IntroductionAnion-exchange membrane fuel cell (AEMFC) has a potential use of non-precious metal catalysts under an alkaline environment. Furthermore, anion-exchange membranes (AEMs) are expected to be used for water electrolysis. However, the performance of AEMFC is still lower than that of proton-exchange membrane fuel cell (PEMFC). During the power generation, the PEMFC only produces H2O at the cathode, whereas the AEMFC consumes H2O at the cathode and produces twice as much as H2O at the anode. The main factors for the increase in polarization of AEMFC are limitation of H2 diffusion due to liquid water flooding at the anode and lack of H2O as a reactant at the cathode. Therefore, elucidating the water distribution in the AEMFC is essentially important. In this study, we imaged liquid water in an AEMFC during power generation. The data were compared with those of a PEMFC.ExperimentalThe water distribution in the AEMFC was imaged by neutron beams in the directions, perpendicular and parallel to the membrane-electrode assembly (MEA) surface during power generation. Imaging was conducted at the beamline BL-22 of the MLF, J-PARC(2). The cell size was 100×70 mm2, and the active area was 20×20 mm2. Single serpentine flow channels were used for the gas flow channels. The depth and the width of the gas flow channels were both 1 mm. The end plates were made of aluminum to avoid neutron scattering. The thickness of the end plate was10 mm. Both the H2 and O2 gas flow rate was 100 ml/min. Pt/CB was used as a catalyst. For the electrolyte membrane and the binder, QPAF-4(1) was used. The temperature of the cell was 60°C and the humidity was 100% RH.Results and DiscussionFig. 1 shows the water distribution inside the AEMFC parallel to the MEA. In Fig. 1(a), the left side is the anode and the right side is the cathode. In the image, three vertical blue lines are seen. The two outer lines are of a Kapton tape made of silicone. The middle line is of the water on the anode side of the AEMFC. There are blue squares on both side the anode and cathode; the bule squares on the anode side are very clear, whereas those on the cathode side are vague. These blue squares are gas flow channels. The relative thickness of water is represented by the color bar shown on the right in Fig. 1. In the dark area, water hardly existed. As the water increased, the color changed to red. Fig. 1(b) is an enlarged image of the gas flow path and the MEA. In the middle of the image, an MEA (0.55 mm thick) existed with GDLs (anode side 330 µm thick, cathode side 210 µm thick) sandwiching the CCM using a QPAF-4 membrane (30 µm thick). At the cathode, water was hardly imaged, but a small amount of water was observed in the gas flow channels near the outlet, mostly on the wall of flow channel in purple. In the GDL at the cathode, a small amount of water was observed on the CCM. At the anode, water was observed both in the gas flow channel and under the ribs. A larger amount of water was observed in the gas flow channel especially near the outlet, whereas near the inlet, a smaller amount of water was observed in the gas flow channel and under the ribs. In the gas flow channel, a large amount of liquid water was seen on the wall surface of the gas channel of the anode. Therefore, In the gas channel of the AEMFC, the gas path existed on the GDL side. Fig. 2 shows the water distribution inside the AEMFC perpendicular to the MEA. The thickness of liquid water in the gas channel and the entire MEA area at different current densities are calculated. In AEMFC, at low current density, liquid water existed near the outlet of the gas channels. As the current density increases, liquid water was also generated near the ribs and near the inlet of the gas flow channels. In PEMFC (not shown), at low current density, almost no liquid water existed in the gas channel and MEA. As the current density increased, liquid water is generated only near the gas channel outlet. The water distribution trends of PEMFC and AEMFC were significantly different. References Ono et al. J. Mater. Chem. A, 5 (47), 24804-24812 (2017)Shinohara et al., Rev. Sci. Instrum. 91, 043302 (2020). Figure 1