Despite rapid progress in halide perovskite solar cells, a comprehensive understanding of recombination processes and agreement about the most appropriate models to explain recombination and interpret experimental data are still lacking yet will be critical to further performance and stability improvements. Here, we analyze the evolution of recombination models in halide perovskites, highlighting how the background of many researchers in dye-sensitized and organic photovoltaics triggered an early emphasis on the inclusion of excitons followed by a phase of simplifying the mathematical description until experiment and theory were clearly becoming inconsistent with each other. Recent years have seen the trend of going back to the original models for classical semiconductors. These approaches are more frequently combined with sophisticated approaches for fitting and quantifying confidence in data analysis that make increasing use of machine learning algorithms. We outline the need for global recombination models and the challenges and opportunities in attaining this.
Stranks et al. (Tue,) studied this question.