AI Maps 67,000 Magnets—Rare Earth-Free Future in Sight

A Database That Could Rewire Global Supply Chains

Scientists at the University of New Hampshire have used artificial intelligence to create the most comprehensive searchable database of magnetic materials ever assembled — cataloguing 67,573 compounds and identifying 25 previously unrecognized materials that retain strong magnetic properties even at elevated temperatures. The work, published in Nature Communications, could help sever one of the most strategically dangerous dependencies in modern industry: the reliance on Chinese-controlled rare earth elements to build the magnets that power electric vehicles, wind turbines, and consumer electronics.

How the AI Did It

Rather than testing millions of potential material combinations in the laboratory — a process that would take decades — the UNH team built an AI system that automatically reads scientific literature and extracts experimental data. The system then feeds that information into machine learning models that predict whether a given material is magnetic and estimate its Curie temperature, the threshold above which a magnet loses its properties.

"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 lead author Suman Itani, a physics PhD student at UNH. The resulting resource, called the Northeast Materials Database, is already being used by researchers across the field.

The Rare Earth Problem Is Getting Worse

The urgency behind this research is hard to overstate. China controls roughly 60% of global rare earth mining, 85–90% of processing capacity, and manufactures over 90% of the high-performance sintered permanent magnets used in cars, wind turbines, and defense systems. A typical electric vehicle motor contains 1–2 kilograms of neodymium and praseodymium — elements overwhelmingly processed in China.

In April 2025, Beijing tightened its grip further, introducing export controls on seven heavy rare earth elements and all related compounds. European rare earth prices subsequently surged to up to six times Chinese domestic prices, and some carmakers were forced to temporarily halt production lines. According to the Center for Strategic and International Studies, the controls also threaten U.S. defense supply chains.

A Parallel Breakthrough From Georgetown

The UNH database is not the only front in the rare-earth-free race. Researchers at Georgetown University separately synthesized a new class of magnets based on high-entropy borides — combining earth-abundant transition metals and boron in a largely unexplored C16 crystal structure. The resulting materials achieve magnetic anisotropy approaching that of rare-earth permanent magnets, without a single atom of neodymium or dysprosium. Their approach, detailed by Phys.org, also eliminates precious metals entirely.

From Lab Bench to Factory Floor

Both breakthroughs still face the same steep climb: laboratory results must survive industrial scaling. Magnet performance often degrades when compounds are produced in bulk, and manufacturing processes for entirely new material classes take years to optimize. Professor Jiadong Zang at UNH acknowledged the challenge but expressed confidence: the team is "optimistic that our experimental database and growing AI technologies will make this goal achievable."

If even a handful of the 25 flagged compounds prove viable at scale, the implications are significant. The clean energy transition currently depends on a supply chain concentrated in a single country — one that has already demonstrated willingness to use that leverage. A credible rare-earth-free alternative would not just cut costs; it would fundamentally change the geopolitics of the green economy.