From computers to maglev trains, power tools to MRI scanners, rare earth permanent magnets are all around us. Modern life would be difficult without them, and their importance cannot be overstated. However, extracting the rare earth elements that make up rare earth permanent magnets is often labor- and energy-intensive. Scientists have been searching for a better way, and thanks to machine learning algorithms, they may have found one.
Materials Nexus, in collaboration with researchers from the Henry Royce Institute and the University of Sheffield, developed Magnex, a rare earth-free permanent magnet. Magnex is reportedly manufactured at one-fifth the cost of a regular permanent magnet. The new magnet reduces carbon dioxide emissions by 70 percent (kilograms CO2 equivalent).2 Compared to rare earth permanent magnets, the magnetic force per kilogram of material is more than 1.5 times greater.
That’s great in itself, but there’s reason to be excited in a broader sense. The permanent magnets we have were developed in the 1970s and ’80s from alloys of rare earth elements. Finding these materials with the right properties was a long process of trial and error. With MagNex, development from design to testing moved 200 times faster.
“I am really pleased that our initial interaction with Materials Nexus has had such a positive outcome,” Professor Ian Todd FREng, professor of metallurgy and materials processing at the University of Sheffield, said in a statement.
“Materials Nexus’ approach to utilizing AI [artificial intelligence] “Our AI research and development for materials discovery, combined with our world-class facilities for producing advanced alloys at the Henry Royce Laboratory in Sheffield, has enabled us to develop a new magnetic material at incredible speed. This achievement signals a bright future for materials and manufacturing. The next generation of materials unlocked by the power of AI holds huge potential for research, industry and the planet.”
The AI system identifies and analyzes over 100 million potential alloy compositions that have the right properties for permanent magnets, are free of rare earth elements, and meet the requirements for affordability and sustainability. The potential for this approach to yield new materials is enormous.
“This is a groundbreaking discovery using innovative machine learning software, whose development has been made possible with funding from Innovate UK,” said Bruce Adderley, director of Make & Use – Net-Zero at Innovate UK. “Eliminating the need for rare earth elements in high-performance permanent magnets could have a major future impact on our net-zero ambitions through renewable energy and low-carbon transport.”
Computers are already making huge contributions to the discovery of new materials, and thanks to machine learning algorithms, this and similar approaches are now being used in laboratories such as the Self-Driving Lab developed at the University of Toronto.
