We are pulling together climate related resources.

Early Detection of Lithium Plating on Lithium-ion Batteries

Stage: Prototype

This invention is a machine-learning (ML) framework for detecting lithium plating (Li-plating) in batteries by analyzing global and conventional cell electrochemical (EC) signatures. Li-plating causes serious degradation and safety issues and must be identified early and avoided. Earlier inventions for detecting Li-plating, which monitor any single EC signature, have serios limitations. The issue is that some of these EC signatures may not be detectable in realistic operating conditions, despite the existence of Li-plating. Therefore, combining multiple EC signatures in a ML framework offers a more robust and accurate approach to detect Li-plating. 

Instead of relying on one electrochemical signature, the proposed method uses multiple signatures to avoid false negative cases while providing more reliable and earlier Li-plating detection. This ML framework distinguishes Li-plating from normal solid electrolyte interface (SEI) dominant battery degradation. It uses a variety of physically meaningful signatures including capacity loss, coulombic efficiency, end of charge rest voltage, and post-charge open circuit voltage relaxation profiles. This framework is directly applicable to full cells without any special measurement requirements or additional sensors , and the classification can be made as early as 25 life cycles. 

Development Status: TRL 3, method demonstrated in lab using R&D lithium-ion battery. 

Provisional Patent Application No. 63/116,032, “An Electrochemical signature-based machine learning framework for early detection of lithium-plating in lithium-ion batteries.” 

Applications and Industries

• Battery development and optimization

• Electric vehicle battery health monitoring

• Stationary and grid storage utilities

• Battery management system


• Early, robust detection of Li-plating.

• Applicable to a variety of Li-ion battery uses.

• Redure expensive post testing cost

• Shorten battery development and optimization cycle

• Easily implemented into battery management systems.

• Reduce liability of costly battery failures.