GOAT: A Global Optimization Algorithm for Molecules and Atomic Clusters

Area of Science:

• Computational Chemistry
• Materials Science
• Quantum Chemistry

Background:

• Finding the global energy minimum of atomic and molecular systems is crucial for understanding their properties.
• Traditional methods like molecular dynamics (MD) are computationally expensive, requiring millions of gradient calculations.
• Existing global optimization algorithms may struggle with complex potential energy surfaces (PES).

Purpose of the Study:

• Introduce a novel Global Optimization Algorithm (GOAT) for efficient global energy minima searching.
• Demonstrate GOAT's capability to find global minima without relying on computationally intensive molecular dynamics (MD).
• Validate GOAT's performance against state-of-the-art methods across diverse chemical systems.

Main Methods:

• Development of the Global Optimization Algorithm (GOAT).
• Application of GOAT to organic molecules, water clusters, metal complexes, and metal nanoparticles.
• Comparison of GOAT results with the Conformer-Rotamer Ensemble Sampling Tool (CREST) and other advanced techniques.
• Analysis of GOAT's theoretical underpinnings and success mechanisms in challenging scenarios.

Main Results:

• GOAT successfully identifies global energy minima for various molecular and atomic cluster systems.
• The algorithm demonstrates higher efficiency and accuracy compared to existing global optimization methods.
• GOAT overcomes limitations of other algorithms, particularly in navigating complex potential energy surfaces (PES).
• The method is compatible with various quantum chemical approaches, including hybrid DFT.

Conclusions:

• GOAT provides a robust and efficient alternative to molecular dynamics for global energy minimization.
• The algorithm's flexibility allows its use with diverse quantum chemical methods.
• GOAT represents a significant advancement in computational chemistry for structure prediction and optimization.