A research team has developed a database using artificial intelligence that includes more than 67,000 magnetic materials. Through their analysis, they identified 25 new compounds that retain their magnetic properties at high temperatures. This approach could help reduce dependence on rare earth elements and accelerate the development of electric vehicles and clean energy technologies.

Creating a Database of Magnetic Materials

Experts from the University of New Hampshire applied artificial intelligence to speed up the search for promising magnetic materials. As a result, they created a database containing 67,573 magnetic compounds, among which they discovered 25 previously unidentified materials capable of maintaining magnetic properties at elevated temperatures.

Significance for Industry and Technology

Magnets are essential components in smartphones, medical devices, generators, electric vehicles, and many other systems. Today’s powerful magnets are mainly produced using rare earth elements, which are expensive, imported, and increasingly scarce. Despite the large number of known magnetic compounds, no entirely new permanent magnets had been found among them until now.

Application of Artificial Intelligence

The published study describes the process of creating an AI system capable of analyzing scientific publications and extracting experimental data. This information was used to train computer models that determine the magnetic properties of materials and the temperatures at which they lose these properties. The resulting data was combined into a unified searchable database.

Prospects and Future Development

Testing all possible combinations of elements in laboratory conditions requires significant time and resources. The use of artificial intelligence allows for a faster search for stable alternatives to permanent magnets. The developed database and AI technologies can be applied not only to create new materials but also in education—for example, to convert images into modern text formats and update library collections.

Project Support

The project was carried out with support from the Office of Basic Energy Sciences, the Materials Science and Engineering Division of the U.S. Department of Energy.