Analysis of Buried Interfaces in Photovoltaic Device Stacks

Interfaces are a critical component of modern photovoltaic devices but are challenging to directly probe experimentally. Interfacial components are generally a very low volumetric fraction of an already thin, micron-scale device stack. Interface materials can be comprised of thermally or energetically labile elements such as halogens or pnictides that are easily perturbed by ion beams, mechanical polishing, or electron beams. In this invited contribution, we present examples of problems faced in using electron spectroscopy for the analysis of contacts and interfaces in photovoltaic materials. Solutions to these problems generally have involved novel methods of accessing the buried interface of interest. As a starting point, we discuss an artifact in which focused ion beam (FIB) cross sectioning of a chalcogenide solar cell caused complete exchange of cadmium and copper ions in adjacent sulfide layers. The true structure of this chalcogenide solar cell was ultimately revealed by selective chemical etching and subsequent Auger electron spectroscopy (AES) and X-ray photoelectron spectroscopy (XPS) analysis. Chemical etching is directly compared to a sputter depth profiling in the case of a metal/organic semiconductor interface. The chemical method used to access this buried interface revealed the rich chemical features intrinsic to the organic layer, in contrast to sputter profiling where ion damage erased the variety of chemical environments initially present. Continuing the theme of chemically accessing buried interfaces, a failed attempt to use in-situ atomic hydrogen etching for depth profiling will be discussed. More recently, we have turned to thermomechanical peeling to analyze interfaces in CdTe-based solar cells. This sample preparation method, when combined with AES, XPS, UPS, and low energy inverse photoemission (LEIPS), has revealed several aspects of CdTe interfaces that had previously evaded detection despite likely being present since the inception of CdTe photovoltaic technology. In addition to chemical analysis, we show how thermomechanical peeling can be used in a unique way to measure band alignments in fully completed solar cells, allowing direct comparison between photoemission-determined band offsets and those indirectly measured by transport modeling. Possible future directions in the analysis of buried interfaces will close out the discussion.