A fast and accurate method for inferring solid-state diffusivity in lithium-ion battery active materials: improving upon the classical GITT approach

TL;DR

This paper introduces ICM, a novel, faster, and more accurate method for inferring solid-state diffusivity in lithium-ion battery materials, improving upon the classical GITT approach by using a consistent physical model.

Contribution

The paper presents ICM, a new methodology that infers diffusivity from a physical model, reducing data collection time and increasing accuracy compared to classical methods.

Findings

ICM yields more accurate diffusivity estimates.

ICM requires five times less data collection time.

The method is robust across different data types.

Abstract

Data collected using the galvanostatic intermittent titration technique (GITT) and application of the Sand equation is a ubiquitous method for inferring the solid-state diffusivity in lithium-ion battery active materials. However, the experiment is notoriously time-consuming and the Sand equation relies on assumptions whose applicability can be questionable. We propose a novel methodology, termed Inference from a Consistent Model (ICM), which enables inference of solid-state diffusivity using the same physical model employed for prediction, and is applicable to more general and quick-to-measure data. We infer the diffusivity (as a function of inserted lithium concentration) by minimising the residual sum of squares between data and solutions to a spherically-symmetric nonlinear diffusion model in a single representative active material particle. Using data harvested from the NMC cathode of a commercial LG M50 cell we demonstrate that the ICM is robust, and yields more accurate diffusivity estimates, while relying on data that are five times faster to collect than that required by the classical approach. Moreover, there is good reason to believe that further speed ups could be achieved when other types of data are available. This work contributes towards developing faster and more reliable techniques in parameter inference for lithium-ion batteries, and the code required to deploy ICM is provided to facilitate its adoption in future research.