Interplay between PEMFC Performance, Gas Diffusion Electrode Structure and Mass Transport Properties

Abstract

Due to substantial research and development in the last two decades, energy conversion systems based on proton exchange membrane fuel cell (PEMFC) technology are emerging to the market. However, cost of the fuel cells is still high compared to the conventional energy generating systems like internal combustion engines. A decrease in Pt loading is a vital requirement for cost reduction and commercialization of the fuel cell technology. Operation of low-Pt PEMFCs showed that fuel cell performance is declining with a decrease in Pt loading and the voltage drop becomes very significant at high current density due to increasing mass transfer losses [1]. Thus, it is critical to understand this voltage loss phenomenon and mitigate it. Previously we developed a method for determining the oxygen mass transfer coefficient which is based on measurements of limiting current distributions using O2 mixtures with different diluents (from He to C3H8) [2]. In this work we studied effects of gas diffusion electrodes (GDEs) composition and structure such as cathode catalyst and microporous layer (MPL) loadings on oxygen mass transport and PEMFC performance.A segmented cell and data acquisition system were operated with commercially available 100 cm2 MEAs with cathode loadings of 0.1, 0.2 and 0.4 mgPt cm-2, while the anode Pt content was kept 0.1 mg cm-2. Cathode gas diffusion layer (GDL) was also varied to separate contributions from MPL and catalyst layer. 25BA (no MPL) and 25BC (with MPL) diffusion medias from Sigracet SGL were used for the cathode electrode. The experiments were performed at 80°C, fully humidified conditions and varied back pressures. Details of experimental procedures are presented in [2].A comparison of the IV curves showed that 25BA cannot provide good water and gas management at high current operation (Fig. 1 a). At the same time, an increase in Pt loading significantly improves the performance at low and high current regimes. The high performance of the sample MEA-0.4-25BC with Pt loading of 0.4 mgPt cm-2 and 25BC cathode GDL indicates on a possibility that part of catalyst layer acts as an MPL and actively participates in water removal from electrode structure to macro-porous layer of carbon paper.Using our previously described method we separated mass transport coefficients in gas phase (km, N2 ) and a combination of Knudsen diffusion and transport through ionomer/water films in overall GDE (kK+film ). Comparing results for MEA-25BA and MEA-25BC we extracted contribution from MPL (kK, MPL ) and cathode layer (CL) (kK+film, CL ). Further variation of back pressure allowed us to eliminate impact of Knudsen diffusion in CL and extract pressure independent contribution (kfilm,CL ) which represents a diffusion through the ionomer/water films in CL.The obtained values of pressure-independent oxygen mass transfer coefficients in CL (kfilm, CL ), Knudsen diffusion in MPL (kK, MPL ) and CL (kK, CL ) and molecular diffusion in N2 media (km, N2 ) are compared at Fig. 1 b. A reduction in Pt loading leads to a decrease in kfilm, CL as it was previously reported [3-6]. Interestingly, that MEA with the highest Pt content (MEA-0.4-25BC) is characterized by the greatest value of kK, MPL as well. All samples have nearly the same values of the O2 mass transfer coefficient in gas phase varying from 0.040 to 0.043 m s-1.Performances of the studied samples in terms of cell voltage at 1.5 A cm-2 and mass transport voltage losses (permeability and diffusion) are presented at Fig. 1 c. A direct comparison of the data shows that MEA-0.4-25BC has the best performance due to the greatest values of kfilm, CL and kK, MPL insuring a good gas transport within catalyst layer and MPL as well as water transport. Detailed analysis of the PEMFC performance, voltage losses and oxygen mass transfer coefficients at various operating conditions will be presented and discussed.ACKNOWLEDGEMENTSWe gratefully acknowledge funding from ONR (N00014-19-1-2159) and ARO (W911NF-15-1-0188). The authors are thankful G. Randolf and J. Huizingh for support of the operation as well as T. Carvalho for assisting with SEM.References A. Kongkanand, M. Mathias, J. Phys. Chem. Lett. 7 (2016) 1127.T. V. Reshetenko, J. St-Pierre, J. Electrochem. Soc., 161, F1089-F1100 (2014).Y. Ono, T. Mashio, S. Takaichi, A. Ohma, H. Kanesaka, K. Shinohara, ECS Trans. 28 (27) 69-78 (2010).N. Nonoyama, S. Okazaki, A.Z. Weber, Y. Ikogi, T. Yoshida, J. Electrochem. Soc. 158 (4) B416-B423 (2011).T. Greszler, D. Caulk P. Sinha, J. Electrochem. Soc. 159 (12) F831-F840 (2012).J.P. Owejan, J.E. Owejan, W. Gu, J. Electrochem. Soc. 160 (8) F824-F839 (2013). Figure 1