What can J–V hysteresis tell us about defect mediated phenomena in perovskite based solar cells?

J–V scans of lead-halide perovskite solar cells can show higher fill factors and open-circuit voltages during the reverse sweep depending on measurement conditions,[1] a phenomenon known as J–V hysteresis. A consensus has emerged that this behaviour is due to the halide ion vacancy distribution adjusting to the preconditioning voltage and screening the built-in potential. During the J–V protocol the ion distribution lags behind the applied voltage and the bulk electric field changes in size – modulating the extracted current.[2] The inverted scenario of the reverse to forward bias sweep showing ostensibly improved J–V characteristics, has also been measured.[3] The presence of hysteresis is typically seen as a negative, but due to its dependence on the measurement protocol hysteresis can be used to help characterise the cell.

Here we simulate both n-i-p and p-i-n cells that show normal and inverted hysteresis within a single parameter set using drift-diffusion modelling. The key parameters responsible for these effects are demonstrated. The voltage programme for the J–V experiment is shown to be crucial in determining the behaviour seen, and measurement protocols where J–V curves cannot be used to accurately characterise the cell are identified. The J–V scan is described by differences in the fill-factor, open-circuit voltage, and short-circuit current and the underlying mechanisms are explored. The effect of recombination and generation location is investigated – outlining how voltage protocol and illumination wavelength can help determine defect site location