Among reliability studies on perovskite photovoltaics (PV) cells and modules, partial shading degradation is a crucial and under-investigated topic. In the present work, we use a combination of mapping electroluminescence (EL), photoluminescence (PL), and illuminated lock-in thermography (ILIT) to gain insight into the reverse bias degradation mechanisms induced by partial shading on a monolithically interconnected module. Spatial inhomogeneities across the cell length are shown to play an important role in the degradation. A perovskite module was subjected to partial shading, causing, in the lower region of the shaded cells, a PL signal intensity increase and EL decrease. We suggest the formation of a barrier at one of the perovskite/transport layer interfaces, preventing both carrier extraction in PL and carrier injection in EL. A simple model for the current flow in the presence of the barrier can satisfactorily explain the EL, PL, and ILIT behavior and point to some possible propagation mechanisms. In summary, we show that studying partial shading degradation at module level draws a more complex and realistic picture of the interplay between material and electrical parameters than cell-level studies. We also demonstrate that luminescent and thermal imaging techniques can be combined to draw meaningful conclusions on the degradation mechanisms, their formation, and propagation.

In realistic partial shading scenarios, the impact of low-intensity illumination needs to be considered. However, there is barely any research available and the published results are contradictory. Here, it is shown that the reverse bias behavior of perovskite solar cells under low-intensity illumination strongly depends on the voltage scan rate. As explanation, a hypothesis is developed and experimentally verified that is based on two antagonistic mechanisms: on one hand, illumination affects the mobile ions conductivity, decreasing the breakdown voltage. On the other hand, an electrochemical reaction caused by the reverse bias current increases the breakdown voltage. Since the two mechanisms occur on slightly different time scales, it depends on the voltage scan rate which mechanism dominates. These findings emphasize that more detailed research into the mechanism occurring in reverse bias and the factors affecting the reverse bias breakdown is still necessary. Firstly, a deeper understanding would be helpful for investigating the effect of realistic, non-ideal partial shading scenarios on perovskite modules. Secondly, knowing how cell properties and external factors influence the breakdown voltage is necessary for defining standardized measurement procedures that allow the comparison of different perovskite solar cells in regards to their reverse bias stability.