Prediction of Reversible Lithium Plating during Fast Charging with a Pseudo-3D Lithium-Ion Battery Model
Serena Carelli,Wolfgang G. Bessler
DOI: https://doi.org/10.1149/ma2022-012344mtgabs
2022-07-15
ECS Meeting Abstracts
Abstract:Mathematical modelling and numerical simulation have become standard techniques in Li-ion battery research and development, with the purpose of studying the issues of batteries, including performance and ageing, and consequently increasing the model-based predictability life expectancy. The efficient and fast charging of Li-ion batteries remains a delicate challenge for the automotive industry, being seriously affected by the formation of lithium metal on the surface of the anode during charge. This degradation process, lithium plating, is very damaging for the mechanical and chemical integrity of the battery, which not only sees its capacity lowered but could also incur serious damage and the risk of thermal runaway. It is very difficult to detect lithium plating in situ without a direct observation of the open cell, but it is possible to deduce its presence by analyzing the cell behavior during cycles of charge/discharge in critical conditions and detecting some peculiarities which have been shown to indicate plating. The most common hints are a voltage plateau due to lithium oxidation during discharge at constant temperature and a voltage drop due to re-intercalation of metallic lithium during heating of the cell. On the other hand, the absence of any evidence of changes in voltage should not be considered as proof of evidence of a complete absence of lithium plating. Following our development of a comprehensive modelling and simulation framework for a commercial 0.35 Ah high-power lithium-ion pouch cell with LCO/NCA blend cathode 1 , here we present an extended pseudo-3D (P3D) model 2 in which a lithium plating reaction has been integrated and parameterized. The model is able to describe and predict both the equilibrium potentials and the non-equilibrium kinetics of the competing intercalation and plating reactions for arbitrary macroscopic operating conditions (C-rate, temperature, SOC). A relatively simple and common way to assess plating risk with P2D models is to compare the simulated local anode potential Δφ an with the thermodynamic plating condition of Δφ Li eq = 0 V, but this approach shows several pitfalls that have not been well discussed in literature, including the effects of temperature, pressure and ion concentration on the thermodynamics and kinetics of the plating reaction (see Figure 1a). An extra reaction, simulating explicitly the re-intercalation of the plated lithium, has also been included and can be freely switched on to simulate a case in which the cell is likely not showing macroscopic plating hints. The models allow the creation of operation maps (see Figure 1b) and an accurate spatiotemporal analysis of the competing reactions and lithium plating formation at the electrode-pair scale (1D, mesoscale) and intraparticle scale (1D, microscale) over a wide range of conditions. The governing equations for this model are implemented in an in-house multiphysics software. The electrochemistry model is based on the use of the open-source chemical kinetics code CANTERA, enabling the thermodynamically consistent description of the main and side reactions. To validate our extended model, we simulated and successfully reproduced our own experimental data on our modelling reference cell (0.35 Ah high-power lithium-ion pouch cell with LCO/NCA blend cathode - where no macroscopic plating hints are present) and the published experimental data from Ecker et al. 3 (40 Ah high-power lithium-ion pouch cell with NMC cathode - where the plating hints are instead clearly visible). References S. Carelli, M. Quarti, M. C. Yagci and W. G. Bessler, J. Electrochem. Soc., 166(13), A2990-A3003 (2019). S. Carelli and W. G. Bessler, J. Electrochem. Soc., 167(100515) (2020). M. Ecker, Lithium Plating in Lithium-Ion batteries: An experimental and simulation approach, RWTH Aachen University (2016). Figure 1