Aprotic lithium-oxygen (Li–O2) batteries are promising candidates for next-generation energy storage devices, because of their much higher potential energy density than Li-ion batteries. However, the practical application of rechargeable Li–O2 batteries has been limited by poor cycle performance, especially the side reactions that lower the oxygen reaction efficiency at the positive electrode. Ideally, reversible two-electron reaction (with O2/e- = 0.5) should proceed during repeated discharge/charge cycles. However, in practice, various side reactions can occur at the same time. Especially, reactive oxygen species (such as O2 -, LiO2, Li2O2, and singlet O2) tend to decompose the electrolyte (both the solvents and anions) via irreversible chemical reactions that form lithium alkyl carbonates as side products [1,2]. Although various organic solvents have been investigated as electrolytes for Li–O2 batteries, an ideal one has yet to be identified [3,4]. In the present study, we tested new types of triethylphosphate-based electrolytes in Li–O2 batteries to possibly avoid the side reactions by the reactive oxygen species, which could lead to highly efficient oxygen evolution and enhanced life span of the battery. Properties of TEP, such as non-flammability, high oxidative stability, moderate viscosity and high boiling point, are advantageous for use in Li–O2 batteries. Although several investigations have been conducted on TEP-based electrolytes with regard to use in Li–O2 batteries [5-7], to our knowledge, this study is the first demonstration that such electrolytes offer highly efficient oxygen evolution reaction during charging process in the actual batteries.We performed in situ mass spectrometry (MS) analyses to quantitatively evaluate O2 evolution during the charging process. For the Li–O2 cell with TEP-based electrolyte containing 3 M LiNO3, the analysis clearly revealed that the O2 evolution rate was almost 100 % of that of the ideal two-electron reaction (O2/e- = 0.5) during the major part of the charging process. However, there was a sharp decrease in O2 evolution rate at around 0.9 mAh/cm2, accompanied by a sharp increase in the CO2 evolution rate. As a result, the total O2 evolution yield during the charging process was about 90 % that of the ideal two-electron reaction.Next, we focus on clarifying the mechanism of CO2 evolution reaction in the final stage of the charging process. Although in situ MS analysis revealed that the O2 evolution rate was almost 100 % of that of the ideal two-electron reaction during the major part of the charging process, it suffered a sharp drop later, especially when the voltage was higher than 4.0 V. Importantly, the decreased O2 evolution rate coincided with an increased CO2 evolution rate, suggesting the need of suppressing CO2 evolution reaction. In our experiment, the carbon electrode and TEP solvent are both possible carbon sources for the CO2. To further clarify the origin of CO2, in situ MS analysis was performed using an electrode containing 13C. When using this electrode, both 13CO2 and 12CO2 were generated when the voltage was higher than 4.0 V, and the 13C/12C ratio in the evolved CO2 was almost identical to the effective surface area ratio of 13C-powder and 12C-CNT in the electrode. These results clearly revealed that the major part of CO2 came from the carbon electrode, not the TEP solvent. This result suggest that this carbon related side reaction is a crucial issue that must be addressed in order to develop rechargeable Li–O2 batteries with high energy densities and long cycle life.Refereces.[1] McCloskey, B. D. et al. Phys. Chem. Lett. 2011, 2, 1161.[2] Feng, S. et al. Mater. Chem. A 2017, 5, 23987.[3] Adams, B. D. et al. Energy Mater. 2014, 5, 1400867[4] Huang, Z. M. et al. Chem. Int. Ed., 2019, 58, 2345.[5] Bryantsev, V. S. et al. Electrochem. Soc. 2013, 160, A160.[6] Xu,W. et al. Journal of Power Sources, 2012, 215, [7] Reeve, Z. E. et al. Phys. Chem. C, 2015, 119, 26840