In recent years, clean energy vehicles have been getting more attention. Especially fuel cell vehicles (FCVs) attract garnering attention from the viewpoint of energy source diversification, zero emissions and cruising range. To popularize FCVs, fuel cell stack downsizing is needed with higher performance. In other words, high power density fuel cell stacks are required. In a polymer electrolyte fuel cell (PEFC), a challenge for high power density is reducing hydrogen and oxygen transport loss. However, it is known that the liquid water generated with Oxygen Reduction Reaction blocks oxygen or hydrogen transport. The flow field channel has capability to drain liquid water with gas flow, however, the gas diffusion layer (GDL) has a porous structure and blocks liquid water. The effective oxygen diffusion-coefficient in the GDLs decreases with increasing of the water saturation in the GDLs and may cause fuel cell performance loss. Therefore, for achieving a high power density, it is important to improve the effective oxygen diffusion-coefficient in the saturated GDLs. A lot of groups reported in this area, measuring effective oxygen diffusion-coefficient, however, generated liquid water influence to effective oxygen diffusion-coefficient is not clear quantitatively. Therefore, we developed the measuring technique for the effective oxygen diffusion-coefficient in the water-containing GDLs.

In this study, we developed unique equipment for measuring effective oxygen diffusion-coefficient that uses the oxygen sensor and the zinc-air battery, and elucidate the correlation of the effective oxygen diffusion-coefficient and the liquid water saturation in the GDLs. As a result, it turned out that the decrease of the effective oxygen diffusion-coefficient with the water saturation depended on the GDL hydrophobicity.In this study, the effective oxygen diffusion-coefficient of water-containing GDL was got by measuring the increase of the effective oxygen diffusion-coefficient and the decrease of the water saturation with the drying, and investigating correlation of them.

At the beginning, the effective oxygen diffusion-coefficient was measured by our equipment that uses an oxygen sensor and zinc-air batteries. In the equipment, the sample separates apparatus inside rom atmosphere, and the oxygen sensor and the zinc-air batteries are located independently. The zinc-air batteries are located on the circumference of the oxygen sensor, and these are connected to an electronic load. The oxygen in the apparatus is consumed by the circuit. The oxygen sensor monitors the oxygen concentration of apparatus inside, and the electronic load control the oxygen consumption rate of zinc-air batteries. As the oxygen is consumed by the zinc-air batteries, the oxygen concentration in the apparatus begins to reduce. At that moment, the oxygen diffuses into the apparatus inside by passing through the sample from the atmosphere. The oxygen concentration of apparatus inside becomes stable when the oxygen consumption rate balances with the oxygen diffusion rate. The oxygen concentration at this time is applied to the calculation of the effective oxygen diffusion-coefficient.

On the one hand, the liquid water was vacuum-impregnated into the GDL, and the water saturation was got from the measuring sample weight change in the drying process from a state filled with water.

A correlation of the effective oxygen diffusion-coefficient and the water saturation was got from a result of these measurements. And we compared the difference of the correlation by the presence of the giving hydrophobicity and confirmed the influence that hydrophobicity gave to the effective oxygen diffusion-coefficient with water-containing. In addition, the GDLs were added hydrophobicity by impregnating a carbon paper (Toray TGP-H-060) with 30wt% PTFE.

By this measurement, it was found out that it was confirmed that the effective oxygen diffusion-coefficient increased with a decrease of the water saturation in the drying process. Also, it was found out by comparing the different hydrophobicity sample that the decrease of the effective oxygen diffusion-coefficient became gentle by adding hydrophobicity. It means that hydrophobicity of the GDL affect the fuel cell performance, because of the difference of the effective diffusion-coefficient with water. Therefore, adding the hydrophobicity appropriately is one of the solutions to achieve high power density fuel cell stacks.