Surface potential variation across (hk1) and non-(hk1) grain boundaries of antimony triselenide

Abstract

Electronic materials that can tolerate defects or create self-passivated interfaces expand the horizons to more sustainable electronic devices. Two heuristic models suggest that (i) antimony triselenide is a defect tolerant semiconductor (to certain point defects) and that (ii) the boundaries of grains with a specific orientation are self-passivated. Yet, experimental support or scrutiny of these models is lacking. Partly because measuring a passive quality of a material is perplexing. A possible way forward is to measure defect activity in correlation with other experimental evidence for defect presence, for example, by relating the grain boundaries’ (GBs) nanostructure and electronic quality. One measure of the electronic quality is the magnitude of the surface potential variation across grain boundaries—a high (charged) defect concentration and low defect tolerance typically leads to a significant surface potential variation. On the other hand, direct observation of defects (e.g., by electron microscopy) and a small surface potential variation indicates that the material is defect tolerant. In parallel, a comparison of the surface potential variation across grain boundaries in films with grains of different orientations provides a qualitative measure of the self-passivation of material because only certain crystallographic surfaces are expected to be self-passivated. Herein we report that regardless of the grains’ mutual respective orientation, the grain boundaries of antimony triselenide have a relatively benign electronic quality. This is evident in a shallow surface potential variation across grain boundaries (10–50 mV; 195 GBs were investigated). Yet, a high-resolution TEM investigation reveals that a variety of defects is prevalent at and near the GBs. Additionally, we found that GBs in films with distinct (hk1) orientation indeed have a lower surface potential variation than grains with (hk0) orientation. These findings support both heuristic models and provide a practical way to scrutinize defect tolerance in other polycrystalline semiconductors.