PGA: A new particle swarm optimization algorithm based on genetic operators for the global optimization of clusters

Area of Science:

• Computational chemistry
• Materials science
• Quantum mechanics

Background:

• Determining the ground-state structures of atomic clusters is crucial for understanding their properties.
• Traditional methods often struggle with the complexity and vast search space of cluster structures.

Purpose of the Study:

• To develop and validate a novel global optimization program, PGA, for identifying the lowest-energy structures of atomic clusters.
• To assess the program's efficiency and accuracy by comparing simulated and experimental data.

Main Methods:

• Developed the PGA (Particle Swarm Optimization with Genetic Operators) program.
• Applied PGA to known systems like Au20 and B20.
• Used photoelectron spectroscopy (PES) for validation.
• Searched for global minima of silicon clusters (Sin, n=3-30).

Main Results:

• Successfully identified known structures of Au20 and B20.
• Validated PGA by matching simulated and experimental PES for ground-state clusters.
• Discovered new structures for Sin clusters (n=6, 7, 12, 14).
• Determined structures for medium-sized Sin clusters (n=21-30) for the first time.
• Explored structural evolution and electronic properties of Sin clusters.

Conclusions:

• PGA is an effective and efficient tool for global structure optimization of atomic clusters.
• The program shows promising potential for exploring global minima in other cluster systems.
• The code is freely available for further research.