Isolating Surface Reactivity of High Energy Cathode Materials through Model Thin Films

Abstract

Increasing demands for improved energy density in lithium ion batteries for electric vehicles and grid storage has spurred the development of high energy cathode materials. Ni-rich LiNixMnyCozO2 (x > 0.5, x + y + z = 1) cathode systems are frontrunners among the next-generation of Li-ion battery cathode active materials. Despite the material’s high theoretical capacity (~200 mAh/g), commercial adoption has been hindered by a lack of understanding of complex interactions between the electrolyte and the binder, conductive additives, and active material present in composite electrodes. In this study, we isolated the surface reactivity of LiNi0.6Mn0.2Co0.2O2 (NMC622) by physical vapor deposition of thin film electrodes (thickness < 1.5 µm) for characterization and electrochemistry without influence from conductive additives and binders. Radio Frequency Magnetron Sputtering of homemade NMC622 targets was used to deposit thin films of NMC622 onto Pt-coated Al2O3 substrates which were annealed to achieve the desired layered structure. LiCoO2 thin films were deposited in a similar manner as a baseline for the process. The dependence of crystal structure and stoichiometry on deposition and annealing conditions of these films were determined by XRD, Raman, and ICP-AES. Cathodes were cycled against Li metal in a standard commercial electrolyte of 1.2 M LiPF6 in EC:EMC (3:7 wt%) and extracted in an inert atmosphere glove box for surface characterization after rinsing in DMC. Ex situ XPS and FT-IR analysis of pristine and cycled cells traced a substantial difference in surface species dependent on annealing environments and cycling conditions of thin films. This provided a model system to investigate the influence of thin film coatings such as Al2O3, LiPON, and ZrO2 on the active material alone. The approach developed here may be extended to other promising cathode systems for which the surface chemistry of the active material is not well understood. Figure 1