Unlocking Enhanced Ionic Transport: A Machine Learning-Driven AIMD Study on Doping, Defects, and Strain in...

Area of Science:

• Materials Science
• Computational Chemistry
• Solid-State Physics

Background:

• Anti-perovskite solid-state electrolytes (AP SSEs) are promising for next-generation batteries.
• Understanding the factors influencing lithium-ion transport is crucial for optimizing their performance.

Purpose of the Study:

• To investigate the effect of doping (Cl/Br ratio) and defects on lithium-ion transport in Li3OClxBr1-x AP SSEs.
• To identify optimal compositions and defect structures for enhanced ionic conductivity.

Main Methods:

• Ab initio molecular dynamics (AIMD) simulations were performed on Li3OClxBr1-x structures.
• Analysis included various defects (vacancies, interstitials, Schottky, Frenkel) under biaxial strain.
• Machine learning (ML) and SHAP analysis were applied to predict and interpret lithium diffusivity and ionic conductivity.

Main Results:

• The highest lithium diffusivity was observed for a 0.5/0.5 Cl/Br ratio with double lithium ion interstitials.
• Lithium diffusivity and conductivity are primarily influenced by lithium ion vibration amplitude and concentration.
• Doping and defect engineering showed a more significant impact than biaxial strain.

Conclusions:

• Compositional tuning (Cl/Br ratio) and defect control are key strategies for designing high-performance AP SSEs.
• ML and SHAP analyses provide valuable insights into structure-property relationships for material design.
• This research offers a pathway for developing advanced solid-state electrolytes for energy storage applications.