Introduction Lithium-air batteries (LABs) are one of the promising post lithium-ion battery technologies owing to its remarkably high theoretical specific energy density. Low energy efficiency and poor cyclability due to large charging overpotential should be addressed for realizing practical LABs. Recently, it was reported that the overpotential especially during charging is greatly reduced and thus the cyclability is significantly improved by modification of air electrodes with a noble metal catalyst such as Pt or Ru 1). An appropriate base metal-based electrode catalyst, e.g. Co3O4 and MnOx 2), is also capable to effectively reduce the overpotential. In this study, the base metal-based ternary (Fe-Co-Ni) and binary (Fe- Ni, Fe-Co, and Co-Ni) alloy catalysts were prepared. Their catalytic activities on the discharge-charge reactions were evaluated and compared with the RuO2 catalyst. Experimental The base metal alloy-modified carbon, Fe-Co-Ni/KB, Fe-Ni/KB, Fe-Co/KB, and Co-Ni/KB (KB: Ketjen black), were prepared via the impregnation process. RuO2/KB was also prepared by a similar method. The catalyst-modified carbon and polyvinylidene fluoride were mixed together with N-methyl-2-pyrrolidone solution. The obtained slurry was coated on the carbon paper and dried in vacuum at 80 oC overnight. The coin-type cell of CR2032 consisting of the carbon cathode, lithium metal anode, and 1 mol dm-3 LiTFSA/tetraglyme electrolyte was assembled in an Ar filled glovebox. For the electrochemical measurement, the discharge-charge tests were carried out in dry air atmosphere at 25 oC in the voltage range of 2.0 - 4.3 V at the current density of 50 μA cm-2. Results and Discussion Figure 1 shows the 1st and 20th discharge-charge curves of the coin cells with the prepared catalyst-modified carbon electrodes. The discharge behavior was similar, ca. 2.7 V of stable plateau observed, regardless of the carbon electrodes used. On the other hand, a remarkable difference was observed in the charging process, depending on the type of the catalysts. The RuO2 catalyst effectively suppresses the charging overpotential over 20 cycles. The overpotential of the Fe-Ni/KB electrode was also small during the initial charging stage at the 1st cycle, while the catalytic activity of this catalyst would be almost completely diminished after 20 cycles. On the other hand, the charging overpotential of the other electrodes with Co was very large irrespective of the cycle number. As the ternary Fe-Co-Ni catalyst is found to be not contributed to decreasing charging overpotential, the catalytic activity of Co would be considerably small and/or Co would impede the catalytic activity of other metals. Unfortunately, the air electrodes are rather spoiled by the presence of the base-metal based catalysts, probably due to oxidation of the catalysts during charging process. Improvement in catalyst durability is pivotal to achieve the cost-effective base metal-based electrode catalysts for LABs. Reference s 1) F. Wu, et al., J. Power Sources, 332, 96 (2016). 2) D. Oh, et al., J. Phys. Chem. C, 121, 1407 (2017). Acknowledgments This research was partly supported by ALCA-SPRING project of Japan Science and Technology Agency (JST). Fig.1. Discharge-charge curves at the 1st (left) and 20th (right) cycles of the metal-modified carbon electrodes. current density: 50 μA cm-2, loading density of active materials: 0.5 mg cm-2. voltage range: 2.0 - 4.3 V, discharge-charge time: 2 h, temperature: 25 ° C, atmosphere: dry air. Figure 1