Modeling degradation in graphite-based electrodes for Li-ion batteries

Abstract

The Lithium-ion battery technology (LIB) revolutionized energy storage systems and enabled the so-called mobile revolution. Due to their characteristics, i.e., high potential, high energy density, and capacity, these batteries have changed and improved our lives but will undoubtedly remain key to our lives in the years to come. From portable electronic devices, as complementary support to renewable energy sources (whose generation fluctuates over time, depending on the conditions), up to the fundamental role they play in implementing electric vehicles. Although the degree of optimization of current batteries is high, ongoing research is focused on: the improvement of battery performance and durability, its scalability, reduction of production costs, and reduction of environmental impacts of batteries. While the principle of operation of batteries is relatively simple (i.e., they are electrochemical systems inwhich redox reactions occur in the active materials that make up the electrodes), the necessary fields of knowledge involved in their analysis, design, and manufacture is broad: not only can it be approached from electrochemistry, but it is required to do so in conjunction with physics, materials science, engineering, etc., as well as advanced computational techniques (e.g., artificial intelligence, big data, machine learning), applied fromthe atomic scale up to the complex system, which is the battery. Specifically, mechanics have a determining influence on the performance and lifetime of LIBs: chemo-mechanical degradation is one of the main problems encountered in today’s batteries. It is precisely in this area that this doctoral thesis focuses on. The rapid development and increased demand for LIBs require further research into the degradation mechanism of anode and cathode materials. The global objective of this work is to present a 3D coupled diffusivemechanical model to gain insight into the degradation processes of graphite active particles (APs) by including fracture formulation. This thesis proposes a novel lattice model approach for simulating the fracture processes driven by diffusion-induced stress in electrode active particles of LIBs. The numerical framework analyzes the mechanical degradation and capacity loss of graphite particles in LIBs anode. The numerical models developed are based on the finite element method — specifically discrete models, also known as lattice-model. From the computational point of view, this choice raises additional issues, such as the discretization process, the size of the representative element, and the incorporation of the damage model. To account for material inhomogeneities, the lattice model approach includes a randomness parameter and a stochastic characterization of material properties. In contrast to most existing works that assume perfect spherical particles, the proposed methodology can include the effect of the shape of particles, their internal structure, and preexisting defects in crack initiation and propagation. Furthermore, the model is used to analyze the impact of particle coating as a strategy to diminish the effect of transient cracking (which leads to early capacity fade). The obtained results capture most high-level observations on particle cracking, showing crack patterns consistent with the experimental results. It provides the basis for the battery degradation modeling community to integrate it into a much more detailed and broader degradation module. It can be very useful for developing improved lithium-ion batteries by introducing new components.