A thermally driven differential mutation approach for the structural optimization of large atomic systems.

Area of Science:

• Computational materials science
• Condensed matter physics
• Chemical physics

Background:

• Topological amorphous systems present challenges for structural optimization.
• Finding low-lying energy structures is crucial for understanding material properties.
• Multimodal structural optimization requires robust computational methods.

Purpose of the Study:

• To present a novel computational method for obtaining low-lying energy structures of topological amorphous systems.
• To demonstrate the method's effectiveness on amorphous graphene.
• To assess the method's performance with a small population size.

Main Methods:

• Merging a differential mutation genetic algorithm with simulated annealing.
• Incorporating a thermal selection criterion for reliable minima identification.
• Applying the method to unbiased atomic starting configurations of amorphous graphene.

Main Results:

• The method successfully obtained energetically very low structures for amorphous graphene.
• Despite distinct atomic arrangements, the optimized structures exhibited similar properties.
• Key properties analyzed include energy, ring distribution, radial distribution function, coordination number, and bond angle distribution.

Conclusions:

• The presented computational method reliably identifies low-lying energy structures in amorphous systems.
• The approach is efficient, requiring only a small population size.
• The method is suitable for multimodal structural optimization, as demonstrated with amorphous graphene.