Yb-Doped Double Halide Perovskite Cs₂AgBiCl₆ for Lead-Free High-Efficiency Downconversion

The mismatch between the solar spectrum and band gap in silicon solar cells limits the light to electrical energy conversion to 33.6%, the Shockley-Queisser (SQ) limit. One approach to surpassing this limit is shifting the solar spectrum entering the silicon solar cell to better match the silicon’s band gap by downconverting ultraviolet (UV) and blue photons to near-infrared (NIR) photons. In some materials, downconversion can create two NIR photons from each UV/blue photon, a phenomenon called quantum cutting. Quantum cutting can eliminate significant energy loss due to mismatch and raise the SQ limit to 41%, with prospects to translate this to record solar module efficiencies and improve solar module lifetime. Yb-doped halide perovskite CsPbCl_3-xBr_x(x<1) has been shown to absorb nearly all light above the perovskite’s bandgap and downconvert it to NIR at 1.25 eV close to the silicon’s bandgap energy via quantum cutting with nearly 200% photoluminescence quantum yield (PLQY). However, lead is toxic, and thus, there is a need to find alternatives. We have been exploring bismuth and silver-based halide double perovskite Cs₂AgBiCl₆ to fulfill this need. Previous studies on Cs₂AgBiCl₆arrived at conflicting conclusions regarding its bandgap and the origin of its characteristic visible orange emission. Here, we address these knowledge gaps and report on undoped and Yb-doped Cs₂AgBiCl₆thin films synthesized through reactive thermal evaporation. Specifically, CsCl, BiCl₃, and AgCl precursor powders are co-evaporated onto glass substrates and subsequently annealed (200-250 ^oC) either in air or in a nitrogen-filled glovebox, resulting in 400 nm thick polycrystalline thin films. Films are doped by adding YbCl₃ to the co-evaporation. Our optical characterization analysis of the undoped films reveals an indirect bandgap of 2.77 eV and a direct transition at 2.9 eV. X-ray diffraction (XRD) and Raman spectroscopy were used to verify the phase purity and structure of the perovskite. Like in previous work, we observed an orange emission at around 650 nm. We explore the origin of this emission using time-resolved photoluminescence and lifetime measurements coupled with a physically based kinetic mathematical model of exciton decay channels that considers various recombination processes, including radiative and nonradiative recombination on defects and self-trapped excitons. With only a few adjustable physically-based time constants, photoluminescence lifetime data could be fit with high fidelity. Our analysis suggests that the orange emission comprises both self-trapped exciton and radiative defect emissions with fractional contribution depending on the details of the material synthesis conditions. We also doped the Cs₂AgBiCl₆perovskite with YbCl₃ to explore the possibility of quantum cutting. The perovskite host absorbs the ultraviolet energy and transfers it to Yb, which then relaxes and emits photons in the NIR region. Doping with Yb resulted in PLQY of 50%, the highest reported in the literature for this material. We investigated various post-deposition treatments, such as annealing in air versus in a nitrogen-filled environment, and concluded that annealing in air, in the presence of moisture, results in the formation of bismuth oxychloride (BiOCl), confirmed by XRD and Raman spectroscopy. While these results suggest energy transfer to Yb could be efficient in Cs₂AgBiCl₆, quantum cutting, and PLQY exceeding 100% remains to be achieved by fine-tuning the synthesis conditions.