Bridging Particle-Scale Lithiation Mechanisms and Macroscopic Performance in High-Energy Density Si Anodes via Time-resolved 3D Visualisation Methods

Anodes with high Silicon (Si) content, combined with Nickel-Manganese-Cobalt (NMC) cathodes provide interesting prospects for Li-ion batteries fostering possible energy densities well beyond the state of the art. However, when Si alloys with Lithium (Li), it undergoes significant volume changes, raising the critical question how exactly the Si particles and anode respond to the lithiation dynamics in such cells and thus impacting the battery performance. Here, we integrate in situ synchrotron X-ray computed nano-tomography, deep learning algorithm for image classification and Digital Volume Correlation-based analysis to uncover explicitly in three dimensions (3D) the lithiation behaviour of an 89 wt% Si anode in a high-energy-density 811NMC|Si cell. The custom-designed spring-loaded electrochemical cell allows temporal resolved 3D tracking of the individual silicon particle dynamics as well as full-field displacement and strain mapping at nanoscale resolution. We discover diverse lithiation behaviours, anisotropic strain evolution and mechanically distinct transformation modes across hundreds of particles. Rather than following uniform core–shell transformation, we witness that those particles develop internal lithiated networks, indicating more complex, directionally dependent pathways. We reveal that the origin of the lithiation behaviour is conditioned by stress concentrator and fracture nucleation sites, initial particle morphology as well as by the particle size and their location. Consequently, degradation manifests by localized strain heterogeneities, rather than correlated with the overall electrode expansion, underscoring the need to move beyond simplistic lithiation models. These insights provide practical design guidelines for developing robust silicon anodes, including defect screening, particle size optimization and architecture engineering.