Light management in highly-textured perovskite solar cells: From full-device ellipsometry characterization to optical modelling for quantum efficiency optimization.

While perovskite solar cells (PSCs) are now reaching high power conversion efficiencies (PCEs), further performance improvement requires a fine management and an optimization of the light pathway and harvesting in the cells. These go through an accurate understanding, characterization and modelling of the optical processes occurring in these complex, often textured, multi-layered systems. In the present work, we have considered a typical methylammonium lead iodide (MAPI) solar cell built on a fluorine-doped tin oxide (FTO) electrode of high roughness (43 nm RMS). By variable-angle spectroscopic ellipsometry (VASE) of the full PSC device, we have been able to determine the optical constants of all the device layers. We have designed a one-dimensional (1D) optical model of the stacked layers where the rough texture is described as layers of effective-medium index. We have supported the model using data extracted from scanning electron microscopy, diffuse spectroscopy and photovoltaic efficiency measurements. We show that the 1D model, while insufficient to describe scattering by the FTO plate alone, gives an accurate description of the full device optical properties. By comparison with the experimental external quantum efficiency (EQE), we estimate the internal quantum efficiency (IQE) and the effect of the losses related to electron transfer. Based on this work, we finally discuss the optical losses mechanisms and the possible strategies that can be implemented to improve light management within PSC devices and further increase their performances.