Illuminating the Potential of Thin-Film Photovoltaics

Abstract

Widespread adoption of photovoltaics (PV) as an alternative electricity source will be predicated upon improvements in price performance compared to traditional power sources. Solution processing of thin-film PV is one promising way to reduce the capital expenditure (CAPEX) of manufacturing solar cells. However, it is imperative that a shift to solution processing does not come at the expense of device performance. One particularly problematic parameter for thin-film PV has historically been the open-circuit voltage (VOC). As such, there is a pressing need for characterization tools that allow us to quickly and accurately evaluate the potential performance of solution-processed PV absorber layers. This work describes recent progress in developing photoluminescence (PL) techniques for probing optoelectronic quality in semiconductors. We present a generalized model of absorption that encompasses ideal direct-gap semiconductor absorption and various band tail models. This powerful absorption model is used to fit absolute intensity PL data and extract quasi-Fermi level splitting (maximum attainable VOC) for a variety of PV absorber technologies. This technique obviates the need for full device fabrication to get feedback on optoelectronic quality of PV absorber layers and has expedited materials exploration. We then use this absorption model to evaluate the thermodynamic losses due to different band tail cases and estimate tail losses in Cu2ZnSn(S,Se)4 (CZTSSe). The effect of sub-bandgap absorption on PL quantum yield (PLQY) and voltage is elucidated, and new analysis techniques for extracting VOC from PLQY are validated that reduce computation time and provide us even faster feedback on material quality. We then use PL imaging to develop a mechanism describing the degradation of solution-processed CH3NH3PbI3 films under applied bias and illumination.