Introduction In recent years vanadium redox-flow battery (VRFB) draws attention for the energy storage system, which is particularly useful for conjunction with renewable energy sources such as wind and solar. The VRFB consists of electrodes, membranes, electrolytes and only vanadium ions as active substance in both positive (VO2 +/VO2+) and negative (V3+/V2+) electrolytes [1]. This constitution makes it convenient to recover battery capacity to initial state by adjusting the concentration and valence of vanadium ions between positive and negative electrolytes, and makes it difficult to directly evaluate vanadium ions crossover of membrane, simultaneously. In the case of VRFB, the crossover of vanadium ions during an operation cannot be avoided. Thus, the dynamics of vanadium ions through the membrane are important for understanding both self-discharge and coulombic efficiency during charge-discharge processes of a VRFB. Since the permeability of four oxidation states of vanadium ions through the ion exchange membrane are known to be dissimilar to each other, and might be a function of the concentrations of other coexisting ions and electrostatic potential among other parameters. In a permeability cell for the determination of concentration gradient-induced crossover, the solution containing vanadium (enriched side) is circulated/resting on one side of membrane while the other side (deficient side) has either sulfuric acid aqueous solution or sulfuric acid containing magnesium sulfate to equalize the ionic strength between two solutions, minimizing the effect of the osmosis of the water [2]. Although the environment above mentioned are clearly different from the composition in a VRFB. In this study, we product radioactive vanadium 48V as isotope tracer and investigate the direct observation method of dynamics of vanadium ions through the membranes in VRFB. Experimental The vanadium oxide target was subjected to irradiation with bremsstrahlung radiation of 50 MeV electrons converted by Pt for about 12 h at linear accelerator of Research Center for Electron Photon Science, Tohoku University, Sendai, Japan, with an average beam current of 125 microA. This target contained a number of radionuclides, 46,47,48Sc with 48V. For the preparation of the 48V source for radioactive vanadium tracer, it is necessary to separate the contaminating radioactive scandium from this target. For the separation we applied known method [3] with some modifications. After the cooling time, the target was dissolved in conc. HCl by repeated evaporation to near dryness and finally dissolved in 1 M HCl for the adjustment of vanadium concentration to 1 M. The solvent extraction was carried out from HCl solution above mentioned with cyclohexane dissolved 0.1% HDEHP to separate scandium from bulk vanadium. Pretreatment method of CEM (Cation Exchange Membrane) is 1) immersed in 3 wt% H2O2 solution at 80 degree for 1 hour, 2) immersed in deionized water at 80 degree for half hour, 3) immersed in 1 M H2SO4 solution at 80 degree for 1 hour, 4) immersed in deionized water at 80 degree for half hour, 5) stored in deionized water before use. In order to observe the diffusion of the vanadium ions penetrating through the membrane, a flange-type cell with Nafion type membrane (DuPont, USA), was employed in this experiments. The effective area of the membrane was 2.35 cm2. As shown in Fig. 1, radioactive vanadium 48V was put in the B-side and radiological measurement for the identification of 48V was carried out in the A-side. The radiological measurement was made using Geiger-Mueller counter or Germanium semiconductor detector. Results and Discussion Figure 2 indicates the change of VO2+ concentration (refer to right axis) and change of 48V activity (refer to left axis) in the receiver side with time under identical condition for different membranes. The VO2+ concentration and 48V activities increase with time for N112 is the highest of them all, and N115 is nearly comparable to that for N117. These unities between VO2+ concentrations and 48V activities indicate that the hypothesis that makes no difference between isotopes is true. It’s to be noted that the 48V activities in the receiver side increase with a margin of large error due to the statistical variation cursed by incidental 48V decay process. In our presentation, we will discuss permeability obtained from radioactive measurements. Acknowledgements We would like to acknowledge Dr. Hidetoshi Kikunaga, Research Center for Electron Photon Science, Tohoku University with his kindly support for preparation of 48V. This work was supported by JSPS KAKENHI (Grant-in-Aid for Scientific Research) Grant Number 26820417.