Optimizing activity and morphology of the cathode catalyst layer is one of the major challenges to overcome cost and durability of polymer electrolyte fuel cells (PEFCs). The PEFC performance is primarily determined by the losses in the catalyst layer. These are kinetic and mass-transport overpotentials associated with the sluggish oxygen reduction reaction (ORR), transport of oxygen and water flooding at high current densities.

A combination of characterization, modeling and multi-modal imaging techniques is used to understand morphology and to decouple the transport processes that limit electrocatalsyt utilization and mass transport within the catalyst layer. Hydrogen pump technique is used to measure ionic conductivity of the catalyst layers without Pt. The technique is a modified version of that previously reported [1], where a pseudo-catalyst layer is fabricated with an ionomer and carbon. The interlayer thickness is varied to plot the catalyst layer ionic resistance as a function of the layers thickness (Figure 1). The slope of the curve produces an inverse of ionic conductivity, whereas the intercept with y-axis is due to membrane and contact resistances. We varied the inter-layer thickness (three to four thicknesses), ratio of ionomer to carbon (I/C) content and the type of ionomer within the carbon black support (Vulcan XC-72). We observed an increase in ionic conductivity with increase in I/C ratio, and also PFIA ionomer at low RH showed higher conductivity compared to PFSA ionomer for the same I/C ratios. Increased ionic conductivity of PFIA ionomer explained some of the improvements in the polarization behavior of the PEFC with PFIA-containing ionomer. Hydrogen pump “DC” method is also compared to the electrochemical impedance spectroscopy (EIS) method to explain some of the differences observed in literature with DC vs AC methods [2]. To probe local ionomer interaction with the electrocatalyst a CO-displacement experiment [3] is performed on the catalyst layers with varied ionomers, I/C ratios and carbon supports. A combined rotating disc electrode (RDE) and membrane electrode assembly (MEA) studies are performed.

Electrochemical studies for ionic conductivity measurements are compared to direct imaging of ionomer distribution in the catalyst layers using nano X-ray computed tomography (CT) and spectroscopy. Current methods to image ionomer rely on staining it with heavy Cesium ions to enhance the absorption contrast imaging. However, distinguishing between the Cs- stained ionomer and Pt is challenging because of similar X-ray attenuation coefficients. We use X-ray absorption near edge spectroscopy (XANES) to image samples at Cs and Pt K-edges, enabling distinction between these two phases. Figure 1 shows images above and below Pt L3-edge, as well as volume-rendered representation of Pt and ionomer-phases. Ionomer tortuosity values obtained with imaging and hydrogen pump technique show good agreement.

Lastly, direct imaging of water formation within the catalyst layer is still challenging due to either resolution, ambient environment or field of view limitations of the imaging techniques. Over the last two years we have been developing operando nano X-ray computed tomography PEFC hardware to enable imaging with 30-60 nm resolution and 80 um field of view. Challenges of X-ray beam damage to membrane at low energies were solved by using synchrotron beamline ID 16B at ESRF with higher X-ray energy (17.5 keV) and Kirkpatrick-Baez optics. We observed no beam damage to the MEAs at these operating conditions. Complementary continuum and pore-network model bridging was conducted to explain water distributions in catalyst layers.

Acknowledgement:

This material is based upon work supported by the Department of Energy, Office of Energy Efficiency and Renewable Energy (EERE), under Award Number DE-EE0007270

References:

[1] H.O. Iden, A.; Shinohara, K., Journal of the Electrochemical Society, 156 (2009) B1078-B1084.

[2] A. Kusoglu, A.Z. Weber, Chemical Reviews, 117 (2017) 987-1104.

[3] T.R. Garrick, T.E. Moylan, V. Yarlagadda, A. Kongkanand, J Electrochem Soc, 164 (2017) F60-F64.

Figure 1. A:a) 11.7 keV image b) 11.5 keV image c) result of subtracting (b) from (a) d-f) thresholded images of respective scans. The bright spot in (a) could be either Pt or Cs. By subtracting (b) which Cs only, (c) shows Pt only B) Mass attenuation coefficient as a function of photon energy for Pt and Cs. C) Volume renderings of a) Platinum b) Ionomer c) Platinum and Ionomer. d) Cross-section image with Pt (green) and Ionomer (blue) highlighted.