Characterization of Li2s-P2S5 Solid Electrolytes/Alloy Negative Electrodes Interface Using Microelectrode Techniques

Abstract

Introduction All solid-state Li-ion batteries (LIBs) are considered as promising candidates for post LIBs. Recently, some inorganic sulfides have been reported as highly conductive Li-ion conductors. Hayashi et al. found that Li2S-P2S5 glassy solid electrolytes showed high ionic conductivity comparable to liquid electrolytes [1]. Furthermore, the Li2S-P2S5 solid electrolytes are so soft and flexible that a closer contact interface between an electrode and solid electrolyte without any voids or cracks is formed by hot pressing technique, which is quite different from the oxide solid electrolyte/electrode interphase. Previously, we investigated the charge transfer reaction at the interface between Li2S-P2S5 solid electrolyte and Li metal electrode by electrochemical methods with microelectrodes [2, 3]. In this study, deposition/dissolution of Li at various metal negative electrodes (In, Al, Sn) with alloying was investigated. Activation energy (E a) for charge transfer reaction at the interface between these electrodes and solid electrolyte was evaluated to discuss suitability for the negative electrode of all solid-state LIBs. Experiments Li2S-P2S5 solid electrolyte powders were prepared with Li2S and P2S5 powders by mechanical milling technique in an Ar-filled glove box [1]. For electrochemical measurements, a two-electrode cell was prepared. Four metal disk electrodes (φ 500 μm) were used as the working electrode and Li disk with 8 mm in diameter was used as the counter and reference electrodes. Li was deposited on the working electrode at -150 mV (vs. Li/Li+) and the potential was stepped as -10, +10, -20, +20 mV (vs. OCV) until the potential reached at ±150 mV. Results and discussion To evaluate exchange current density i0 , steady-state current i and overpotential η were inserted into Allen-Hickling equation [Eq (1)]; ln[i/{1-exp(Fη/RT)}] = ln i0-αF/RT (1) where F, R, T and α are Faraday constant, gas constant, absolute temperature and transfer coefficient, respectively. The value of i0 was evaluated from the y-intercept of each plot and replotted against the reciprocal of absolute temperature to calculate the Ea for the Li+/Li couple reactions using the following Arrhenius equation; ln i0 = lnA-(E a /RT) (2)where A is frequency factor. Figure 1 shows Arrhenius plots of i 0 for four metal negative electrodes and Figure 2 shows E a values calculated with Arrhenius plots. The In and Al electrodes had E a value close to Li electrode. On the other hand, the Sn electrode had the E a value larger than other metals. To investigate the reason why the Sn electrode showed high Ea value, we tried to find the rate determining steps of the charge transfer reaction between Li2S-P2S5 solid electrolyte and alloy negative electrodes by using electrochemical methods. We found the rate determining steps for the In and Al electrodes were formation of Li alloy but that for the Sn electrode was the diffusion of Li in Sn metal. This difference suggests the diffusion rate of Li in Sn will be lower than that in In and Al. These results suggest that the In and Al electrodes are suitable for the negative electrode of all solid-state LIBs compared to the Sn electrode.