Efficient models and algorithms for mass conservation and morphology evolution in lithium metal batteries

The demand for energy storage devices with high energy density, coulombic efficiency, long-term stability, and high capacity while ensuring safety has never been higher. However, the efforts towards carbon neutrality and exploration of next-generation batteries for electric vehicles and mobile applications are still insufficient to meet the demands. The development of Lithium-ion batteries is a giant leap in achieving the utilization of lithium, which has high reactivity, mobility, and superior energy density along with high output voltage. However, there is one major area of improvement to advance the conventional lithium-ion batteries: the anode electrode. Lithium metal has the highest energy density among the other potential candidates for anodes, ahead of conventional graphite electrodes which are based on the intercalation of lithium ions. Despite the early research interests, the metal anode was not commercially successful due to safety concerns and inferior cyclability. Even today, those defects are challenging and need further research. Thus, to resolve the above-mentioned difficulties, there is a significant need for a fundamental understanding of the morphology changes during the deposition and stripping, specifically, the anomalies in the microscale, such as the formation of dendrites, local cavitation, and initial surface defects. These translate into macroscale as dendrite growth, depletion of the electrolyte by the continuous solid-electrolyte interface (SEI) layer growth, and formation of isolated regions called 'dead lithium'. This thesis focuses on the physics-based models and algorithms at different scales and varying complexity of the system to simulate the evolution of the anode surface in lithium metal batteries. This includes continuum, mesoscale and multiscale models conserving the system's total mass. The different approaches, such as coordinate transformation and phase-field model, are discussed with proper mathematical reformulation. Lastly, an effective algorithm for fast and accurate simulation is proposed with selected examples.