Exploration and Optimization in Crystal Structure Prediction: Combining Basin Hopping with Quasi-Random Sampling.

Area of Science:

• Crystallography
• Computational Chemistry
• Materials Science

Background:

• Predicting molecular crystal structures is essential for understanding material properties.
• Global optimization methods are key to exploring the complex energy landscapes of crystal structures.
• Existing methods like quasi-random sampling can be inefficient for locating low-energy polymorphs.

Purpose of the Study:

• To implement and evaluate a hybrid Monte Carlo basin hopping (BH) and quasi-random (QR) structure generation method (QR-BH) for molecular crystal structure prediction.
• To assess the efficiency of QR-BH in locating low-energy crystal structures compared to pure QR sampling.
• To demonstrate the utility of QR-BH in identifying important polymorphs.

Main Methods:

• Implementation of a Monte Carlo basin hopping global optimization algorithm.
• Integration of quasi-random (QR) sequences for initial structure generation.
• Hybridization of QR sampling with BH optimization to create the QR-BH method.
• Testing QR-BH on single-component molecular crystals and co-crystals.

Main Results:

• The QR-BH method significantly accelerates the location of low-energy molecular crystal structures compared to pure QR sampling.
• QR-BH maintains the efficiency of finding higher-energy structures, which are critical for polymorph identification.
• Successful application demonstrated on both single-component and co-crystal systems.

Conclusions:

• The QR-BH method offers an efficient and effective approach for molecular crystal structure prediction.
• This hybrid strategy enhances the discovery of relevant polymorphs by combining broad sampling with focused optimization.
• QR-BH represents a valuable advancement in computational materials science for crystal engineering.