Neutron Radiography Analysis of Water Management in a Passive PEMFC with Superhydrophobic Electrosprayed Catalyst Layers

EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476301 B1701 Page 1/7 B1701 Neutron Radiography Analysis of Water Management in a Passive PEMFC with Superhydrophobic Electrosprayed Catalyst Layers L. Duque (1), E. Diaz-Alvarez (1), M. Sarma (2), T. Lappan (2), N. Weber (2), J. Lee (3), P. Boillat (3), P. García-Salaberri (4), M.A. Folgado (1), A.M. Chaparro* (1) (1) CIEMAT, Madrid/Spain; (2) Helmholtz-Zentrum Dresden – Rossendorf, Dresden/Germany; (3) Paul Scherrer Institut, Villigen PSI/Switzerland; (4) University Rey Juan Carlos, Madrid/Spain; *Contact corresponding authors: www.EFCF.com/ContactRequest Abstract Superhydrophobic catalyst layers may help to improve water transport in the porous electrodes of passive, air-breathing, proton-exchange membrane fuel cells (p-PEMFCs). Slow water transport gives rise to high water saturation in pores and becomes the main limitation for power production. By using superhydrophobic catalyst layers based on Pt/C catalyst and ionomer phase, such water saturation can be decreased [1]. Superhydrophobic catalyst layers can be obtained by electrospray deposition, resulting in macroporous and dendritic morphologies that provide superhydrophobic character [2]. In this communication, neutron transmission images are used to study water distribution in p-PEMFCs with superhydrophobic catalyst layers in anode, in cathode, and in both sides. Based on neutron attenuation, water thickness differences can be distinguished in profiles between anode and cathode (see Figure), showing a decrease in water amount in the superhydrophobic electrosprayed electrode. More homogeneous water distribution in the cell plane is also observed with electrosprayed layers. As a consequence, a change in performance is observed with polarization curves and impedance spectroscopy. Experiments carried out under different neutron incidences and cell orientations demonstrate that the superhydrophobic catalyst layers accelerate water removal from the cell and improve the performance. Cross-sectional neutron images (left), and water thickness profiles (right) with four MEAs configurations: ES: electrospray; COM: commercial electrode. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476301 B1701 Page 2/7 Introduction Passive proton exchange membrane fuel cells (p-PEMFCs) are characterized by feeding with ambient air and a static hydrogen pressure in cathode and anode, respectively [1,2]. Such operation mode requires changes in the design of cathodic and anodic plates, responsible for the clamping pressure to minimize resistive contacts in the membrane – electrode assembly (MEA). The cathodic plate must allow sufficient volume of ambient air contact and water drops elimination from the surface of the electrode, which require a design like the columnar plate proposed in a previous work [3]. The anodic plate should provide a gas-tight volume (‘anodic chamber’) able to store the static hydrogen pressure. With the appropriate configuration, the p-PEMFC can attain current and power densities comparable to those of a convective cell at the same temperature and pressure working with flowing gases [3]. Such results suppose an important gain in the specific power densities, per volume and weight, of the whole system, and makes of p-PEMFC most promising for small and portable electric applications, requiring tenths to hundreds of watts [4]. Despite the optimization attained, p-PEMFCs have still limitations to surpass power densities per active area above 200 mWcm-2 (400 mAcm-2 current density), which hinders their use in higher power applications. One principal limitation is the slow, passive, transport of water inside the porous electrodes and its elimination from the cell, which rely on weak natural forces, like the capillary diffusion in porous layers, evaporation, and dragging of water drops on the cathode surface. Larger power densities in passive operation require, therefore, not only optimization of the cell architecture but also of the transport properties of the electrodes. One possibility is the use of catalyst layers (CLs) with superhydrophobic character able to minimize the interaction of water drops with the pores walls. Superhydrophobic CLs can be prepared by using the electrospray deposition technique [5] which gives rise to macroporous structures and dendritic growths of Pt/C and ionomer chains responsible for the superhydrophobicity of the pores walls (Cassie-Baxter mode [6]). The effect of superhydrophobic CLs on water dynamics in a p-PEMFC can be tested by using techniques sensitive to liquid water inside the cell, like the neutron radiography [7–9]. 1. Scientific Approach In this work, p-PEMFCs with electrosprayed CLs have been assembled with superhydrophobic catalyst layers. Water distribution inside the cells in-operando has been observed with neutron radiography, and compared with conventional, hydrophilic, CLs in order to assess the hydrophobic effect. 2. Experiments For experimental details of the p-PEMFC and neutron radiography experiments were carried out at the NEUTRA beamline of Paul Scherrer Institut. For cell preparation and experiments configuration we refer to previous works [3,9], and communication B1009 in this conference. Passive PEMFCs (p-PEMFCs) have been studied with electrosprayed CLs in anode, in cathode, and both in anode and cathode, and compared with a p-PEMFC without electrosprayed electrodes. A summary of the studied cells is given in Table I. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476301 B1701 Page 3/7 Table I. MEA configurations of p-PEMFCs studied in this work. FCSTORE: commercial GDE with Pt/C (40wt%) + Nafion ionomer on GDL woven carbon cloth with microporous layer; ES0.05 and ES0.17: electrosprayed CLs with 0.05 mgPtcm2 and 0.17 mgPtcm-2, respectively, deposited on the membrane (Nafion 211). The GDL used with these CLs is woven carbon cloth with microporous layer (ELAT GDL LT1200W). Neutron radiography of p-PEMFCs has been carried out in-operando under frontal and lateral neutron incidences, using dry hydrogen at 0.8bar from a cannister (Hydrostik Pro, 1g capacity, Horizon), and ambient air (23 oC, 40%RH). Cells were started up from dry conditions, resulting from previous storage at 50 oC for 12h, and run in progressive current steps [B1009]. In this way, an initial ‘dry-cell’ image can be obtained, while the water observed in-operando in neutron images is fully produced by the cell. Water thickness images and water volume determination. Images of the water thickness in the cells were obtained from neutron radiographies by using Pyerre Framework (PaulScherrer Institut) that includes: 1) processing of individual neutron radiography images (1 per 10s), including dark current substraction, scattered background correction, and dryimage subtraction, to obtain the neutron transmission images; 2) merging images every 5 frames; and 3) water thickness calculation using Beer Lambert expression: 𝐼 = 𝐼0𝑒−𝑢 𝑥, (1) where I is the transmitted intensity, I0 is the incident intensity, x the water thickness, and u (= 0.496 mm-1) the experimental water attenuation coefficient [7]. The water volume in the cell was calculated from the water thickness images using ImageJ software, by selecting a region-of-interest (RoI) that excludes erroneous bright pixels generated where the neutron transmission level is too low. Surface integration of water thickness over the RoI yields the water volume in neutron images (WN). On the other hand, the water volume generated electrochemically (WFarad) is calculated from current measured during image acquisition. 3. Results Fig. 1 shows water thickness images of the four cells after faradaic generation of fixed amounts of water, 10, 50, 100, and 150 mm3. The figure shows images under frontal (Fig. 1a) and lateral (Fig. 1b) neutron incidences. Cell ref. nº Anode Cathode Membrane Configuration AL-m-01 1 FCSTORE FCSTORE Naf 211 single AL-mES-01 2 ES0.05 FCSTORE Naf 211 single AL-mES-03 3 FCSTORE ES0.17 Naf 211 single AL-mES-05 4 ES0.05 ES0.17 Naf 211 single AL-D-01 5 FCSTORE (x2) FCSTORE (x2) Naf 211 dual EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476301 B1701 Page 4/7 Fig. 1. Water thickness images of the cells in Table I, after generation of fixed amounts of water indicated on the left, under frontal (a) and lateral (b) neutron incidences. During current generation, water drops emerge at the cathode surface where they grow and evaporate. The columnar cathodic plate allows large surface mobility of the drops, that coalesce by hydrophilic interactions among them and with the hydrophilic columns of the cathodic plate. Neutron images in Fig. 1 show that the electrosprayed catalyst layers allow for more homogeneous water distribution in the cell (C2, C3, C4), compared with the commercial electrodes (C1). This effect reflects different transport properties in cells with electrosprayed CLs that favor homogeneous distribution of liquid water over the cathode surface. A closer inspection of the images under lateral incidence shows that the water accumulates in the anodic and cathodic sides of the MEA and their grid contacts. Plotting the water profiles shows two corresponding maxima in water thickness (Fig. 2a). b) a) EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476301 B1701 Page 5/7 Fig. 2. Water thickness profiles obtained on the four cells (Table I) after generation of 100 mm3 of water. a) Profile generation. b) Profiles of the p-PEMFCs with different MEA configuration (C1-C4 in Table I). Fig. 2b compares water profiles of the p-PEMFCs after 100 mm3 water generation, showing preferential water accumulation in the cathode side for C1, C2 and C4 cells, and in the anode side for C3 cell with electrosprayed electrode in the cathode. In this cell, the superhydrophobic effect of the cathodic CL decrease water saturation of the cathode which benefits more stable and higher power production in the p-PEMFC [B1009]. Integration of water thickness in neutron images provides the evolution of liquid water volume inside the cell, as plotted in Fig.3 as a function of the water volume generated by the faradaic current. Fig. 3. Plots of the water inside the cell and cell temperature, as a function of water produced by the faradaic current, for a) frontal, and b) lateral neutron incidences. (A short-circuit problem with C2 cell in a) was corrected but gave rise to high initial temperature of this cell). As expected, water volume inside the cells is below the faradaic water (indicated by the dotted line in the plots), the difference corresponding to the water eliminated from the cell. Frontal (Fig. 3a) and lateral (Fig. 3b) neutron incidences show different water contents evolutions that must be explained as a result of the different water contributions with each incidence type, as explained by Eqs. 2 and 3: EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476301 B1701 Page 6/7 Frontal incidence: 𝑊𝑁,𝐹 = 𝑊𝐶𝑎𝑡ℎ + 𝑊 𝐴𝑛 + 𝑊𝑀𝑒𝑚 = 𝑊𝐹𝑎𝑟𝑎𝑑 − 𝑊𝐸𝑣𝑎𝑝 − 𝑊𝐷𝑟𝑎𝑔 (2) Lateral incidence: 𝑊𝑁,𝐿 = 𝑊𝐶𝑎𝑡ℎ = 𝑊𝐹𝑎𝑟𝑎𝑑 − 𝑊𝐸𝑣𝑎𝑝 − 𝑊 𝐴𝑛 − 𝑊𝑀𝑒𝑚 − 𝑊𝐷𝑟𝑎𝑔 (3) where WCath, WAn, WMem, WEvap and WDrag are the water volume in cathode, anode, membrane, water evaporated from the cell and water dragged outside the cell, respectively. Hence, under frontal incidence (Fig. 3a), WN,F includes water in the cathode, anode and membrane of the cell, and excludes water evaporated or dragged from the electrode surface (Eq. 2). In this mode larger water volume with commercial electrodes (C1) reflects slower elimination rate. Electrospray electrodes (C2, C3 and C4) allow faster elimination of water either in the cathode or in the anode. For lateral incidence images (Fig. 3b), the RoI selected for integration only includes the water in the cathode of the cell and over its surface (Eq. 3). The evolution of water contents in the cathodes provides new details of the passive water management in the p-PEMFCs, showing an initial period, until faradaic production of 20-30 mm3, where no water is detected with neutrons in any of the cell. We assign this initial period to water filling the superhydrophilic and initially (quasi-)dry membrane. After the initial period, a second one shows a linear increase of WN,L with WFarad with almost same slope for all the cells, up to about 150mm3 faradaic production. This second period may reflect the filling of the cathodic GDL pores once the membrane has been saturated. By the end of this period, a divergence among the different cells is observed in water evolution. In this period, starting from 150mm3 water production, again the commercial cell, C1, accumulates larger amounts of water in the cathode. Eletrosprayed electrodes favor water elimination in this period leading to more stable and higher performance of p-PEMFCs. Conclusions Neutron transmission images have been used to study water distribution in passive PEMFCs with superhydrophobic catalyst layers in anode, in cathode, and in both sides. The calculated water thickness images show differences due to the MEA configuration. More homogeneous water distribution in the cell plane is observed with electrosprayed CLs while through-plane water profiles show a decrease in water amount in the electrosprayed CL. Water contents in the cell, calculated from integration of water thickness images, shows less water accumulation in cells with electrosprayed CLs. This effect is a result of the superhydrophobicity of electrosprayed CLs, and leads to higher and more stable performance of p-PEMFCs. Acknowledgements: Funded by PORHYDRO1 (TED2021-131620B-C21) and PORHYDRO2 (TED2021131620B-C22)/AEI European Union Next Generation EU/PRTR, Ministry of Science and Innovation of Spain. Neutron imaging was performed at the NEUTRA beamline of SINQ, PSI (P20240329). References [1] A. Heinzel, C. Hebling, M. Müller, M. Zedda, C. Müller, Fuel cells for low power applications, J Power Sources 105 (2002) 250–255. https://doi.org/10.1016/S03787753(01)00948-X. [2] J.C. Kurnia, B.A. Chaedir, A.P. Sasmito, T. Shamim, Progress on open cathode proton exchange membrane fuel cell: Performance, designs, challenges and future directions, Appl Energy 283 (2021) 116359. https://doi.org/10.1016/j.apenergy.2020.116359. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476301 B1701 Page 7/7 [3] M.A. Folgado, L. Duque, A.M. Chaparro, Air-breathing fuel cell with a columnar cathodic plate for the passive removal of water, Int J Hydrogen Energy 52 (2024) 1315–1324. https://doi.org/10.1016/j.ijhydene.2023.08.291. [4] P. Ferreira-Aparicio, A.M. Chaparro, Portable Hydrogen Energy Systems, 1st Ed., Elsevier, London, 2018. https://doi.org/10.1016/C2016-0-04605-3. [5] J. Conde, P. Ferreira-Aparicio, A. Chaparro, Electrospray Deposition: A Breakthrough Technique for Proton Exchange Membrane Fuel Cell Catalyst Layer Fabrication, ACS Appl Energy Mater 4 (2021) 7394–7404. https://doi.org/10.1021/acsaem.1c01445. [6] D. Quéré, Wetting and roughness, Annu Rev Mater Res 38 (2008) 71–99. https://doi.org/10.1146/annurev.matsci.38.060407.132434. [7] P. Boillat, V. Faucheux, D. Alincant, P. Capron, G. G, Water management in a planar air-breathing fuel cell array using operando neutron imaging, 331 (2016) 535– 543. https://doi.org/10.1016/j.jpowsour.2016.09.041. [8] P.A. García-Salaberri, D.G. Sánchez, P. Boillat, M. Vera, K.A. Friedrich, Hydration and dehydration cycles in polymer electrolyte fuel cells operated with wet anode and dry cathode feed: A neutron imaging and modeling study, J Power Sources 359 (2017) 634–655. https://doi.org/10.1016/j.jpowsour.2017.03.155. [9] A.M. Chaparro, P. Ferreira-Aparicio, M.A. Folgado, R. Hübscher, C. Lange, N. Weber, Thermal neutron radiography of a passive proton exchange membrane fuel cell for portable hydrogen energy systems, J Power Sources 480 (2020) 228668. https://doi.org/10.1016/j.jpowsour.2020.228668. Keywords: EFCF2025, H2, Low-Temp. Fuel Cells & Electrolysers, PEMFC, electrospray, passive cell, neutron imaging, water management Remark: This work is licensed under Creative Commons Attribution 4.0 International