Uncertainty in solid phase diffusivity determination: Comparing electrode parameterization workflows for P2D modeling of commercial battery cells

The pseudo-two-dimensional (P2D) battery model is a cornerstone of battery research, striking a critical balance between accuracy and computational efficiency for mathematical modeling of battery cells. It captures essential mechanistic parameters while remaining applicable for optimizing battery design and operation. It also often serves as the basis for aging modeling, a key challenge in the advancement of lithium-ion battery technology. However, wider adoption of P2D models is hindered by parameterization challenges, in particular the demanding time and experimental requirements of direct measurement strategies that preserve the full value of physically meaningful parameters. Among these, the solid phase diffusion coefficient, which is highly dependent on the lithium concentration in the active material particles, is particularly challenging. While Galvanostatic Intermittent Titration Technique (GITT) is widely used to measure it, its assumptions break down in phase-change materials and thick porous electrodes as used in commercial battery cells. This work evaluates different methods for determining diffusion coefficients, focusing on their impact on parameter plausibility and model accuracy under dynamic driving conditions. A novel approach is introduced to identify invalid GITT measurements and probability distribution of diffusion coefficient, allowing the construction of accurate models with plausible parameters.