Researchers at Ames National Laboratory are advancing the discovery of materials for rare-earth-free permanent magnets by combining fundamental physics with artificial intelligence. The work contributes to the U. S. Department of Energy's (DOE) Genesis Mission, which unites DOE National Labs, industry, and academia to harness artificial intelligence (AI) for scientific breakthroughs in energy, discovery science, and national security, including a focus on securing America's supply of rare earths and other critical minerals.
Rare earth elements give today's high-performance magnets their exceptional strength and resistance to de-magnetization, properties essential for many technologies and particularly for energy generation and defense. But they can be quite expensive and often require dependence on unstable sources outside the United States. Reducing or eliminating the need for rare earth elements will lower costs and enhance domestic production of high-performance magnets.
For more than 20 years, researchers have been looking for permanent magnet materials that do not use rare earth elements. Investigations have largely been material-focused, with researchers relying on a trial-and-error approach that involves synthesizing materials in the lab, testing them, and then using the data to build knowledge one data point at a time.
In a recently published article, Ames Lab Scientist Prashant Singh outlines a faster, more systematic path by combining physics-based modeling, high-throughput simulations, and reasoning-based AI tools to guide discovery before materials are made in the lab.
This approach focuses on understanding how a material's atomic structure and electronic behavior to determine key magnetic properties such as magnetization strength, energy storage capacity, resistance to demagnetization, and performance at elevated temperatures. By embedding that physics knowledge into computational models, researchers can more efficiently identify promising material candidates and reduce the need for costly experimental iteration.
Ames Lab researchers are well-positioned as leaders in addressing this challenge.
"Ames Lab's strength comes from its deep expertise and a long history of data in the magnet space that no other institution has," Singh said. "That's what makes our role critical. In any material design problem, you need to know how combining two elements will change their performance before you ever run an experiment. We have been building both theoretical and analytical tools to answer that question, and we are now bringing AI into that process to make it faster and broader in scope."
Building on this foundation, the team is developing new AI-assisted ways for researchers to interact with these models more directly, allowing them to pose design questions, refine requirements, and explore potential materials more efficiently. These emerging capabilities draw on earlier work such as Singh's recently developed tool DuctGPT, an agentic AI designed for interactive exploration and materials design.
The challenge to making these approaches effective is ensuring AI models are trained in the right kind of data. Rather than relying on generalized data, models need to be trained on experimentally measured and scientifically calculated material properties to enable predictions that remain grounded in real-world behavior.
"Understanding the physics of materials is important to include in AI frameworks when you are trying to design new materials. If you just use the data to train your models, you are going to get only the predictions within the range of information you have," Singh said. "But once you understand the physics of what controls specific properties, then you and your agentic tools or AI frameworks can search arbitrary material space."
These AI tools offer an important addition benefit: they can take into consideration material availability and costs.
"Supply chain conditions shift by the hour, material costs fluctuate, availability changes daily, and the market never stands still," Singh explained. "By factoring those conditions into the discovery process, we can better target materials that are not only high-performing but also practical to produce and scale."
This roadmap addresses the full critical materials pipeline from accelerating discovery to ensuring industrial viability could reduce America's reliance on outside sources for critical materials and build a more resilient domestic supply chain. By combining Ames National Laboratory's long-standing strengths in theory, simulation, and proprietary magnetic materials data with emerging AI capabilities, researchers hope to dramatically expand the pace and scope of magnetic materials innovation.
"What sets Ames apart is seven decades of work in critical materials-a depth of knowledge no other lab can match," said Singh. "We have strong foundations in theory and simulation, and proprietary magnetic material data exclusive to Ames. Adding AI to that base lets us explore new materials and magnetic possibilities faster and directly address the supply chain vulnerabilities the country faces."