AI-based techniques to predict crystal orientation improve

A team led by researchers from Nagoya University in Japan have successfully predicted crystal orientation by teaching artificial intelligence (AI) using photo optics of polycrystalline materials. The results are published in APL Machine Learning.

Crystals are a vital component of many machines. Familiar materials used in industry contain polycrystalline components, including metal alloys, ceramics, and semiconductors. Since polycrystals are made up of many crystals, they have a complex microstructure, and their properties vary greatly depending on how the crystal grains are oriented. This is especially important for the silicon crystals used in solar cells, smartphones and computers.

“To obtain polycrystalline materials that can be used effectively in industry, it is necessary to control and measure the grain orientation distribution,” said Professor Noritaka Usami. “However, this is hindered by the expensive equipment and time-flow techniques required to sample large areas.”

The Nagoya University team consisting of Professor Usami (he, she) from the Graduate School of Engineering and Professor Hiroaki Kudo (he, she) from the Graduate School of Informatics, in collaboration with RIKEN, have implemented a machine learning model that assesses photos taken by illuminating the surface of a material polycrystalline silicon from various directions. They found that AI succeeded in predicting the grain orientation distribution.

“It took about 1.5 hours for this measurement to take the optical photo, train the machine learning model, and predict orientation, which is much faster than conventional techniques which take around 14 hours,” said Usami. “It also enables wide material measurements that are not possible with conventional methods.”

Usami had high hopes for the use of team techniques in the industry. “This is a technology that will revolutionize materials development,” said Usami. “This research is intended for all researchers and engineers who develop polycrystalline materials. It is possible to create polycrystalline material orientation analysis systems that package image datasets and crystal orientation prediction models based on machine learning. We hope that many companies dealing with polycrystalline materials will install such equipment.”