Mixed-halide perovskites exhibit ideal band gaps for use in perovskite-based multijunction photovoltaics, but stable performance is compromised by light-induced halide segregation. Photoluminescence (PL) tracking is universally used to monitor such photoinstability; however, here we reveal that such data do not accurately quantify halide segregation. We utilize a combination of simultaneously recorded PL and X-ray diffraction (XRD) measurements to explore CH3NH3Pb(I1–xBrx)3 films across 18 different halide ratios. While PL data suggests that segregation rates increase exponentially with bromide fraction x, XRD patterns reveal that they are actually unchanged. We demonstrate that PL cannot accurately reflect the rate and extent of halide segregation because it is governed by charge funneling to iodide-rich minority domains, which is strongly influenced by additional factors, including luminescence efficiency, band energetics, and charge extraction. To assess the efficacy of treatments to suppress such photoinstabilities, it is therefore essential to probe changes across the full material volume, e.g. by monitoring XRD or absorption spectra.
Lilly et al. (Tue,) studied this question.