How AI is revolutionizing material discovery

Rate this post

Metal Magnet Materials Science Art Concept

Ames National Laboratory scientists have developed a machine learning model to predict new magnet materials without using rare elements. This innovative approach focusing on the Curie temperature of the material provides a more sustainable path for future technological applications.

Scientists use AI to discover new magnetic materials without critical components.

A team of researchers at the US Department of Energy’s Ames National Laboratory developed a new Machine learning A model for finding critical-element-free permanent magnet materials. The model predicts the Curie temperature of new material combinations. Using artificial intelligence to predict new permanent magnet materials is an important first step. The model adds to the team’s recently developed ability to detect thermodynamically stable rare earth materials.

The importance of high-performance magnets

Technologies such as wind power, data storage, electric vehicles and magnetic refrigeration require high-performance magnets. These magnets contain critical elements such as cobalt and rare earth elements such as neodymium and dysprosium. These materials are in high demand but their availability is limited. This situation is motivating researchers to find ways to design new magnetic materials with less critical materials.

Magnet puck

Photo of the magnet. Credit: US Department of Energy Ames National Laboratory

The role of machine learning

Machine learning (ML) is a form of artificial intelligence. It is driven by computer algorithms that continuously improve its predictions using data and trial-and-error algorithms. The team used experimental data on the Curie temperature and theoretical modeling to train the ML algorithm. The Curie temperature is the maximum temperature at which a material retains its magnetism.

“Finding compounds with a high Curie temperature is an important first step in the search for materials that can retain magnetic properties at elevated temperatures,” said Yaroslav Mudrik, an Ames lab scientist and senior leader of the research team. “This aspect is important for the design not only of permanent magnets but also of other functional magnetic materials.”

According to Mudrick, discovering new materials is a challenging activity because discovery has traditionally been based on experiments, which are expensive and time-consuming. However, using the ML method can save time and resources.

Developing the model

Prashant Singh, an AIIMS lab scientist and member of the research team, explained that developing ML models using basic science was a key part of this effort. The team trained their ML model using experimentally known magnetic materials. Information on these materials establishes a relationship between several electronic and atomic structure features and the Curie temperature. These patterns provide a basis for the computer to search for potential candidate materials.

Model testing and validation

To validate the model, the team used compounds based on cerium, zirconium and iron. Andriy Palasyuk, an Ames Lab scientist and member of the research team, came up with the idea. He wanted to focus on unknown magnet materials based on Earth-abundant elements. “The next super magnet must not only excel in performance, but also rely on abundant domestic elements,” Palasyuk said.

Palasyuk worked with Tyler Del Rose, another scientist at the Ames lab and a member of the research team, to synthesize and characterize the alloys. They found that the ML model was successful in predicting the Curie temperature of the material candidates. This breakthrough is an important first step in creating a high-throughput way to design new permanent magnets for future technological applications.

“We are writing machine learning that informs physics for a sustainable future,” Singh said.

Reference: Prashant Singh, Tyler Del Rose, Andriy Palasyuk, and Yaroslav Mudrik, 2 August 2023, “Physics-informed machine-learning prediction of the Curie temperature and its promise to guide the discovery of functional magnetic materials.” Chemistry of Materials.
DOI: 10.1021/acs.chemmater.3c00892

Leave a Comment