Fast Molecular Crystal Structure Prediction Using Sampling by Analogy to Previously Predicted Landscapes.

Area of Science:

• Crystallography
• Computational Chemistry
• Materials Science

Background:

• Accurate molecular crystal structure prediction (CSP) is crucial for materials design.
• Conventional CSP methods often require extensive computational resources and sampling.
• Developing faster and more efficient CSP approaches is an ongoing challenge.

Purpose of the Study:

• To present a proof-of-concept for a fast molecular crystal structure prediction method.
• To utilize analogy to hypothetical crystal structures of related molecules for prediction.
• To reduce computational cost and sampling requirements in CSP.

Main Methods:

• Constructing valid analogues of known crystal structures for related molecules.
• Employing these analogues as trial structures in CSP workflows for new molecules.
• Comparing the analogue-based approach with quasi-random sampling methods.

Main Results:

• The analogue-based method successfully predicted known crystal structures and low-lying energy structures.
• Achieved greater success than quasi-random methods with comparable minimal sampling.
• Generated trial structures closer to local minima, reducing computational cost.

Conclusions:

• The proposed analogue-based approach offers a promising strategy for rapid molecular crystal structure prediction.
• This method has potential applications in high-throughput studies and preliminary testing.
• Demonstrates significant improvements in efficiency and accuracy compared to conventional methods.