Multiscale Experimental Approaches to Li-ion Battery Research: From Particle Analysis to Optimized Battery Design.

Abstract

We approach the challenges in Li-ion battery research through multiscale experiments: a small but macro scale Li-ion battery was designed for an implantable surgical device for distraction osteogenesis, while in particle- to micro-scale, the baseline cathode materials for Li-ion batteries were investigated for their structural and electrochemical characteristics. For the optimized battery design study, we first identified the power / energy requirements for a common clinical protocol using a novel distraction device developed in parallel to its battery design, and then ran an algorithm to select a commercially available battery with minimal volume that satisfied the system demands. A polymer Li-ion battery was selected due to high power and energy densities as well as its favorable geometry. A bench-top prototype device, integrating an actuator, a control circuit, and a battery, was fabricated to test its functionality and reliability, and eventually will be ready for animal implantation studies. Particle- to micro- scale experimental studies of Li-ion insertion metal oxide cathode materials were conducted using simple forms of the baseline materials, such as thin films and dispersed single particles, aiming to understand their structural characteristics and electrochemical properties. Various characterization techniques including SEM, TEM, XRD, and AFM were used to observe external and internal microscopic morphology of primary particles from candidate cathode materials for EV applications, such as LiFePO4, Li[Ni1/3Co1/3Mn1/3]O2, and LiMn2O4. Their anisotropic and inhomogeneous nature was revealed due to the hierarchic structure consisting of crystal grains and grain boundaries. Thin film study of LiMn2O4 also showed similarly complex microstructures that were found to be determined by their fabrication conditions, including substrate material and annealing temperature. In an experimental study with single LiMn2O4 particles, we take one step toward precise modeling and control of large format cells in EV applications by generating and incorporating accurate model parameters, including diffusion coefficients from CV and PITT methods, and realistic particle geometries from AFM scanning data. Simulation of Li-ion intercalation with the implemented experimental measurements showed that LiMn2O4 particles could be under higher intercalation-induced stress due to slower diffusion and local stress concentration at the grain boundaries.