High-temperature defect engineering for silicon solar cells : predictive process simulation and synchrotron-based microcharacterization

Abstract

Efficiency is a major lever for cost reduction in crystalline silicon solar cells, which dominate the photovoltaics market but cannot yet compete subsidy-free in most areas. Iron impurities are a key performance-limiting defect present in commercial and precommercial silicon solar cell materials, affecting devices at concentrations below even one part per billion. The lack of process simulation tools that account for the behavior of such impurities hinders efforts at increasing efficiency in commercial materials and slows the time-to-market for novel materials. To address the need for predictive process modeling focused on the impact of impurities, the Impurity-to-Efficiency kinetics simulation tool is developed to predict solar cell efficiency from initial iron contamination levels. The modeling effort focuses on iron because it is known to limit most industrial solar cells. The simulation models phosphorus diffusion, the coupled diffusion and segregation of iron to the high phosphorus concentration emitter, and the dissolution and growth of iron-silicide precipitates. The ID process simulation can be solved in about 1 minute assuming standard processing conditions, allowing for rapid iteration. By wrapping the kinetics simulation tool with a genetic algorithm, global optima in the high-dimensional processing parameter space can be pursued for a given starting metal concentration and distribution. To inform and test the model, synchrotron-based X-ray fluorescence is employed with beam spot sizes less than 200 nm to identify iron-rich precipitates down to 10 nm in radius in industrial and research materials. Experimental X-ray fluorescence data confirm model predictions that iron remains in heavily-contaminated multicrystalline materials after a typical industrial phosphorus diffusion. Similar measurements of the iron-silicide precipitate distribution in multicrystalline silicon samples before and after higher-temperature gettering steps confirm that the higher the process temperature, the larger the reduction in precipitated iron, leading to marked lifetime improvement. By combining the impurity kinetics modeling with the experimental assessment of metal distribution, design guidelines for process improvement are proposed: the high-temperature portion of the process can be designed to enhance dissolution of precipitated iron, while the cooldown from the high-temperature process is crucial to the reduction of the interstitial iron concentration. Finally, while precipitated iron reduction improves with higher temperatures, some regions of multicrystalline silicon samples degrade with higher-temperature gettering steps. To investigate the effect of gettering temperature on the remaining lifetime-limiting defects, spatially-resolved lifetime, interstitial iron concentration, and dislocation density are measured. The detailed defect characterization and analysis provide insight into the limitations of high-temperature phosphorus diffusion gettering.