Li-air batteries (LABs) have a theoretical energy density (3500Wh kg-1) and are attracting worldwide attention for longer cruising range of electric vehicles. However, the discharge product Li2O2 is difficult to be decomposed during charging with a large overvoltage, and repeated discharge/charge cycles gradually cause clogging of the air electrode, decomposition of the electrolyte, and air electrode corrosion, which shorten the cycle life. Recently, redox mediator (RM) to reduce the charging overvoltage is widely studied by adding it to the electrolyte solution, while RM+ generated at the air electrode during charge process diffuses and reacts with Li negative electrode, causing shuttle effect [1], [2]. We have found that coating RM on the air electrode suppressed the shuttle effect and stabilized the cell performance. In this study, we used both RM to reduce the discharge/charge overvoltages and LiNO3 dissolved in tetraglyme (G4)-based electrolyte solution to suppress the shuttle effect by forming Li2O layer on the Li negative electrode. [3] We introduced two kinds of RM, LiBr and 10-methylphenothiazine (MPT), to electrolyte solution or air electrodes coated on a carbon paper substrate with Ketjen black (KB)-PVdF/NMP-based slurry. The amount of RM dosed on the air electrode was the same as that in the electrolyte solution in the cell. The cells consisted of the air electrode of 2 cm2, a separator, a Li negative electrode and an electrolyte solution, which were tested under pure O2 gas flow at 1.0 mL s-1 at 50 oC. Discharge/charge cycle tests were conducted at constant current of 200 mA gKB - 1 and constant capacity of 500 mAh gKB -1 between 2.0 V and 4.5 V.Figure 1 shows discharge/charge curves of the LAB cells using a LiBr-containing air electrode in 1.0 M LiTFSI/G4 and LiNO3/G4 electrolyte solutions. The LiTFSI/G4 cell shows capacity fading before 10 cycles, while the LiNO3/G4 one shows limited fading even after 10 cycles. The difference should be due to the suppressed shuttle effect by Li2O protective layer formed on the Li electrode and the prolonged RM function as demonstrated by a plateau at around 3.5 V. Figure 2 shows comparison of cycle performance among four LAB cells, which includes RM-free, LiBr on the air electrode (AE), MPT in EL, and MPT on the air electrode (AE) cells. MPT-containing cells revealed better performance in the less overvoltage and the longer cycle life than the others. SEM-EDS and XPS analyses of the electrode surface after cycle tests showed a good consistency with the above results. Details will be presented on the day.This study was supported by JST Project ALCA-SPRING (JPMLAL1301) and NIMS Joint Research Hub Program, Japan.[1] S. Ha et al., J. Mater. Chem. A, 5, 10609 (2017).[2] Y. Hayashi, et al., J. Electrochem. Soc., 167, 020542 (2020).[3] M. Saito et al., J. Electrochem. Soc., 168, 010520 (2021). Figure 1