Preparation and Electrochemical Characterization of Composite Cathodes Prepared By Aerosol Deposition for 5V Class All-Solid-State Lithium Rechargeable Batteries

Abstract

Introduction Oxide-based all-solid-state batteries (SSBs) are candidates for the next-generation of rechargeable batteries because of their advantages in terms of both high energy density and safety. However, a significant disadvantage of this type of batteries is large contact resistance between active materials and solid electrolytes, which decreases Li+ conductivity in electrodes. To solve this problem, we prepared composite powders of 5V-class active material: LiNi1/2Mn3/2O4 (LNM) and a Li+-conductive solid electrolyte (Ohara Inc. -prepared proprietary composition of Li2O-P2O5-Nb2O5-B2O3-GeO2). We deposited these composite powders on Pt substrates by an Aerosol Deposition (AD) method. AD is a powerful approach to produce composite films at room temperature [1] while suppressing reactions of active materials with solid electrolytes. After preparation, we performed electrochemical measurements of the resultant composite film electrodes. Experimental Composite powders of LNM (Toshima D50=10-15 µm) and Li2O-P2O5-Nb2O5-B2O3-GeO2 solid electrolyte (Ohara Inc.) were prepared by a dry-powder mixer (NOB-MINI, Hosokawa Micron Co.). The mixing ratio of LNM to Li2O-P2O5-Nb2O5-B2O3-GeO2 solid electrolyte was 20:1, weight basis. The obtained composite powders were deposited on Pt substrates by AD to prepare the composite electrodes. Pure LNM powder was also deposited on Pt by AD to compare to the composite electrodes. AD films were characterized by Raman spectroscopy, scanning electron microscopy and energy dispersive X-ray spectrometry (EDX). The amounts of active materials in the composite electrodes were estimated by Image-J image processing software based on cross-sectional-SEM images. Weights of LNM in AD films were estimated by using the volume and density of LNM. In this procedure, we assumed that films were considered fully densified in these estimates. The resultant films were covered by LiPON solid electrolyte films by radio frequency magnetron sputtering. Subsequently, Li anode films were deposited on the top of LiPON films by vacuum evaporation. The characterization studies of fabricated SSBs (Pt | LNM or the composite electrodes | LiPON | Li) included cyclic voltammetry (1.0 mV sec-1), impedance spectroscopy (frequency range: 10 mHz to 500 kHz), and galvanostatic charge-discharge measurements (1.0 μA cm-2). All electrochemical measurements were carried out at 25 °C in an Ar filled glovebox. Results and Discussions The SEM images of the composite powder showed that the solid electrolyte powder uniformly distributed on LNM particle surfaces. In Figure 1, the cross-sectional SEM images of AD films showed that AD films have dense structures, and the thicknesses were more than about 1 μm. In the case of AD films of the composite electrode, LNM and Li2O-P2O5-Nb2O5-B2O3-GeO2solid electrolyte were uniformly distributed with layer-like structures. Nyquist plots at 4.7 V (Li/Li+) showed that the charge transfer resistance was lower for composite cathodes compared to pure LNM cathodes. Figure 2 shows the first charge-discharge curves of SSBs. The SSB with a composite electrode had a specific capacity of 62.7 mAh g-1 which was higher value compared to the SSB with pure LNM electrode (4.3 mAh g-1). Approximately 43% of the theoretical specific capacity of LNM (147 mAh g-1) could be obtained by fabricating the composite electrolyte. It is thought that Li2O-P2O5-Nb2O5-B2O3-GeO2 solid electrolyte filled in space among the LNM particles produce Li+conductive pathways, which increased the amounts of effective active materials, and reduced apparent resistance. In this way, the discharge capacity for the composite electrodes increased. Conclusion As a result, for electrodes fabricated by an AD technique, the composite electrodes had lower charge transfer resistance and higher discharge capacity compared to pure LNM electrodes. Acknowledgement This work is financially supported by NEDO-Post LiEAD. Figure 1