Introduction The heavy hydrogen isotopes, deuterium (D) and tritium (T) are important materials in the field of nuclear energy. Particularly in recent years, development of new effective separation method is required for the water contaminated by the T. Every day the enormous volume is accumulating in Fukushima Daiichi Nuclear Power Plant. The water electrolysis, which is the industrially most general separation method, has disadvantage that consume energy is tremendous. To solve this point, we have investigated novel hydrogen isotope separation system named Combined Electrolysis Fuel Cell (CEFC) [1]. In this system, hydrogen and oxygen produced by electrolysis is used for power generation in fuel cell. From this, we can save energy and get further isotope effect by electrochemical reaction in fuel cell. Actually we reported that the D isotope was enriched in water phase by fuel cells with anion or cation conducting polymer [2, 3]. Here, we newly focused on another type of fuel cell, Alkaline Fuel Cell (AFC), and investigated the separation mechanism by several electrochemical methods. Experimental The anode was Ru catalyst supported by carbon and the cathode was Pt one. The electrolyte was 30 wt.% KOH solution. Pure H2 or D2 gas was supplied to anode and air was did to cathode. The reference electrode was Hg/HgO. Linear sweep amperometry was measured at the scan rate of 20 mA/sec until the cell voltage reached 0.4 V. Simultaneously electrochemical impedance spectroscopy (EIS) was carried out by frequency analyzer. The frequency range was from 0.1 Hz to 1 x 104 Hz. The experimental temperature was 50 ºC. 3.Results Figure 1 shows the cell voltage-current curves of AFC. The same values of cell voltage at 0 A did not clearly demonstrate that the equilibrium potential of hydrogen oxidation reaction depended on the isotopes. The activation overvoltage, which can be detected as the voltage drop when the current was applied, was about 0.2 V for both gases. On the one hand, the cell voltage decreased linearly until the current reached 1400 mA when H2 gas was used. On the other hand, the voltage of D2 result was remarkably fallen at 1160 mA. The difference of the cell voltage was mainly attributed to the anode potential. This result suggested that the H2 gas transportation in the gas diffusion layer was faster than that of D2 probably. In the meeting, we report the separation factor of the D in a ddition to the EIS results. [1] H. Matsushima, T. Nohira, T. Kitabata, Y. Ito, Energy. 30 (2005) 2413. [2] S. Shibuya, H. Matsushima, M. Ueda, J. Electrochem. Soc. 163 (2016) F704 [3] R. Ogawa, H. Matsushima, M. Ueda, Electrochem. Commun. 70 (2016) 5. Figure 1