Short communicationCharacterization of pore network structure in catalyst layers of polymer electrolyte fuel cells
Graphical abstract
Introduction
In current and past physical models of Catalyst Layer (CL) in Polymer Electrolyte Fuel Cells (PEFCs), there has been a lack of clear CL nanostructure picture, in particular, structure and function of ionomer, nature and distribution of active interfaces, structure and composition of agglomerates, porosity, mixed wettability, and water distribution [1]. Simulations of microstructure formation, including self-organization of components in CL, predict the interplay between fabrication parameters, structure and transport properties [2], [3]. Moreover, understanding CL structure is imperative for development of novel materials and improvement of CL physicochemical properties. Nevertheless, in order to evaluate the impacts on “pseudo” steady state performance and durability, modeling efforts are needed to couple basic structural parameters obtained from modeling the CL microstructure formation to the fuel cell performance modeling [4], [5].
In order to complete basic structural aspects, the remaining issues include establishing a relationship between composition and microstructure in terms of ionomer and Pt content, pore size distribution, and pore network characteristics. The Pore Network Modeling (PNM) approach is widely used to study the transport phenomena inside porous materials and its application has increased recently in the field of Proton Exchange Membrane Fuel Cell (PEMFC) materials. The main input of the PNM is the pore size distribution (PSD). Nitrogen physisorption is a common tool used to evaluate microstructures of porous materials and obtain pore size distribution curves. The PSD can be estimated either from adsorption or desorption isotherms using modified forms of Kelvin equation for capillary condensation.
In this work, we used the PSDs extracted from the nitrogen adsorption isotherms for various catalyst layer compositions as an input parameter for the PNM. The experimentally-derived isotherms were reported in Ref. [6] and explained in the Experimental section. Then the adsorption isotherms are simulated using the model described below to reproduce the experimental isotherms. The flexibility of the PNM allows for simulation of the nitrogen adsorption process with the consideration of pores blockage due to nitrogen condensation. A comparison between the experimental and simulated isotherms validates representation of the CL using the pore network model.
Section snippets
Experimental
In a previous study by Soboleva et al. [6], Ketjen Black carbon and a catalyst powder with 46 wt% Pt on Ketjen Black from Tanaka Kikinzoku Kogyo (TKK) [1], [6] were used for N2-adsorption studies and for fabrication of the catalyst coated membranes (CCMs). All CCMs were fabricated by spray-deposition of catalyst inks on Nafion 211 membrane, used as received. CCMs with 5, 10, 30 and 50 wt% of ionomer in the CL corresponding to 0.047 mg cm−2 (0.1 I/C), 0.094 mg cm−2 (0.2 I/C), 0.375 mg cm−2 (0.8
Model development
The PNM is based on using a network of pores and throats (links between adjacent pores) (Fig. 1) to represent the structure of a porous medium [10], [11]. The pore network used in this work is constructed from spherical shape pores and cylindrical shape throats. The distance between two adjacent pores, i.e., network spacing, is constant. Thus, the pore network is regular. We use only 3D networks where every pore is connected to 6 adjacent pores (Fig. 1). The CL consists of agglomerated
Results and discussion
The results obtained with the PNM are presented in Fig. 4. In general, it is possible to distinguish three different regions in the adsorption isotherms. The effect of the PSD and in particular the amount of primary pores is evident from the low partial pressure section of adsorption isotherms (<0.05 p/p0) where the pore partial filling process is occurring (no capillary condensation). At the values of relative pressure >0.05, the amount of adsorbed nitrogen increases almost linearly, which is
Acknowledgment
The authors gratefully acknowledge partial funding from CEA, Grenoble, France, as well as the NRC-CEA collaboration program and Simon Fraser University. The CEA funding was received from the European Union's Seventh Framework Programme (FP7/2007–2013) for the Fuel Cells and Hydrogen Joint Technology Initiative under grant agreement n° [256798] (PEMICAN).
References (19)
- et al.
Electrochim. Acta
(2011) - et al.
Int. J. Hydrogen Energy
(2012) - et al.
Phys. A
(2003) Carbon
(1987)- et al.
Carbon
(1988) - et al.
Carbon
(1989) - et al.
Electrochemical Materials for PEM Fuel Cells: Insights from Physical Theory and Simulation in Modern Aspects of Electrochemistry
(2009) - et al.
J. Phys. Chem. C
(2007) - et al.
Electrocatalysis
(2011)
Cited by (34)
Removal and recovery of silver nanoparticles by hierarchical mesoporous calcite: Performance, mechanism, and sustainable application
2020, Environmental ResearchCitation Excerpt :The H3 hysteresis loop can disclose the existence of the aggregates of plate-like particles or the assemblages of slit-shaped pores (Chen et al., 2018b). The rapid increase in nitrogen adsorption at higher relative pressure (P/P0 > 0.9) further indicates the presence of an appreciable amount of secondary porosity in the hierarchical structure (El Hannach et al., 2014). The results are consistent with the FESEM observations (Fig. 1a).
PEFC catalyst layers: Effect of support microstructure on both distributions of Pt and ionomer and cell performance and durability
2020, Current Opinion in ElectrochemistryUnderstanding the interplay between cathode catalyst layer porosity and thickness on transport limitations en route to high-performance PEMFCs
2019, International Journal of Hydrogen EnergyCitation Excerpt :While the micrographs show similar morphologies for all CLs, N2-sorption porosimetry reveals significant differences in the BET surface area and pore volume (Fig. 2 and Table 2). All CLs have a BET surface area that is lower than that of the Pt/C catalyst powder (651 m2 g−1), attributed to occlusion of micropores in the carbon support by ionomer, as also reported by other researchers [27,37–39]. The thinnest CL evaluated, 3.8 μm, has the highest BET surface area, 644 m2gCcathode−1, which decreases by 45% and 63% as the CL thickness increases to 7.5 μm and 11.8 μm, respectively (Fig. 2a, and Table 2).
In-plane and through-plane non-uniform carbon corrosion of polymer electrolyte fuel cell cathode catalyst layer during extended potential cycles
2017, Journal of Power SourcesCitation Excerpt :al. have reported reverse current decay mechanism for the degradation of MEA which mimic start-stop event [20], where the difference among the several region in the MEA is not evident by the simple cross-section image. Several groups have studied the structural features of the PEFC catalyst layer pore-network using 3D tomographic image analysis [21–23]. All such image analysis involves segmentation of 2D scanning electron microscopy (SEM) images, where the pore area is distinguished from the carbon network based on the brightness difference of the pixels belonging to those areas.
Prediction of pore properties of hierarchical porous silica templated on natural rubber
2016, Microporous and Mesoporous Materials
- 1
Laboratoire de Réactivité et de Chimie des Solides (L.R.C.S.), Université de Picardie Jules Verne, CNRS UMR 7314-33 rue Saint Leu, Amiens, France.
- 2
Réseau sur le Stockage Electrochimique de l'Energie (RS2E), FR CNRS 3459, France.