Design of a Growth Process of High-Quality Quasi-Monocrystalline Silicon Ingot Integrating Experimental, Theoretical, Computational, and Data Sciences
| Autor: | Usami, N., Liu, X., Fukuda, Y., Tanaka, H., Kutsukake, K., Kojima, T. |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2022 |
| Předmět: | |
| DOI: | 10.4229/wcpec-82022-1ao.4.3 |
| Popis: | 8th World Conference on Photovoltaic Energy Conversion; 11-13 We report on a methodology to design a growth process of high-quality quasi-monocrystalline silicon ingot integrating experimental, theoretical, computational, and data sciences. First, the model of crystal growth simulation was validated by utilizing experimental data, and the resulting model was used to collect training data showing the relationship between multi-dimensional process parameters and the dislocation density distribution after growth. A machine learning model of the simulation was created to instantly predict the dislocation density from the process parameters, allowing the application of an optimization algorithm. As a result, efficient design of the optimal growth process was possible. Furthermore, by using seed crystals with artificial grain boundaries and their spontaneous change during the growth, we have clarified the effect of the grain boundary structure on the direction of grain boundary growth and the carrier recombination velocity at the grain boundary. These permits to design of an appropriate seed crystal and process, which will result in the growth of high-quality quasi-monocrystalline silicon ingot. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|