Advances in Methods of Synthesis and Spatial Characterization for Lithium-Ion Batteries

Lithium-ion batteries currently have the highest power and energy densities compared to other battery chemistries, making them the choice solution for renewable energy storage and electric vehicles. Unfortunately, cathode materials available for commercial batteries still suffer from capacity fade caused by mechanisms of irreversible degradation. The electrode-electrolyte interface presents an area of increased instability at high voltage and cycling rates. To improve the stability of cathode materials, it is essential to explore both structural and chemical influences. The following thesis details synthetic routes, advanced X-ray imaging, and deep learning methods of quantification to develop a comprehensive understanding of the synthetic, electrochemical influences, and modes of degradation that affect capacity retention for Li-ion batteries. The first study examines the synthetic optimization of a new cathode composition following the influence of synthetic parameters on chemistry and structure. Tomographic transmission X-ray microscopy (TXM) with methods of deep learning segmentation were used to quantify changes in three-dimensional morphology during high-temperature lithiation. The results led to the second synthetic study with the goal of surface stabilization. A novel cathode material with an aluminum gradient structure was synthesized through continuous growth from core to surface layer, which exhibited longer life capability and decreased impedance over the bare material during repeat cycling. Further surface stabilization efforts have led to the development of complex cathode heterostructures, targeting architectures of multiple compositional shells. For the third project, multi-shelled cathodes were used to develop a new tomographic TXM method to fully quantify the spatial dependence of elemental composition, allowing the complete interrogation of the role of transition metal diffusion after high-temperature synthesis. Lastly, signs of morphological degradation in anode and cathode composites were explored in an 18650 battery during fast charge cycling using operando micro-computed tomography. Deep learning segmentation was used to identify signs of electronic isolation and irreversible electrode expansion. A greater understanding of the structure and underlying influences of battery chemistry was achieved using approaches spanning multiple imaging scales with the development of statistically significant quantification. The combination of synthesis and imaging creates a feedback loop that informs the design of new cathode materials with increased stability.