Scanning Tunneling Spectroscopy Method for the Prediction of Semiconductor Heterojunction Performance as a Prequel for Device Development.

ABSTRACT

The prediction of semiconductor device performance is a persistent challenge in materials science, and the ability to anticipate useful specifications prior to construction is crucial for enhancing the overall efficiency. In this study, we investigate the constituents of a solar cell by employing scanning tunneling microscopy (STM) and spectroscopy (STS). Through our observations, we identify a spatial distribution of the dopant type in thin films of materials that were designed to present major p-doping for germanium sulfide (GeS) and dominant n-doping for tin disulfide (SnS2). By generating separate STS maps for each semiconductor film and conducting a statistical analysis of the gap and doping distribution, we determine intrinsic limitations for the solar cell efficiency that must be understood prior to processing. Subsequently, we fabricate a solar cell utilizing these materials (GeS and SnS2) via vapor phase deposition and carry out a characterization using standard J-V curves under both dark/illuminated irradiance conditions. Our devices corroborate the expected reduced efficiency due to doping fluctuation but exhibit stable photocurrent responses. As originally planned, quantum efficiency measurements reveal that the peak efficiency of our solar cell coincides with the range where the standard silicon solar cells sharply decline. Our STS method is suggested as a prequel to device development in novel material junctions or deposition processes where fluctuations of doping levels are retrieved due to intrinsic material characteristics such as the occurrence of defects, roughness, local chemical segregation, and faceting or step bunching.