Integrating machine learning interatomic potentials with batched optimization for crystal structure prediction

Key Points

• This research aims to enhance crystal structure prediction using machine learning interatomic potentials and optimization techniques.
• Developed a machine learning model for interatomic potentials.
• Employed a batched optimization approach to sample packing landscapes.
• Analyzed energy differences within molecular crystal structures.
• Significant reduction in computational resources needed for crystal structure prediction.
• Improved accuracy in resolving energy differences between competing structures.
• Demonstrated efficiency across various molecular systems.