Toward Low-Symmetry Systems: An Adaptive Differential Evolution Algorithm for Global Structure Searching

The reduction in the symmetry of nanomaterials can produce unexpected properties, while the determination of atomic structures is a sizable challenge in related fields, including low-dimensional materials, surface science, defects, etc. Herein, we develop an adaptive algorithm based on the differential evolution algorithm, which provides benefits for structure searching on low-symmetry systems. The dynamic strategy pool and the island concept are proposed to accelerate the efficiency in the full search space. With several test examples, the designed program not only locates reported structures but also affords new stable configurations that were not located by previous structure search algorithms. Moreover, we provide frameworks and interfaces for stable structure searching on complex systems like grain boundaries, supported clusters, surfaces, and edges. The success in repeatable structure searching with high efficiency demonstrates the reliability and practicability of our algorithm and ensures its potential applications as an advanced technology in many newly arising fields.

Automated summary

In this study, we propose an adaptive algorithm based on the differential evolution algorithm for structure searching in low-symmetry systems. Our algorithm can not only locate known structures but also discover new stable configurations that were previously unknown. Additionally, we provide frameworks and interfaces for stable structure searching in complex systems.