Introduction The large scale installations of renewable energy are required to prevent global warming problem. Since the cost of solar cells has reduced remarkably, the key issue becomes the management of unstable power generation based on renewable energy, which depends on time, weather conditions and location. Therefore, large scale energy storage becomes more important for stable power supply and for further installations of renewable energy. Secondary batteries such as lithium ion battery are the most popular one but their high cost limits the large scale installations. Novel batteries such as lithium-air battery have been reported, but the critical issues are their large overvoltage and instability during charge/discharge cycles due to their complex redox reactions with dissolution/precipitation. “Power to gas system” is another approach that is producing hydrogen by water electrolysis. By transporting hydrogen, surplus energy can be delivered. But the storage system requires ultra-low temperature (< 20 K) or high pressure (several tens MPa). In the paper, we proposed “Carbon-air secondary battery” system as another approach, and we demonstrated the power generation and charging repeatedly. Furthermore, we evaluated the charge/discharge efficiency. Carbon-air secondary battery system The system is focused on the redox reaction of carbon, C + O2 ⇄ CO2. Because carbon has high specific energy (2500 Wh/kg and 1940 Wh/L based on liquid CO2) and CO2 can be much easier and safer to be stored by liquidation (5.7 MPa at 293 K) than hydrogen, the system is expected as large scale energy storage to solve the problems of hydrogen storage. To construct the system with high efficiency, the electrochemical devise requires high temperature. Direct carbon fuel cells (DCFC) using solid carbon as fuel, such as molten carbonate type, solid oxide type (SOFC), and their hybrid type, have been studied [1]. Because the ratio of Gibbs free energy to enthalpy of the reaction (C + O2 → CO2) is near to 1 and carbon fuel utilization ideally can reach almost 1, DCFC potentially has high efficiency. The way to deliver solid carbons to the three phase boundary (TPB) are critical in development of DCFC. We had proposed rechargeable DCFC (RDCFC) based on SOFC delivering solid carbon to TPB by using hydrocarbon pyrolysis [2-4]. As shown in Fig. 1 (left), three electrochemical reactions (1-3) occur on the fuel electrode. Direct carbon electrochemical oxidation reactions (1, 2) require larger overvoltage than that of the reaction (3), but the produced CO2 induces Boudouard equilibrium reaction (4) that leads the reaction (3) to proceed preferentially, which contributes to the reduction of total overvoltage with high fuel utilization. In this paper, we propose a novel “secondary battery” system with power generation by RDCFC, and charging by CO2 electrolysis and store of both solid carbon and liquid CO2. In the case of CO2 electrolysis (Fig. 1 (right)), the reaction (5) has large overvoltage. But if the reaction (5) proceeds easier, the efficient secondary battery system can be constructed. Experiment al The cells were consisted of Ni / GDC (Gd0.10Ce0.90O2- δ) composite fuel electrode, LSM (La0.8Sr0.2MnO2) air electrode and YSZ (8 mol% Y2O3 – stabilized ZrO2) or ScSZ (10 mol% Sc2O3 and 1 mol% CeO2 –stabilized ZrO2) disk electrolytes. The image of experimental setup is shown in Fig. 2. The fuel electrode side and the air electrode side were filled with Ar and air respectively at 800℃. The initial fuel of solid carbon was deposited on the fuel electrode by supplying propane for 5 minute, and the fuel electrode side was subsequently filled by CO2 and closed. Finally power generation and electrolysis at constant current were carried out repeatedly. Results and discussions We succeeded in stable cycle of power generation at 100 mA/cm2 over 30 times. The coulomb efficiency was ranged from 0.8 to 1 and the maximum charge/discharge efficiency was 0.35. When the current density were varied between 50 mA/cm2 to 175 mA/cm2, the maximum power density and average power density reached to 88 mW/cm2 at 150 mA/cm2 and 47 mW/cm2 at 100 mA/cm2 respectively. The power generation of our proposed “Carbon-air secondary battery” system could be demonstrated. Reference [1] T. M. Gür, Chem. Rev, 113, (2013), 6179-6206 [2] M. Ihara, K. Matsuda, H. Sato and C. Yokoyama, Solid State Ionics, 175, (2004), 51-54 [3] M. Ihara and S. Hasegawa, J. Electrochem. Soc., 153 (8), (2006), A1544-A1546 [4] S. Hasegawa and M. Ihara, J. Electrochem. Soc., 155 (1), (2008), B58-B63 Figure 1