AI unlocks secrets of polycrystalline materials

Scientists at Nagoya University in Japan have used supercomputer technology to discover a new method of detecting tiny imperfections called dislocations in polycrystalline materials. These materials are widely used in electronics, solar panels, and other tech devices, but their effectiveness can be hindered by the presence of dislocations.

Polycrystalline materials are a vital component in many devices we use daily, such as smartphones, computers, and cars. However, because of their complex structures, they are challenging to use effectively. Besides their composition, factors like microstructure, dislocations, and impurities can affect the performance of these materials. One significant issue in using polycrystals is the formation of dislocations caused by stress and temperature changes, which can disrupt the arrangement of atoms and affect performance. It is crucial to understand the formation of these dislocations to prevent failures in devices that use polycrystalline materials.

A team of researchers at Nagoya University, led by Professor Noritaka Usami and including Lecturer Tatsuya Yokoi and Associate Professor Hiroaki Kudo, utilized AI to analyze image data of a commonly used material called polycrystalline silicon, which is used in solar panels. The AI created a 3D model in virtual space, allowing the team to identify areas where clusters of dislocations were affecting the material's performance.

The researchers used electron microscopy and theoretical calculations to analyze dislocation clusters and determine how they formed. They found stress distribution in the crystal lattice and staircase-like structures at the boundaries between crystal grains, which contribute to dislocations during crystal growth. This discovery has implications not only for practical applications but also for the study of crystal growth and deformation. The Haasen-Alexander-Sumino (HAS) model is commonly used to understand dislocation behavior in materials, but the researchers believe that their work uncovered previously unrecognized types of dislocations not accounted for by the HAS model.

Furthermore, the team made another surprising discovery while examining the atomic arrangement of these structures. They found significant tensile bond strains along the edges of the staircase-like formations which triggered the generation of dislocations. Usami, one of the experts on this subject, stated that they were amazed and delighted to finally have evidence of dislocations in these structures. This suggests that by controlling the direction in which boundaries spread, we can also control the formation of dislocation clusters. Through a combination of experiments, theory, and AI, they were able to analyze nanoscale regions in polycrystalline materials and shed light on previously unexplained phenomena. This breakthrough research has paved the way for universal guidelines in creating high-performance materials, with potential impacts beyond solar cells to various fields such as ceramics and semiconductors. Improved performance in polycrystalline materials could have a revolutionary effect as they are widely used in society.