Accelerating the discovery of hidden two-dimensional magnets using machine learning and first principle calculations

Area of Science:

• Condensed Matter Physics
• Materials Science
• Computational Materials Science

Background:

• Two-dimensional (2D) magnets are a frontier in condensed matter physics, offering unique magnetic properties.
• Discovering new 2D magnetic materials is crucial for advancing spintronics and quantum computing.
• Traditional discovery methods are often slow and resource-intensive.

Purpose of the Study:

• To leverage data science and first-principles calculations for accelerated discovery of 2D magnetic materials.
• To identify 2D materials with significant magnetic moments.
• To uncover novel, stable 2D magnetic materials.

Main Methods:

• Utilized machine learning with four descriptors to predict magnetic moments for 216 known 2D materials.
• Trained a machine learning model on existing 2D materials data.
• Performed first-principles calculations to validate predicted materials.

Main Results:

• Identified four key descriptors for predicting magnetic moments in 2D materials.
• Predicted 254 2D materials exhibiting high magnetic moments.
• Revealed eight previously undiscovered, stable 2D materials with high magnetic moments.

Conclusions:

• Data science and materials data analytics offer an effective pathway for surfacing novel materials.
• The combined approach of machine learning and first-principles calculations accelerates the discovery of advanced 2D magnetic materials.
• This work presents an innovative strategy for uncovering hidden materials with desirable properties.