Argyrodite configuration determination for DFT and AIMD calculations using an integrated optimization strategy.

Area of Science:

• Computational Materials Science
• Solid-State Chemistry
• Theoretical Chemistry

Background:

• Accurate modeling of partially occupied structures in density functional theory (DFT) and ab initio molecular dynamics (AIMD) requires careful configuration selection.
• Conventional random sampling methods for identifying low-Coulomb-energy configurations are often inefficient and time-consuming.

Purpose of the Study:

• To develop and evaluate a more efficient strategy for selecting low-Coulomb-energy configurations for materials modeling.
• To improve the accuracy and reduce the computational cost of DFT and AIMD calculations for complex materials.

Main Methods:

• Utilized metaheuristics (genetic algorithm, particle swarm optimization, cuckoo search, harmony search), Bayesian optimization, and modified deep Q-learning.
• Applied these algorithms to search the configurational space of the solid electrolyte Li6PS5Cl.
• Developed an integrated optimization strategy combining these advanced computational techniques.

Main Results:

• Identified ten configuration candidates with relatively low Coulomb energy values for Li6PS5Cl.
• Demonstrated significant computational cost savings compared to traditional random sampling.
• The integrated optimization strategy proved superior to conventional random sampling-based selection.

Conclusions:

• The proposed integrated optimization strategy offers a more efficient and effective approach for selecting configurations in DFT and AIMD.
• This method enhances the reliability of computational materials science studies by improving configuration selection.
• The findings have implications for accelerating materials discovery and design.