Uni-Electrolyte: An Artificial Intelligence Platform for Designing Electrolyte Molecules for Rechargeable Batteries

Area of Science:

• Materials Science
• Computational Chemistry
• Electrochemistry

Background:

• Electrolyte innovation is critical for advancing rechargeable battery technology, particularly lithium batteries.
• The vast molecular space (>10^60) and complex solution chemistry present significant challenges for traditional electrolyte design.

Purpose of the Study:

• To introduce Uni-Electrolyte, an artificial intelligence (AI) platform designed for the rapid discovery and design of advanced electrolyte molecules.
• To overcome the limitations of conventional methods in exploring the extensive chemical space for novel electrolytes.

Main Methods:

• The Uni-Electrolyte platform integrates high-throughput screening and generative AI models within its EMolCurator module.
• It utilizes molecular properties like frontier molecular orbital information, formation energy, binding energy with Li ions, viscosity, and dielectric constant for screening.
• Additional modules, EMolForger and EMolNetKnittor, predict retrosynthesis pathways and solid electrolyte interphase (SEI) formation mechanisms.

Main Results:

• The EMolCurator module can screen over 100 million alternative molecules to identify promising candidates.
• The platform facilitates the design of new electrolyte molecules by combining screening and generative AI.
• Predictive modules aid in assessing the feasibility and performance of designed electrolytes.

Conclusions:

• The Uni-Electrolyte AI platform significantly accelerates the discovery of novel electrolyte molecules and underlying chemical principles.
• This AI-driven approach is poised to advance the practical application of next-generation rechargeable batteries.
• Uni-Electrolyte offers a powerful tool for navigating complex molecular landscapes in materials science.