Electrical impact assessment of dislocations in silicon materials for solar cells

Abstract

Cast multicrystalline silicon (mc-Si) makes up about 60% of the global photovoltaics market production, and is favored due to its lower areal and capex costs relative to monocrystalline silicon. This method, however, produces material with a higher density of defects (e.g., dislocations, grain boundaries, metal impurities) than more expensive single-crystalline growth methods. A higher density of defects, particularly dislocations, results in a greater density of charge-carrier recombination centers, which reduce a solar cell's efficiency. Interestingly, the recombination activity of individual dislocations and dislocation clusters can vary by orders of magnitude, even within the same device and a separation of only by millimeters of distance. In this thesis, I combine a surface-analysis approach with bulk characterization techniques to explore the underlying root cause of variations in recombination activity between different dislocation clusters. I propose and validate an optical inspection routine based on dislocations' surface characteristics to predict their recombination activity, and extend this methodology to novel growth processes. Lastly, I explore a spatial dispersion metric to assess its potential as a descriptor for the electrical recombination activity of clusters in silicon. This work provides tools to crystal growers and solar cell manufacturers that facilitate the evaluation of electrical performance at early stages of the cell processing, enabling them to reduce the time required for cycles of learning to improve crystal growth processes.