Modeling Metastability in CdTe Solar Cells Due to Cu Migration

Abstract

Thin-film modules of all technologies often suffer from performance degradation over time. Some of the performance changes are reversible and some are not, which makes deployment, testing, and energy-yield prediction more challenging. Manufacturers devote significant empirical efforts to study these phenomena and to improve semiconductor device stability. Still, understanding the underlying reasons of these instabilities remains clouded due to the lack of ability to characterize materials at atomistic levels and the lack of interpretation from the most fundamental material science. The most commonly alleged causes of metastability in CdTe device, such as “migration of Cu,” have been investigated rigorously over the past fifteen years. Still, the discussion often ended prematurely with stating observed correlations between stress conditions and changes in atomic profiles of impurities or CV doping concentration. Multiple hypotheses suggesting degradation of CdTe solar cell devices due to interaction and evolution of point defects and complexes were proposed, and none of them received strong theoretical or experimental confirmation. It should be noted that atomic impurity profiles in CdTe provide very little intelligence on active doping concentrations. The same elements could form different energy states, which could be either donors or acceptors, depending on their position in crystalline lattice. Defects interact with other extrinsic and intrinsic defects; for example, changing the state of an impurity from an interstitial donor to a substitutional acceptor often is accompanied by generation of a compensating intrinsic interstitial donor defect. Moreover, all defects, intrinsic and extrinsic, interact with the electrical potential and free carriers so that charged defects may drift in the electric field and the local electrical potential affects the formation energy of the point defects. Such complexity of interactions in CdTe makes understanding of temporal changes in device performance even more challenging and a closed solution that can treat the entire system and its interactions is required. In this book chapter we first present validation of the tool that is used to analyze Cu migration in single crystal (sx) CdTe bulk. Since the usual diffusion analysis has limited validity, our simulation approach presented here provides more accurate concentration profiles of different Cu defects that lead to better understanding of the limited incorporation and self-compensation mechanisms of Cu in CdTe. Finally, simulations are presented that study Cu ion’s role in light soaking experiments of CdTe solar cells under zero-bias and forward-bias stress conditions.