Publication date: 17th February 2025
Lead halide perovskite-based photovoltaics has attracted great attention in the last decade since silicon photovoltaic as the dominating technology is approaching its theoretical efficiency limit. Combined in a monolithic perovskite/Si tandem solar cell, they promise a substantial leap in power conversion efficiency beyond the single-junction limit without increasing cost significantly. However, realizing high performance in module-sized formats is limited by the practical challenge of producing uniform high-quality perovskite layers onto large-scale silicon bottom cells. The polycrystalline and mechanically soft nature of perovskite thin films lead to a high defect concentration and inefficient carrier extraction if the crystallization is not properly controlled.
In pursuit of a non-invasive and robust in-line imaging method that locally resolves the quality of the perovskite thin film absorber processed over a silicon bottom solar cell, we advanced a method, called k-imaging, first developed by Hacene et al. for perovskite single junction solar cells.[1] K-imaging is based on intensity-dependent perovskite photoluminescence, eliminating the effect of non-uniform optical in- and out-coupling which regularly complicates the analysis of single-intensity photoluminescence imaging. With a basic power law model a single effective parameter k is extracted which reflects the complex superposition of competing recombination processes, i.e. radiative band-to-band, non-radiative trap-assisted bulk (Shockley-Read-Hall) and interface recombination. This parameter k allows for quantitative analysis simplifying the interpretation in contrast to qualitative PL imaging, while barely raising setup complexity and cost.
We show that k-imaging reproducibly identifies general quality discrepancies as well as local inhomogeneities, which clearly correlate with typical defects in the thin film. Further, we proved its applicability for several relevant perovskite processing techniques and its compatibility with flat as well as industrially relevant textured silicon bottom cells. Overall, k-imaging is a valuable and unique technique meeting the requirements for efficient optimization in academic research as well as quality assessment in large-scale industrialized production of monolithic perovskite/Si tandem solar cells.