Universal Neural Network Potential-Driven Molecular Dynamics Study of CO₂/O₂ Evolution at the Ethylene...

Area of Science:

• Materials Science
• Electrochemistry
• Computational Chemistry

Background:

• High-voltage Li-ion batteries require long-term durability and safety.
• Side reactions at the positive electrode/electrolyte interface, like O2 and CO2 release, limit performance.

Purpose of the Study:

• Investigate the reaction mechanisms between positive electrode materials (LiCoO2, LiNiO2) and ethylene carbonate electrolyte.
• Analyze gas evolution during charging at high voltages.

Main Methods:

• Employed universal neural network potential (UNNP)-driven molecular dynamics (MD) simulations.
• Utilized a solid-liquid interface model with approximately 1700 atoms.

Main Results:

• Molecular CO2 and O2 evolved from Li-deintercalated LiNiO2, but not from LiCoO2.
• LiCoO2 demonstrated greater stability against ethylene carbonate decomposition in the charged state compared to LiNiO2.

Conclusions:

• LiCoO2 is a more stable positive electrode material than LiNiO2 for high-voltage Li-ion batteries.
• UNNP-driven MD simulations offer a robust method for understanding electrochemical device side reactions and guiding material selection.