive altering the material properties.

To address these challenges, we developed Correlation Clustering Imaging (CLIM), a novel noninvasive method that utilizes photoluminescence fluctuations to reveal contrasts associated with defect dynamics in semiconductor materials. In films, CLIM images show the polycrystalline grain structure ideally correlating with electron microscopy images. Analysis of photoluminescence fluctuations suggests the presence of one type of metastable defect dominating the fluctuations. Although the relative amplitude of PL fluctuations is small in films, it is significantly larger in solar cells under short-circuit conditions. The correlated regions within devices are notably larger (up to 10 micrometers) than those in films (up to 2 micrometers). We propose that the regions resolved by CLIM in solar cells possess a common pool of charge extraction channels, which fluctuate and cause PL to vary. Since photoluminescence fluctuations report on the dynamics of non-radiative recombination processes, CLIM offers insights into structural and functional aspects related to carrier transport, ion migration, defect formation and annihilation, and recombination losses, which are crucial for the rational engineering of the next generation of devices.

CLIM provides imaging contrasts based on properties which were never used before for optical non-invasive imaging of luminescent materials and devices based on them in operando. The broad applicability of CLIM, requiring only a standard wide-field microscope and our user-friendly, open-source algorithm, positions it as an important new tool for material chemists, engineers, and device scientists.