Effects of Condensed Water on Oxygen Transport Resistance Components Separated By the Limiting Current Analysis in PEFC

Abstract

To increase the power output of PEFC, management of generated water is extremely important. In previous study, using a heat conduction and water vapor diffusion model, appropriate operating conditions were investigated to reduce oxygen transport resistance (1). However, it clarified only total oxygen transport from catalyst layer (CL) to channel without considering the effect of condensed water. In this study, the oxygen transport resistances in CL and in the other parts were evaluated quantitatively under flooding conditions. The water distribution in the cathode side micro-porous layer (MPL) and CL were observed by using a freezing method and a cryo-SEM (2).In this study, active area of the cell was 1.8 cm2. The anode and cathode flow fields were parallel straight. Two kinds of separators with 0.3 mm or 1.0 mm width rib and channel were employed (we call as 0.3 mm separator and 1.0 mm separator respectively). The anode side was supplied with hydrogen, and the cathode side was supplied with a mixed gas of oxygen and nitrogen. The gas flow rates were 100 sccm for the anode and 2000 sccm for the cathode. All experiments were conducted with 81%RH. We used gas diffusion layer (GDL) with MPL: 28 BC (made by SGL Co., Ltd.). The catalyst coated membrane was produced by Japan Gore-Tex Inc.The oxygen transport resistances were measured by using the limiting current method (3). Until now, decomposition of oxygen transport resistance is limited under dry condition because it is on the assumption that total oxygen transport resistance is constant. Then, this study introduced the separation of oxygen transport corresponding to inside and outside CL under flooding condition. Inside the CL, it was assumed that the water production rate dominates the water accumulation, which is proportional to current densities. Outside the CL, it was assumed that the product of pressure and current density determined the component of oxygen transport resistance. Based on these novel assumption, division of oxygen transport resistance was determined by a least square error calculation of actual total oxygen transport resistances and theoretical values. this experiment was conducted with different pressure (100 kPa, 140 kPa, 180 kPa), and different oxygen concentration (1%, 2%, 4%, 6%, 8%, 12%, 16%, 20%, 24 %). The cell temperature was set at 35°C and 80°C.Figure 1 shows the results of oxygen transport resistance in each inside and outside CL. I Lim is limiting current density. R P and R NP are pressure-dependent or independent oxygen transport resistances, respectively. These are corresponded to the oxygen transport resistance outside and inside the CL. In Figure 2, actual total oxygen transport resistances (R T) are shown as dots, and theoretical values which are obtained from the results of Figure 1 are shown as line. Both results are in good agreement, and the effectiveness of the oxygen transport resistance decomposition was confirmed.From the results at Figure 1, focusing on the temperature condition (1.0 mm separator), R P increases from low current density region at low temperature. But, R NP shows the similar behavior which is suppressed at low current density and increases from high current density region. These results indicate that high temperature is good at water discharge outside CL in case of present GDL.To clarify the reason why R NP increases in high current density region, effect of the difference in rib and channel widths was investigated. In case of 0.3 mm separator, R NP is relatively constant in high current density region. This result suggests that wider rib and channel increase the oxygen transport resistance in the CLFigure 3 shows the water distribution in the vicinity of MPLs at just before limiting current condition. This experiment was conducted with 80°C, and 24% oxygen concentration. No ice is observed in the MPL and CL for 0.3 mm separator (Figure 3 (a)). On the other hand, ice layer was observed at the MPL/CL interface for 1.0 mm separator (Figure 3 (b)). This is related to the result that R NP increases in high current density region in case of 1.0 mm separator (Figure 2). When wider channel separator is employed, the contact of MPL and CL becomes worse. As a result, water is accumulated at the MPL/CL interface and the water prevents oxygen transport to the CL under the channel. These results suggest that a cell structure that improves the MPL/CL interface contact is important to suppress oxygen transport resistance.Reference Kitami, et al., ECS Transactions, 92(8), 213-221 (2019).Y. Aoyama, et al., Electrochem. Commun., 41, 72-75 (2014)David A. Caulk, et al., Electrochem. Soc., 157(8), B1237-B1244 (2010) Figure 1