Defects and degradation are important issues in production and commercialization of low cost photovoltaics. Defects cause intermediate product rejections during quality assurance, while degradation cause shortened device lifetimes. Laterally resolving imaging techniques have already proven their ability for resolving the location of such performance decreasing problems.
A one to one correlation between physical defect types and imaging measurement features cannot be established on a single imaging method alone. However, a combination of several measurement methods is suitable to distinguish different defect types. In this work we classify defects by their electrical properties and correlate them with three different imaging techniques by simulation and experiment. The imaging techniques investigated in this study are Electroluminescence Imaging (ELI), Light-Beam Induced Current mapping (LBIC) and Dark Lock-In Thermography (DLIT). Our software is based on electrical simulations of a resistive network of diodes. The locally resolved current and voltage distributions are used for computing.
Using this systematic approach, we simulate a variety of defect types and correlate their electrical properties to specific fingerprints of the three imaging methods. We confirm our findings by comparing with experiments with artificially induced defects and observe good agreement of the measured and simulated defect patterns. In addition to resolving the defect origin, we analyze the efficiency drop compared to the defect free model.