The team trained their machine learning model using known magnetic materials, establishing a relationship between electronic and atomic structure features and Curie temperature. They tested the model using compounds based on cerium, zirconium, and iron, focusing on unknown magnet materials based on earth-abundant elements. The model successfully predicted the Curie temperature of material candidates, marking a significant step towards a high-throughput method of designing new permanent magnets for future technological applications.
Key takeaways:
- A team of scientists from Ames National Laboratory has developed a machine learning model to discover new permanent magnet materials that do not require critical elements.
- The model predicts the Curie temperature of new material combinations, an important factor in maintaining magnetism at elevated temperatures.
- The machine learning method can save time and resources in the search for new materials, traditionally an expensive and time-consuming process based on experimentation.
- The team successfully tested the model using compounds based on cerium, zirconium, and iron, marking an important step towards designing new permanent magnets for future technological applications.