Inverted Hysteresis in n–i–p and p–i–n Perovskite Solar Cells

A combination of experimental studies and drift-diffusion modeling has been used to investigate the appearance of inverted hysteresis, where the area under the J–V curve for the reverse scan is lower than in the forward scan, in perovskite solar cells. It is found that solar cells in the p–i–n configuration show inverted hysteresis at a sufficiently high scan rate, whereas n–i–p solar cells tend to have normal hysteresis. By examining the influence of the composition of charge transport layers, the perovskite film crystallinity and the preconditioning treatment, the possible causes of the presence of normal and inverted hysteresis are identified. Simulated current–voltage measurements from a coupled electron–hole–ion drift-diffusion model that replicate the experimental hysteresis trends are presented. It is shown that during current–voltage scans, the accumulation and depletion of ionic charge at the interfaces modifies carrier transport within the perovskite layer and alters the injection and recombination of carriers at the interfaces. Additionally, it is shown that the scan rate dependence of the degree of hysteresis has a universal shape, where the crossover scan rate between normal and inverted hysteresis depends on the ion diffusion coefficient and the nature of the transport layers.