Characterization of a multicrystalline silicon ingot using a data science approach
Noritaka USAMI1, Yusuke HAYAMA1
We review our recent attempt to integrate data science with experimental science to establish universal guidelines to improve multicrystalline materials with complicated microstructures. Based on data collection from a large quantity of practical multicrystalline silicon wafers for solar cells and image processing, we succeeded in realizing a three-dimensional (3D) visualization of the microstructures and dislocation clusters in a multicrystalline silicon ingot. This manifested generation, propagation, and annihilation of dislocation clusters in the multicrystalline silicon ingot. The combination of data science and experimental approaches showed that small-angle boundaries are likely to be the source of dislocation clusters in multicrystalline silicon.
- 1 Graduate School of Engineering, Nagoya University