Speaker
Description
Crystalline silicon (c-Si) solar cells are now approaching their fundamental performance limits, with voltages now limited primarily by intrinsic recombination processes. We discuss the key remaining bulk and surface defects relevant for high-efficiency commercial c-Si solar cells, as well as strategies for mitigating these defects in next-generation devices, focusing particularly on work at ANU.
Remaining bulk defects include various defects related to oxygen complexes in Czochralski-grown silicon, including thermal donors and oxygen precipitates, as well as residual levels of metallic contamination (especially Fe). The former manifest for example in so-called “ring” defects, and can be mitigated by high-temperature treatments, while the latter can be reduced to trace levels via gettering by heavily doped regions, dielectric passivation layers, or polysilicon contacts. We show that with the aid of such treatments, the bulk lifetime in commercial c-Si wafers now closely approaches the intrinsic Auger limit.
Surface passivation in recent high-efficiency devices has also improved to the point that surface defects no longer significantly limit the cell voltage. Therefore, future innovation in surface and contact passivation for c-Si devices need not improve on current state-of-the-art levels of passivation, but only maintain these while reducing other losses (i.e. transport and optical losses) or realising other advantages such as reduced process cost or improved stability. In particular, degradation of surface passivation due to defect generation under ultraviolet light has re-emerged as a key concern for device stability in the field. We discuss the potential role of alternative passivation species such as chlorine in place of or in addition to hydrogen, and their potentially beneficial properties for stability. We also highlight some emerging potential alternatives to conventional passivation and contact layers that may offer benefits such as improved optical transparency, contact properties or stability.