Unbiased Structure Prediction of Sophisticated Cage Structures

Area of Science:

• Computational Chemistry
• Materials Science
• Supramolecular Chemistry

Background:

• Current cage structure prediction methods often rely on known structures, limiting the discovery of novel candidates.
• Evaluating all potential structural candidates computationally is crucial for designing new materials.

Purpose of the Study:

• To introduce unbiased structure prediction workflows for novel cage discovery using the cgx software.
• To enable computational design and evaluation of all structural candidates, reducing reliance on prior experimental data.

Main Methods:

• Development of unbiased structure prediction workflows within the cgx software.
• Utilizing exploration algorithms and low-cost minimal models for efficient candidate evaluation.
• Validating predictions against experimental data using building block types, features, and stoichiometry as inputs.

Main Results:

• The cgx software successfully predicts cage structures from basic experimental inputs without prior structural knowledge.
• Demonstrated the ability to predict structures prior to costly experimental commitments.
• The approach is automated, open-source, and applicable to various model resolutions.

Conclusions:

• The developed unbiased prediction method facilitates the efficient, automated discovery of novel cage structures.
• The open-source cgx software and provided documentation with code recipes lower the barrier for chemists to adopt this predictive approach.