Li-air batteries (LAB) have a theoretical energy density of 3500 Wh kg⁻¹, overwhelming that of Li-ion batteries, due to the Li metal anode and atmospheric oxygen fed to the air electrode (AE). Therefore, they have been attracting much attention because of the potential longer cruising range of electric vehicles, although many problems are still left to be solved. The most serious one among them is large charge overvoltage of AE due to the low reactivity of discharge product Li₂O₂, which induces clogging of the air electrode during repeated discharge/charge cycles, decomposition of the electrolyte solution, and air electrode corrosion. Redox mediator (RM) to reduce the charge overvoltage is widely studied nowadays, while RM⁺ generated at AE diffuses and reacts with Li negative electrode to cause shuttle effect ^{[1, 2]}. In this study, we employed both RM to reduce the charge overvoltage and LiNO­₃ as an electrolyte salt in tetraglyme (G4)-based electrolyte solution to modify the Li surface into Li₂O as a protector against shuttle effect ^[3]. We further applied RM-coating on AE to stabilize the cell performance and suppress the shuttle effect ^[4].

AE consisted of Ketjen Black, PVDF binder, and MPT (10-methylphenothiazine) as an RM coated on a carbon paper substrate. Two types of electrolyte solutions were used: 1.0 M LiNO₃/G4 with and without MPT. Three cell configurations with different combinations of AE and electrolyte solution were assembled: “RM-free” without MPT, “MPT-in-EL” containing MPT in the electrolyte solution, and “MPT-on-AE” with MPT coated on AE. The amount of RM dosed on AE was adjusted to be the same as that dissolved in the electrolyte solution of the MPT-in-EL cell. The cell components, AE of 2 cm², a separator, a Li negative electrode and an electrolyte solution, were stacked and contained in an air-tight test cell, which was supplied with pure O₂ gas of 1.0 mL s⁻¹ at 50 ^oC. The discharge/charge cycle test mode was constant current-constant capacity of 200 mA g_KB⁻¹ and 500 mAh g_KB⁻¹ between 2.0 V and 4.5 V.

Figure 1 shows cross sections of discharge/charge voltage profiles at the midpoit of 250 mAh g_KB⁻¹ of the three cells, RM-free, MPT-in-EL, and MPT-on-AE. MPT-on-AE revealed better performance with the less overvoltages in both charge and discharge than the others, although slight degradation was indicated even in MPT-on-AE by the 15th cycle. While we could conceive various causes for the degradation such as dissolution of the coated MPT, decomposition of MPT, and covering of AE with deposits, we started with the quantitative analysis of the dissolved MPT from the coated MPT. Figure 2 shows the results of colorimetric analysis of MPT on AE before discharge, which evaluates the dissolved amount of the coated MPT. A calibration line was obtained by dissolving MPT in 1.0 M LiNO₃/G4 base solution. MPT sample was extracted from MPT-on-AE cell after 1 h rest following cell assembly by means of centrifugal separation. The results showed that half of the coated MPT eluted immediately. However, SEM-EDS analysis evidenced no trace of shuttle effect, suggesting local high concentration of the applied MPT near the AE. Other analytical results concerning degradation causes will be reported on the presentation.

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).

[4] Y. Hayashi, M. Saito et al., Electrochemistry, 89, 557-561 (2021).