Scientists at the University of New Hampshire say artificial intelligence is dramatically speeding up the hunt for advanced magnetic materials. Their new resource, containing 67,573 magnetic compounds, even flagged 25 materials that had never been recognized as magnets capable of staying magnetic at high temperatures.
“By accelerating the discovery of sustainable magnetic materials, we can reduce dependence on rare earth elements, lower the cost of electric vehicles and renewable-energy systems, and strengthen the U.S. manufacturing base,” said Suman Itani, lead author and a doctoral student in physics.
The Northeast Materials Database, unveiled in a study published in Nature Communications, gives researchers a powerful tool for exploring materials at the heart of modern technology. Magnets make smartphones vibrate, power generators spin, and electric vehicle motors run. But the strongest magnets today rely on rare earth elements that are expensive, imported, and increasingly difficult to secure. Even with thousands of known magnetic compounds, researchers still haven’t identified a brand new permanent magnet from that pool.
The team’s AI system was designed to sift through scientific papers and automatically extract experimental data. That information trained computer models to determine whether a material is magnetic and to predict the temperature at which it loses its magnetism. The results were then organized into one searchable database.
The challenge has always been sheer volume. Millions of potential element combinations exist, far too many for researchers to test in a lab. AI compresses years of work into weeks.
“We are tackling one of the most difficult challenges in materials science, discovering sustainable alternatives to permanent magnets, and we are optimistic that our experimental database and growing AI technologies will make this goal achievable,” said physics professor Jiadong Zang, a co-author of the study.
Co-author Yibo Zhang, a postdoctoral researcher in physics and chemistry, said the same large language model behind the database could play bigger roles in science and education. One example is converting old images into rich text formats to help preserve library archives.
The project was supported by the Office of Basic Energy Sciences, Division of Materials Sciences and Engineering at the U.S. Department of Energy.
