Researchers from Tokyo University of Science Developed a Deep Learning Model that can Detect a Previously Unknown Quasicrystalline Phase in Materials Science

The quest to uncover novel crystalline structures in materials has long been a cornerstone of scientific exploration, holding critical implications across diverse industries ranging from electronics to pharmaceuticals. Crystalline materials, defined by their ordered atomic arrangements, play an important role in technological advancements. Identifying and characterizing these structures accurately has conventionally relied on methods like powder X-ray diffraction. However, the emergence of multiphase samples with intricate mixtures of different crystalline structures has posed challenges for precise identification.

Addressing this challenge, a study by researchers from Tokyo University of Science (TUS), Japan, in collaboration with esteemed institutions, introduced a new deep learning model. The research outlines the development of a machine learning-based binary classifier capable of detecting an elusive icosahedral quasicrystal (i-QC) phase from multiphase powder X-ray diffraction patterns.

The researchers constructed a binary classifier employing 80 convolutional neural networks. They trained this model using synthetic multiphase X-ray diffraction patterns designed to simulate anticipated i-QC phase patterns. Following rigorous training, the model exhibited remarkable performance, boasting an accuracy exceeding 92%. It effectively detected an unknown i-QC phase within multiphase Al-Si-Ru alloys, confirming its prowess in analyzing 440 measured diffraction patterns from diverse unknown materials across six alloy systems.

Remarkably, the model’s capability extended beyond detecting predominant components, successfully identifying the elusive i-QC phase even when it wasn’t the primary constituent in the mixture. Additionally, its potential spans beyond i-QC phases, hinting at applicability in identifying new decagonal and dodecagonal quasicrystals and various crystalline materials.

The model showcases an accuracy that promises to expedite the identification process of multiphase samples. This breakthrough, bolstered by the model’s success, is poised to revolutionize materials science by expediting phase identification, which is crucial in mesoporous silica, minerals, alloys, and liquid crystals.

The impact of this study transcends the mere identification of quasicrystalline phases; it introduces a paradigm shift in material analysis. Its potential applications in diverse industrial sectors, from optimizing energy storage to advancing electronics, hold promise for transformative technological advancements.

This research signifies a remarkable stride toward unveiling new phases within quasicrystals, empowering scientists to navigate uncharted territories in material science. The team’s pioneering work enriches our understanding of crystalline structures and heralds a new era of accelerated discovery and innovation in materials science.


Check out the Paper and BlogAll credit for this research goes to the researchers of this project. Also, don’t forget to join our 33k+ ML SubReddit, 41k+ Facebook Community, Discord Channel, and Email Newsletter, where we share the latest AI research news, cool AI projects, and more.

If you like our work, you will love our newsletter..

Niharika is a Technical consulting intern at Marktechpost. She is a third year undergraduate, currently pursuing her B.Tech from Indian Institute of Technology(IIT), Kharagpur. She is a highly enthusiastic individual with a keen interest in Machine learning, Data science and AI and an avid reader of the latest developments in these fields.

🚀 LLMWare Launches SLIMs: Small Specialized Function-Calling Models for Multi-Step Automation [Check out all the models]