Researchers at the University of New Hampshire built an AI system that reads scientific papers, extracts experimental data on magnetic materials, and organizes it into a searchable database. The result: 67,573 magnetic compounds catalogued, with 25 previously unrecognized materials that stay magnetic at high temperatures - potential alternatives to the rare earth magnets that power everything from electric vehicles to smartphones.
The research, published in Nature Communications, addresses a supply chain problem that’s only getting worse. Rare earth elements are expensive, difficult to obtain, and concentrated in a few countries. Finding alternatives that work at the high temperatures inside motors and generators has been a decades-long challenge.
How the AI Found What Humans Missed
Testing every possible combination of elements for magnetic properties would take too long and cost too much. The University of New Hampshire team took a different approach: train AI to mine existing scientific literature for data that researchers had already collected but never connected.
The system works in two stages. First, it reads published papers and extracts key experimental details about magnetic materials. Then predictive models determine whether materials are magnetic and calculate the temperatures at which they lose their magnetism.
“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, the study’s lead author and a doctoral student in physics.
The 25 Materials
The team identified 25 compounds that had been studied but never recognized as potentially useful permanent magnet candidates. These materials maintain their magnetism at the elevated temperatures found inside electric motors - a critical requirement that eliminates most alternatives.
The researchers note this represents the first new permanent magnet discovery from existing research in years. Instead of synthesizing and testing new materials from scratch, the AI found overlooked candidates hiding in plain sight across thousands of published studies.
What This Means for Clean Energy
Permanent magnets are everywhere: wind turbines, electric vehicle motors, medical imaging equipment, consumer electronics. The current options rely heavily on neodymium and other rare earth elements, creating supply vulnerabilities and environmental concerns from mining operations.
The Northeast Materials Database, as the team calls their resource, is designed for other researchers to use. By making the data searchable and accessible, the project could accelerate materials science research beyond the specific compounds the original team identified.
The work was funded by the U.S. Department of Energy’s Office of Basic Energy Sciences - an investment in reducing American dependence on imported critical materials while supporting the transition to electric vehicles and renewable energy.