Consistent Modeling of SEI Growth via Electron Diffusion Validated Across Diverse Cell Specifications and Operating Conditions in LFP/Graphite Cells
Lukas Köbbing, William Black, Claire Floras, Saad Azam, Eniko Zsoldos, Ben Tang, J. R. Dahn, Birger Horstmann
Solid, well-supported validation of an existing electron-diffusion SEI model across a broad LFP/graphite dataset, but incremental over the authors' prior work and with unresolved high-SOC storage failures.
Abstract
Understanding and predicting degradation in lithium-ion batteries is critical for their widespread deployment. The continuous growth of the solid-electrolyte interphase (SEI) remains a dominant aging pathway, yet accurately modeling its kinetics across diverse cell specifications and operating conditions remains a significant challenge. Here, we present and validate a physically grounded SEI growth model mediated by the electron diffusion mechanism. This framework intrinsically captures the state of charge (SOC) dependence of degradation through the impact of the anode potential on the electron concentration, which governs the rate of SEI growth. The presented model explicitly accounts for variations in electrolyte salt, anode material, temperature, and cycling rate. Validation against experimental aging data for LFP/graphite cells demonstrates high predictive accuracy across a broad range of cell specifications and cycling conditions, including confined SOC windows, electrolyte salts (LiFSI, LiPF6), artificial graphites (AGC, AGD), temperatures (40-70°C), and C-rates (C/20-1C). Remarkably, a single, consistent set of parameters is applied to simulate the capacity fade for all experiments without recalibration. While quantitative overestimations occur during high-SOC storage periods, the framework consistently captures the observed degradation trends. Our findings underscore that lower SOC operation mitigates degradation, establishing a physics-based model for predicting long-term lithium-ion battery aging.
AI Impact Assessments
(1 model)Core Contribution
The paper tests a physics-based SEI growth model, in which electrons diffuse through the SEI and the rate is set by the anode potential, against a large LFP/graphite aging dataset from Dahn's group. The model is a compact ODE: dQ_SEI/dt = ξ·exp(−FU/RT)/Q_SEI, extended with multiplicative scaling factors for salt and graphite, an Arrhenius term, and a Butler–Volmer overpotential. SOC dependence follows from the graphite OCV with no extra fitting. The new element is the breadth of the validation. It covers confined SOC windows, two salts, two artificial graphites, 40–70 °C, C/20–1C, and mixed cycle–storage and solar-duty protocols, all with one parameter set of six parameters (Table I). Most of the model's formulation comes from the authors' earlier papers. This paper is best read as an extensive validation and extension to cycling, not a new theory.
Methodological Rigor
Strengths:
Weaknesses:
Potential Impact
The work gives practitioners a simple, physics-motivated, low-parameter model of SEI-driven capacity fade. It could feed lifetime prediction, SOC-window optimization for stationary LFP storage, and screening of electrolytes or graphites through a single ξ. The conclusion that lower SOC reduces degradation is already well known empirically (including from the source dataset in Ref. 24). The contribution is the mechanistic and quantitative explanation. The paper is also useful as a reference for people implementing SEI growth in pseudo-2D or system-level aging frameworks such as PyBaMM. Its reach is mostly within battery degradation modeling, with modest cross-field relevance.
Timeliness
LFP is growing rapidly in stationary storage and EVs, and LFP aging models are less mature than those for NMC. Interpretable, parameter-light aging models that generalize across conditions are a recognized need. The paper also gives timely experimental support for electron diffusion as the SEI growth mechanism, an active debate (cf. the t^0.5 revisit and the generator–collector measurements).
Strengths
Limitations
Overall
This is a solid, competent applied modeling paper with high-quality data and a clear message. Its impact is likely to be moderate: it will be cited within the SEI and battery-aging modeling community, but it is unlikely to shift the field's practices or concepts.
Generated Oct 22, 2026
Comparison History (0)
Solid, well-supported validation of an existing electron-diffusion SEI model across a broad LFP/graphite dataset, but incremental over the authors' prior work and with unresolved high-SOC storage failures.