The Effects of Adding Quinone Redox Mediator on TEMPO-Immobilizing Cathode for Lithium Oxygen Batteries

Abstract

The lithium oxygen battery has received much attention for its high specific capacity (3505 mAh cm-2). However, it has many issues such as the high charge/discharge overpotential and associated by-product formation. To solve this problem, Redox Mediators (RMs) have been widely proposed. RMs were reported to reduce overpotential on charging or discharging and improve cycle performance. However, oxidized RMs may diffuse or migrate from cathode side to anode side and they may be reduced by Li anode (Redox shuttle phenomenon). This phenomenon will lead overcharge and also lead RM’s decomposition. For this reason, the methodology that prevents RMs from diffusing to Li anode is required.In our previous research [1], TEMPO which is known for charge RM was immobilized on carbon paper by silane coupling. This cathode can reduce charge overpotential at very early cycles. However, decreasing charge capacity after several cycles was confirmed. One of possible causes of this problem might be the accumulation of by-product (such as Li2CO3) during discharge/charge cycle and deteriorate immobilized cathode. In this study, quinone RMs which are supposed to suppress the formation of by-product were introduced [2] As quinone RMs, 2,5-Di-tert-butl-1,4-benzoquinone(DBBQ) and Anthraquinone(AQ) are examined.Figure 1 shows 1st charge and discharge curves of O2 cathodes using TEMPO immobilized with/without quinone RMs. As shown Fig.1, with quinone RMs, the effect of reducing discharge overvoltage was confirmed. Moreover, the by-product suppression effect was also confirmed from XPS measurements. SEM images (Figure 2 (b), (d)) illustrates the difference in the morphology of the discharge product by the added quinone RMs. With DBBQ (Figure 2 (b)) about 250 nm particle-like precipitate was confirmed, while, with AQ (Figure 2 (d)) film-like precipitate was confirmed. This difference was considered to be derived from the difference is deposit mechanism [3].The addition of quinone RM also was found to affect charging behavior. In Fig. 1, when using DBBQ, insufficient amount of charge capacity compared to that of without additives, in contrast, when using AQ, sufficient amount of charge capacity was demonstrated. The reason for this was explained in the model shown in Figure 2 (a) and (c). When precipitates occurred in particle form (using DBBQ), the decomposition proceeded preferentially at the contact point between the immobilized TEMPO and the precipitate. For this reason, some of the precipitates may flow to bulk electrolyte without being decomposed, and resulting in insufficient charge capacity.On the other hands, when precipitates occurred in film form (using AQ), immobilized TEMPO could continuously react with precipitate. So, the almost precipitates could be decomposed, and sufficient charge capacity could achieve. However, the deterioration of the TEMPO-immobilized cathode, which was the original purpose, with the cycling was not suppressed. Therefore, there is need to examine other methods to suppress deterioration.In summary, suppress deterioration of TEMPO-immobilizing cathode could not be achieved by the addition of quinone RM, while, the finding as the relationship between the form of discharge product and the charge capacity obtained in this study would be useful knowledge for RM-immobilized cathode. Reference [1] Takuya Naruse, Kae Miyashiro, Toshiyuki Momma, “Immobilization of TEMPO to Carbon Paper Cathode for Li-O2 battery”, The 60th Battery Symposium in Japan, 2019/11/13.[2] Xiang wen Gao, Yuhui Chen, Lee Johnson and Peter G. Bruce, Nature Materials, volume 15, pages882–888(2016).[3] Peng Zhang, Liangliang Liu, Xiaofeng He, Xiao Liu, Hua Wang, Jinling He, and Yong Zhao, J. Am. Chem. Soc. 2019, 141, 6263−6270. Figure 1