Investigation of Membrane Degradation in Fuel Cells through Nanoscale Electron Tomography

Operational reliability of polymer electrolyte fuel cells (PEFCs) depends on the durability of the components used in its construction. Membrane failure is a life-limiting factor in PEFCs [1] which develops across multiple length scales. Hence, an articulated knowledge congregating known degradation mechanisms at various scales is necessary to draw a complete understanding. The molecular level degradation of perfluorosulfonic acid (PFSA) membranes in PEFCs was found to initiate with OH radicals attacking side chain C-S and α-OCF2 bonds proceeding with further side chain degradation through β-OCF2and culminating in main chain cleavage and unzipping [2]. The molecular level degradation has a profound influence on the membrane macroscale stability [3]. The physical integrity of PFSA ionomer membranes depends on their internal morphology and distribution of hydrophilic and hydrophobic phases [4]. The degradation effects at an intermediate length scale therefore requires in depth study to bridge the molecular and macroscale degradation. Our preliminary research on the mesoscale characterization of membrane degradation was conducted using 2-D electron micrographs and compositional analysis and revealed that membrane failure initiated as local damage and cracks propagated along the direction of low concentration of ion-rich regions as a result of combined chemical/mechanical degradation [4]. However, the 2-D nature of these results constrained the ability to determine hydrophilic and hydrophobic phase volume fractions and their shapes, sizes, and distribution. Therefore, the objective of the present work is to investigate the mesoscale morphology and quantify the effects of PFSA membrane degradation with PEFC operation through tomographic reconstruction of hydrophilic and hydrophobic phases in three dimensions. In this work, standard PFSA ionomer membranes were subjected to a combined chemical and mechanical accelerated stress test (AST) used for rapid benchmarking of in-situ membrane stability [3]. The chemical phase of the AST generates hydroxyl radicals that attack both the side chain and main chain of the polymer, while the mechanical stress generated by humidity cycling accelerates mechanical degradation and failure. Contrast enhanced transmission electron microscopy-tomography (TEM-t) of pristine and degraded PFSA ionomer membranes was carried out to explore the overall morphology and mesoscale features induced by the combined chemical and mechanical degradation process. In order to enhance image contrast, the sulfonic acid end group sites in the membrane were selectively exchanged with Pb ions by soaking it in saturated lead acetate solution. The samples embedded in epoxy resin were sliced to thin films (~70-90 nm) using ultra microtome and collected on a Cu grid for imaging. Nanoscale 3-D image reconstruction followed by Marker-based Watershed segmentation was used to identify weak phase boundaries [5] in the tomograms and thereby reveal the hydrophilic and hydrophobic phase distributions in the membrane. The pristine membrane exhibited a randomly interconnected hydrophilic phase with a scaffolding hydrophobic phase (Figure 1(a)), as expected from morphological theory. The membrane subjected to combined chemical/mechanical degradation was found to have similar overall morphology while containing thinner ionomer bundles and reduced hydrophilic volume fraction with smaller hydrophilic pores in regions with elevated chemical degradation as shown in Figure 1(b). Numerical analysis of the phase-segmented tomograms was performed to achieve detailed quantification of the structural properties, which was not possible with previously applied 2-D approaches and therefore contributes important new information for nanoscale analysis of fuel cell membranes. ACKNOWLEDGMENTS Research funding provided by Automotive Partnership Canada (APC), Natural Sciences and Engineering Research Council of Canada (NSERC) and Ballard Power Systems is gratefully acknowledged. We also thank Ballard Power Systems for experimental support. We thank Chan Lim, FCReL and Trevor Lancon, FEI VSG Inc. for technical support. This work made use of the 4D LABS shared facilities supported by the Canada Foundation for Innovation, British Columbia Knowledge Development Fund, Western Economic Diversification Canada, and Simon Fraser University. REFERENCES: M.P. Rodgers, L.J. Bonville, H.R. Kunz, D.K. Slattery, J. M. Fenton, Chem. Rev. 112 6075 (2012). L. Ghassemzadeh and S. Holdcroft, J. Am. Chem. Soc. 135 8181 (2013). C. Lim, L. Ghassemzadeh, F. Van Hove, M. Lauritzen, J. Kolodziej, G.G. Wang, S. Holdcroft, E. Kjeang, J. Power Sources 257 102 (2014). S. Venkatesan, C. Lim, S. Holdcroft, E. Kjeang, J. Electrochem. Soc. 163 F637 (2016). X. Zhang, F. Jia, S. Luo, G. Liu, Q. Hu, Comput. Meth. Prog. Bio. 113 894 (2014). Figure 1

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.