Introduction Polymer electrolyte fuel cells (PEFCs) are considered to be promising for automotive and stationary power applications. One of the most important issues for PEFCs is the reduction of the Pt loading in the cathode catalyst layers (CCLs), which generally consist of carbon-supported Pt catalysts and ionomers. Carbon blacks are widely used as the support materials. It is well known that the performance of the CCL is greatly influenced by the type of carbon black. We need to deeply understand the relationships between the microstructure and the transport properties in order to improve the performance of the low-Pt-loaded CCL. Uchida et al. investigated the relationships between the microscopic (< 10 nm) morphologies of CCLs with different carbon blacks and their performances [1]. They revealed that the existence of Pt particles on the exterior surfaces of the carbon black particles and the uniform coverage of the ionomer on the carbon surfaces had a great impact on the performance at high current density. In the present study, we investigated the mesoscopic (10–200 nm) morphologies and the proton conductivities of the CCLs with different carbon blacks using experimental and modeling approaches. Experimental We used 30 wt% Pt on Ketjen black (surface area ca. 800 m2 g−1, Pt/CB800), graphitized Ketjen black (surface area ca. 150 m2 g−1, Pt/GCB150), an acetylene black (surface area ca. 800 m2 g−1, Pt/AB800), and a second acetylene black (surface area ca. 250 m2 g−1, Pt/AB250) as cathode catalysts. CCLs were fabricated by spraying inks with the cathode catalysts and ionomers (equivalent weight = 909 g eq−1) on 30-μm thick membranes. The Pt loading and the ionomer/carbon (I/C) weight ratio were controlled to be 0.4 mgPt cm−2 and 0.8, respectively. The pore size distributions (PSDs) of the CCLs were evaluated by mercury intrusion porosimetry. Anode catalyst layers with 0.1 mgPt cm−2 Pt loading and 0.8 I/C weight ratio were fabricated by spraying inks with 50 wt% Pt on Ketjen black catalysts and the ionomers. AC impedance measurements were performed at 0.5 V with a perturbation of 2 mV from 100 kHz to 0.1 Hz at 80°C and a 100% RH hydrogen feed to the anode (reference, counter electrode) and nitrogen to the cathode (working electrode). The proton conductivities of the CCLs were calculated from the 45° line of Nyquist plots, according to the literature [2]. Modeling The mesoscopic structure models of the CCLs were generated by the following two steps. First, the porous structures were reconstructed from the FIB-SEM images with a resolution of 4 nm. Then, the solid surfaces were uniformly coated with ionomer. The amount of the coated ionomer was determined by converting the I/C weight ratio to the I/C volume ratio, assuming that the apparent volume of the carbon black was the total volume of the solid phase and nanopores less than 10 nm. The PSDs were calculated from the volume of a non-wetting fluid which was pressed into the medium. The proton conductivities were calculated by Ohm’s Law, assuming that the protons transport through the ionomer regions with a proton conductivity of 10 S m−1, which is the bulk proton conductivity of the ionomer at 80°C and 100% RH. Results and Discussion Figure 1 shows the experimental and modeling results of the PSDs. The modeling results for the Pt/CB800 and the Pt/AB800 corresponded well with the experimental results. These results suggest that the uniform-ionomer-coating model is a good approximation for the CCLs using Pt/CB800 and Pt/AB800, while there were differences between the modeling results and the experimental results for Pt/GCB150 and Pt/AB250. It is considered that the ionomer was unevenly distributed in the CCLs using Pt/GCB150 and Pt/AB250. Figure 2 shows the experimental and modeling results of the proton conductivities. The modeling results for the Ketjen black-supported catalysts underestimated the proton conductivities, while those for the acetylene black-supported catalysts were overestimated. In particular, the deviations for Pt/CB800 and Pt/AB800 were large, in spite of the correspondences in these PSDs. It was suggested that the proton conductivities were influenced not only by the mesoscopic morphologies of the CCLs but also other factors, such as the microscopic (< 10 nm) morphologies and the chemical properties of the carbon surfaces. Acknowledgements This work was partially supported by funds for the SPer-FC Project from the New Energy and Industrial Technology Development Organization (NEDO) of Japan. References Y. C. Park, H. Tokiwa, K. Kakinuma, M. Watanabe, M. Uchida, J. Power Sources, 315 (2016) 179.R. Makharia, M. F. Mathias, D. R. Baker, J. Electrochem. Soc., 152 (5) (2005) A970. Figure 1
Read full abstract