GAMaterial-A genetic-algorithm software for material design and discovery.

Area of Science:

• Computational Chemistry
• Materials Science
• Chemical Physics

Background:

• Genetic algorithms (GAs) are powerful stochastic global search methods inspired by biological evolution.
• GAs are extensively applied in chemistry and materials science, often coupled with theoretical methods like DFT and force fields.
• Accurate structural determination of molecules, clusters, nanoparticles, and surfaces is crucial for understanding chemical processes.

Purpose of the Study:

• To introduce GAMaterial, a novel Python-based software implementing genetic algorithms for global structure searches.
• To enable automated structural elucidation of atomic clusters, doped clusters, materials, and surface-supported clusters.
• To integrate machine learning (ML@GA) for accelerating GA searches, particularly for dopant distributions.

Main Methods:

• Development of GAMaterial software in Python3.x for global structure optimization.
• Application of genetic algorithms coupled with density functional theory (DFTB and DFT) calculations.
• Implementation of a machine learning accelerated genetic algorithm (ML@GA) for enhanced population generation.

Main Results:

• GAMaterial successfully performed automated structural searches for Ti6O12, 4Al@Si11, and Na10@graphene systems.
• The ML@GA method demonstrated potential for accelerating GA searches, specifically for dopant distributions in atomic clusters.
• Successful integration with DFTB and deMon2k (DFT) codes for comprehensive structure elucidation.

Conclusions:

• GAMaterial provides a versatile platform for automated, high-throughput structural discovery in materials science.
• The ML@GA approach offers a significant speed-up for complex structure searches, reducing computational cost.
• This work facilitates advancements in catalysis, environmental science, and materials design through efficient structural determination.