Alan Bowman1 Edoardo Ruggeri1 Miguel Anaya1 Mojtaba Abdi-Jalebi1 Samuel Stranks1

1, University of Cambridge, Cambridge, , United Kingdom

Metal halide perovskites have properties including strong absorption coefficients and long charge diffusion lengths1 (relative to film thickness) making them close to ideal for solar cells. For perovskite devices to achieve their efficiency limits, charge carrier recombination and the role of photon recycling within films must be better understood and controlled.
Current approaches to measure charge carrier recombination rates are typically based on time intensive and expensive transient spectroscopic measurements2. Here we demonstrate that we can employ steady state approaches to extract these parameters, thereby allowing for faster screening of materials. We show the potential of this approach by a comparison with results from transient absorption spectroscopy.
To further understand photon recycling we use our methodology to quantify the photon escape probability. While there are some calculations of the escape probability in the literature3, this quantity has not been previously quantified in metal halide perovskites, despite its importance for solar cell operation including photon recycling. We present measured values and compare with those previously calculated.
Using results from our rapid screening process we calculate the limiting efficiency and number of photon recycling events per initially absorbed photon within devices. We carry out calculations for a range of materials including low bandgap and wider gap mixed-halide perovskites. Our results show that in addition to minimising the charge trapping rate it is necessary to maximise the escape probability for optimal efficiency, thereby reducing the number of photon recycling events. We extend our analysis to two and three bandgap tandem configurations, demonstrating that photon recycling between absorber layers has a relatively small effect on efficiency at the maximum power point. Our combined experimental and modelling approach allows us to target specific device parameters to reach limiting efficiencies.
1. M. A. Green, A. Ho-Baillie and H. J. Snaith, Nat. Photonics, 2014, 8, 506–514.
2. J. M. Richter, M. Abdi-Jalebi, A. Sadhanala, M. Tabachnyk, J. P. H. Rivett, L. M. Pazos-Outón, K. C. Gödel, M. Price, F. Deschler and R. H. Friend, Nat. Commun., 2016, 7, 13941.
3. T. Kirchartz, F. Staub and U. Rau, ACS Energy Lett., 2016, 1, 731–739.