Experimental Evaluation of Dominant Transport Resistances of Oxygen in Catalyst Layers of PEFC

Abstract

In polymer electrolyte fuel cells (PEFCs), the cathode catalyst layer (CL) needs a large amount of Pt because of the slow oxygen reduction reaction. Since electron, proton, and oxygen are necessary for the cathode reaction, achieving the optimum structure of electrode CL and the efficient transport of the reactants is significantly effective to reduce the usage of Pt catalyst. This study investigated the effects of cathode CL structure on the cell performance experimentally, and evaluated the dominant transport resistances of oxygen in the CL. The catalyst layers with various structures, such as various densities of Pt catalyst on carbon and various thicknesses, were fabricated. To distinguish the oxygen transport resistance in the CL from the other resistances in gas diffusion layers and channels, IV characteristics were measured with different total pressures, and the limiting current density method (1)(2) was applied. Pore size distributions (PSDs) of the CLs were also measured by nitrogen physisorption method, and the CL structures were estimated. Using a model analysis of oxygen transport resistances in the CL (1), the contributions of various oxygen transport resistances in the CL were evaluated in further detail. Table 1 shows examples of the measured structural parameters of the fabricated CLs in this study. Here, to evaluate the transport resistance at Pt surface, R Pt , Pt-supported carbons with three different Pt densities were used. The weights of carbon were set to be similar, and ionomer/carbon (I/C) ratios were 0.8. Then the Pt loadings of the CLs were 0.079, 0.20, and 0.37 mg/cm2, and the estimated Pt surface areas were 158, 349, and 475 m2 -Pt/m2 -CL, respectively. Figure 2 shows the PSDs in the CLs measured by nitrogen physisorption method for these CLs. The PSDs are similar, indicating similar structure of the CLs. The measured porosities were also similar around 0.6, and the CL thicknesses were estimated around 5 μm, as shown in Table 1. These similar structures indicate that the difference in oxygen transport resistances for these CLs is mainly caused by the different Pt loadings with different surface areas. The oxygen transport resistances in the CL distinguished using limiting current density method (1)(2) are plotted in Figure 2. The ordinate is the pressure independent resistance corresponding to the oxygen transport resistance in the CL, and the abscissa is the inverse of the Pt surface area. In the analysis model developed by the authors, the oxygen transport resistance at Pt surface is inversely proportional to the Pt surface area, and the result in Figure 2 shows that this analysis model is valid. The gradient corresponds to change in the transport resistance at Pt surface, and the intercept corresponds to sum of the other resistances. This result shows that the contribution of transport resistance of Pt surface is very large, and from the gradient, the relationship between the resistance and the Pt surface area can be determined. This study focused on three oxygen diffusion resistances in the CL: the oxygen transport resistance at the Pt surface, RPt , the oxygen dissolution resistance into the ionomer, Rdiss , and the oxygen diffusion resistances in the CL pores, Rpore . The effect of the oxygen diffusion resistances through the ionomer surrounding the carbon agglomerate were neglected based on the latest researches (1)(3). Using the model analysis, the oxygen transport resistances in the CL with various structures were evaluated. Figure 3 shows the oxygen transport resistances in the CL, (a) distinguished using limiting current density method and (b) estimated by the model analysis. In the model analysis, the structural parameters of the CL were identified using an estimation model of the agglomerate structure, developed by the authors (4). The relationship between the oxygen transport resistances and the CL structures were determined from the result in Figure 2 and the other evaluations in a similar manner. The calculated results in Figure 3(b) can simulate the experimental results in Figure 3(a) pretty well. It is confirmed from the analysis results that the contribution of the transport resistance at the Pt surface is larger. In the CL with low Pt loading and thin thickness, the contribution of the dissolution resistance into the ionomer also becomes more predominant in addition to the resistance at the Pt surface. Reference(1) T. Hayashi, et al., ECS Trans., 75(14), 373 (2016).(2) D.R. Baker, et al., J. Electrochem. Soc., 156, B991 (2009).(3) K. Kudo, et al., Electrocim. Acta, 209, 682 (2016).(4) S. Akabori, et al., ECS Trans., 64(3), 305 (2014). Figure 1