Researchers at the University of New Hampshire have built an AI-powered database containing 67,573 magnetic compounds, identifying 25 previously unrecognized materials that remain magnetic at high temperatures. The breakthrough, published in Nature Communications, could dramatically reduce the world's dependence on rare-earth elements used in electric vehicle motors, wind turbines, and consumer electronics — over 80% of which are currently processed in China.
The team, led by doctoral student Suman Itani and physics professor Jiadong Zang, trained machine learning models to extract experimental data from thousands of scientific papers and predict when magnetic materials lose their magnetism. The resulting "Northeast Materials Database" is now a searchable resource that could cut years off traditional laboratory testing.
"By accelerating the discovery of sustainable magnetic materials, we can reduce dependence on rare earth elements, lower the cost of electric vehicles, and strengthen U.S. manufacturing capabilities," said Zang. The finding is particularly timely as China recently banned exports of critical rare earths to dozens of countries, intensifying the global race for alternatives.
