Tackling complexity in crystal structure determination of metal organic compounds using machine learning interatomic...

Area of Science:

• Materials Science
• Computational Chemistry
• Crystallography

Background:

• Determining crystal structures of metal-organic compounds is crucial for materials development.
• Predicting these structures computationally presents significant challenges.
• Existing methods often struggle with complex or previously unknown structures.

Purpose of the Study:

• To develop and apply an integrated computational framework for predicting challenging metal-organic compound structures.
• To resolve the unknown crystal structures of specific technologically relevant compounds.
• To accelerate the pace of materials discovery and development.

Main Methods:

• Integration of ab initio crystal structure prediction techniques.
• Application of universal machine learning interatomic potentials (UMA and Orb-v3).
• Systematic structure determination for target metal-organic compounds.

Main Results:

• Successfully predicted the crystal structures of lithium phenolate.
• Resolved the previously unknown structure of sodium cyclohexanolate.
• Determined the crystal structure of lithium-benzimidazol-2-one.
• Demonstrated the framework's capability in handling complex metal-organic systems.

Conclusions:

• The integrated computational approach effectively predicts challenging metal-organic compound structures.
• This framework significantly accelerates materials development by resolving unknown structures.
• The methodology provides a powerful tool for discovering novel functional materials.