Large-scale batteries play an important role in the effective use of renewable energy like wind and solar power. Redox flow batteries (RFBs) offer high-speed response, independent design of power and energy, and high safety advantages, which have attracted much attention. RFBs consist of tanks of electrolyte that store chemical energy and electrochemical cells that reversibly convert chemical energy to electricity. Various RFBs can be classified into aqueous RFBs and non-aqueous RFBs, according to the electrolyte. Among aqueous RFBs, all-vanadium RFBs are the most mature technology and have been used in practical applications.

We have reported a Ti (IV) and Mn (II) mixed aqueous electrolyte which is applied for a Ti/Mn RFB. The aqueous Ti/Mn RFB shows a high energy density of 23.5kWh/m^{3 [1]}, and provides a low-cost option. The precipitation of MnO₂ particles generated from instable Mn³⁺ ion can be greatly suppressed by mixing Ti and Mn for both positive and negative electrolytes ^[2]. However, the current density (50mA/cm²) remained low and needed to be improved. In the previous study, we succeeded to enhance the reactivity of both positive and negative electrodes by the thermal treatment of carbon electrodes. The full-cell performance was improved to the energy efficiency of 83.2% at 100mA/cm^2　[3].

For further enhancement of the cell performance, the resistivity of the membrane also needs to be considered. In this study, we investigate various ion exchange membranes for the Ti/Mn RFB. Some membrane samples are pretreated to achieve lower resistivity. The improved cell performance is confirmed with the appropriate membrane. We will also discuss the ion crossover and the coulombic efficiency of various membranes. In addition, the cell performance is affected by the state of the precipitation of MnO₂ in the Ti/Mn RFB. The condition of the electrolyte (SOC, MnO₂precipitation, flow rate, etc.) should be considered at the same time. The cell components and the electrolyte condition will be discussed comprehensively in the presentation.

References
[1] Y. R. Dong et. al., ECS Trans., 69(18), 59 (2015).
[2] H. Kaku et. al., ECS Trans., 72(10), 1 (2016).
[3] Y. R. Dong et. al., PRiME meeting, 26 (October, 2016).

Acknowledgements
This work is financially supported by Ministry of Economy, Trade and Industry (METI). This work is technically supported by The National Institute of Advanced Industrial Science and Technology (AIST).