Fuel Cell Gas Diffusion Layers Research Articles

Communication—Oscillating Oxygen Concentration Transport through PEM Fuel Cell Gas Diffusion Layer: Variation of the Phase Angle

Recently, it was shown that performance of the cathode catalyst layer in PEM fuel cell dramatically improves if the oxygen concentration (OC) at the catalyst/ gas diffusion layer (GDL) interface oscillates in–phase with the overpotential. However, OC oscillations could only be organized in the cathode channel. Does transport through the GDL affect the phase angle of OC oscillations? Below, it is shown that at low frequencies below 1 Hz, the GDL does not change this phase angle. At high frequencies, the phase angle after transport through the GDL is −3π/4 regardless of the GDL parameters.

Hydrophobic Conductive Coatings for Fuel Cell Gas Diffusion Layers

Mass transfer of the reactant gases, water management, thermal and electrical conductivity are all highly depending on the properties of the gas diffusion layer (GDL) in polymer electrolyte fuel cells (PEFCs). GDLs are produced in complex and energy demanding processes including high temperature graphitization and are modified using polytetrafluoroethylene (PTFE) to introduce the hydrophobicity needed to efficiently remove water [1]. PTFE is, however, negatively impacts the required electrical conductivity and is facing an uncertain future with the pending restrictions on the use of perfluoroalkyl substances (PFAS) [2], [3]. Therefore, we have investigated the use of electrically conductive polyaniline (PANI) coatings on carbon fiber substrates with modifications to increase hydrophobicity. As substrate, AvCarb carbon paper EP40 was used and PTFE treated EP40T was used as state-of-the-art reference. EP40 was coated through dip-coating or electro-coating using PANI modified with sodium dodecylbenzenesulphonate as additive to adjust hydrophobicity [4]. With this approach, we have been able to show comparable pore volumes and pore surface areas (Figure 1) through mercury porosimetry as well as comparable hydrophobic properties.

Measuring Thermal Conductivity of Single Thickness Fibrous Gas Diffusion Layers Using Transient Plane Source Method

In a proton exchange membrane fuel cell (PEMFC), the gas diffusion layer (GDL) is a porous material placed between the flow field plates and the catalyst layer. Its primary responsibility is to distribute reactants evenly across the surface of the catalyst layer while allowing for the removal of excess water and heat produced during operation. The GDL is typically made of a carbon-based material, such as carbon paper or carbon cloth, which is treated with polytetrafluoroethylene (PTFE) to become hydrophobic. This treatment allows it to repel water and facilitate the transport of gases such as hydrogen and oxygen during operation. The thermal conductivity of the GDL plays a crucial role in managing the heat and water of PEMFCs.The Guarded Heat Flow (GHF) technique is a well-established steady-state method for measuring the thermal conductivity of carbon fiber papers. This is achieved by testing two different thicknesses of a material that have an identical structure [1-3]. Using this two-thickness method, as per ASTM standard C177-19e1 [4], enables the bulk thermal conductivity to be deconvoluted from the thermal contact resistance at the interface between the sample and the measurement device's surface (an interfacial phenomenon). However, materials with identical structures for two-thickness measurements are not always available for commercial GDLs. In this study, we compare the transient plane source (TPS) thin film method, which measures the thermal conductivity of single thickness GDL samples, with the two-thickness GHF method for measuring thermal conductivity.In this study, GDL samples with the same bulk structure and manufactured in two thicknesses were used: Sigracet series 24BA and 34BA with original thicknesses of 190 µm and 280 µm, respectively. A custom-built Thickness Under Compression and Resistance Under Compression machine (TUC-RUC) was utilized to measure the thickness of the samples under loads ranging from 0.35 MPa to 2 MPa. The thickness under compression was measured for five samples (25.4 mm diameter) of each type of GDL. The standard deviation for each sample thickness was below 4%. As shown in Fig. 1, the stress-strain response of the GDL samples was measured and analyzed to assess the similarity in their mechanical structures. Consistent with the assumptions that both samples have the same structure, the GDL samples with two different thicknesses displayed identical stress-strain behavior under compression, as shown by the curve-fit for stress-strain behavior in proposed Fig. 1.The thermal conductivity of the samples was measured using both the TPS thin films method, as per ISO 2207-2:2022 [5], and the GHFM. On the TPS, three samples of each type of GDL were measured three times. The measured thermal conductivities increased with compression, ranging from 0.15 to 0.47 W/mK, with a standard deviation of less than 5% for each measurement. The results from the TPS technique for a single thickness of GDL were compared with the two-thickness measurements conducted using the GHFM, as shown in Figure 2. The GHFM tests were conducted on three samples of each type of GDL at five compressive loads, and the standard deviation on the deconvoluted thermal conductivity was less than 7% for each measurement. The values measured by the two techniques were found to be in good agreement, with a relative difference of within 6%. This study demonstrates that the transient plane source (TPS) method, when combined with an accurate thickness measurement technique, is a precise technique for determining the bulk thermal conductivity of commercial GDLs produced in a single thickness.[1] H. Sadeghifar, N. Djilali, and M. Bahrami, “Effect of Polytetrafluoroethylene (PTFE) and micro porous layer (MPL) on thermal conductivity of fuel cell gas diffusion layers: Modeling and experiments,” Journal of Power Sources, vol. 248, pp. 632–641, Feb. 2014, doi: 10.1016/j.jpowsour.2013.09.136.[2] M. Ahadi, M. Andisheh-Tadbir, M. Tam, and M. Bahrami, “An improved transient plane source method for measuring thermal conductivity of thin films: Deconvoluting thermal contact resistance,” International Journal of Heat and Mass Transfer, vol. 96, pp. 371–380, May 2016, doi: 10.1016/j.ijheatmasstransfer.2016.01.037.[3] E. Sadeghi, N. Djilali, and M. Bahrami, “Effective thermal conductivity and thermal contact resistance of gas diffusion layers in proton exchange membrane fuel cells. Part 1: Effect of compressive load,” Journal of Power Sources, vol. 196, no. 1, pp. 246–254, Jan. 2011, doi: 10.1016/j.jpowsour.2010.06.039.[4] “Test Method for Steady-State Heat Flux Measurements and Thermal Transmission Properties by Means of the Guarded-Hot-Plate Apparatus.” ASTM International. Doi: 10.1520/c0177-19e01.[5] “ISO22007-2, Plastics-determination of thermal conductivity and thermal diffusivity-part 2: transient plane heat source (hot disc) method”, 2022. Figure 1

Advanced boron-doped carbon papers with excellent electrical conductivity and low graphitization temperature for PEM fuel cells

A facile method is proposed in this work to prepare a carbon paper (CP) with outstanding properties by doping boron to catalyze graphitization at lower temperature. Boron was introduced into CP by impregnating carbon fiber preform into a boron/phenolic resin/ethanol solution, and then CP was obtained after the consecutive procedures of hot pressing, carbonization, and graphitization. Morphology, microstructure, composition, and physical properties of the doped and undoped CPs were thoroughly investigated via various techniques to exam the effects of boron concentration and graphitization temperature. Furthermore, the boron doped CP was fabricated into fuel cell gas diffusion layer (GDL) and membrane electrode assembly (MEA) to evaluate its performance through the single cell performance. The findings indicated CP doping with 2 wt% boron could lower the graphitization temperature by at least 200 °C. The B-doped GDL had a higher power density in comparison with the undoped GDLs made in our lab or from the commercial brands.

Biomimetic and flexible high-performance carbon paper prepared by welding three-dimensional carbon fiber network with polyphenylene sulfide spherical sites for fuel cell gas diffusion layer

High-performance carbon fiber paper (CFP) is usually an important part of the fuel cell gas diffusion layer (GDL). However, the preparation of CFP for GDL is still a challenge due to its arduous preparation process. Here, inspired by the structure of bird's nest in nature, we present the original design of PPS spherical sites welding three-dimensional (3D) carbon fiber network to produce high-performance CFP by facile wet papermaking and heat treatment. With a 3D CF network as the skeleton and PPS spherical sites as the welding points, CFs/PPS composite paper exhibits good mechanical strength and outstanding flexibility. Meanwhile, the excellent conductivity (6.8 S/cm) is integrated with super-hydrophobic performance, electrothermal performance (Tmax = 365 °C), heat resistance (DTGmax>575 °C), and corrosion resistance, becoming a high-performance multifunctional CFP. The mythical combination of these properties makes CFs/PPS composite paper an ideal candidate in the field of fuel cell GDL. In addition, the evolution of chemical structure of CF/PPS during carbonization is emphasized, which is of great significance for the properties of composite paper. This work presents a leap forward in the preparation of multi-property high-performance CFP in a simple method and is a key step toward exploring how the structural changes of CFs/PPS composite paper can be understood during carbonization.

Determining local transport properties of gas diffusion layer land-channel regions via pore network modelling

Targeted, pre-determined values of spatially-resolved liquid water saturation is achieved in the pore networks of polymer electrolyte membrane (PEM) fuel cell gas diffusion layers (GDL). An inlet selection algorithm is presented, where the local saturation of the overall substrate, channel region, and land region match the prescribed input values (within a tolerance of 2% absolute error). With this algorithm, we conduct a novel, comprehensive characterization of the influence of local saturation on local oxygen transport properties using 1721 unique saturation conditions within the GDL. The oxygen transport properties of the GDL are significantly affected by the spatial distribution of liquid water. We find that liquid water accumulation in the channel region leads to an increase in the oxygen transport resistance in the channel region of the substrate and overall substrate. The channel region saturation is a better predictor of substrate transport properties than the substrate saturation alone. The correlation between the land region saturation and the oxygen transport resistance of the substrate is weak and depends on the spatial distribution of liquid water in the land region. Land region saturation only has a strong influence on the oxygen transport resistance in the land region of the substrate.

Water cluster characteristics of fuel cell gas diffusion layers with artificial microporous layer crack dilation

The formation of discrete water clusters in polymer electrolyte fuel cell gas diffusion layers (GDL) can lead to increased resistance for oxygen transport in the catalyst layer. This study investigates the effect of MPL crack propagation on the water cluster development in a X-ray computed tomography (CT) microstructure using the volume-of-fluid method (VoF). The VoF calculation was compared to operando CT data by voxel matching, obtaining a maximum 88 % accuracy. Using 3D contact angle extraction, the local scale heterogeneous wettability in the GDL was investigated. In a simulation study, MPL cracks were created as the boundary sources for water and the effect of increasing the area fraction covered by cracks on the water distribution in the GDL was investigated. The increased cracking, created larger discrete water clusters in the GDL with greater connectivity, due to in-plane coalescence. The in-plane movement leads to coalescence of clusters, forming fewer, larger clusters at later times close to breakthrough to the channel. This phenomena is shown by the decrease in water cluster density (n mm−2) from 10 to 5. This immobile water impacts the distribution of oxygen at the catalyst layer (10 % local difference) and therefore the current density distribution.

Investigating the Influence of Humidity on Liquid Water Transport Mechanisms in Fuel Cell Gas Diffusion Layers Using Operando X-Ray Computed Tomography

The intermittent nature of many renewable energy sources, such as solar and wind, presents the need to acquire on-demand renewable energy either through energy storage or energy production methods. The polymer electrolyte membrane (PEM) fuel cell is quickly becoming an attractive solution as it can produce high power densities under a rapid change in load while producing zero local carbon emissions [1] and is ideal for large-scale applications, such as automotive [2]. However, barriers to their widespread implementation are attributed to the high costs associated with mass transport losses at high current density operation due to inefficiencies in liquid water management in the gas diffusion layer (GDL) [3][4]. Understanding the relationship between liquid water distributions in the cathode GDL and transport properties of PEM fuel cells under varying operating conditions can lead to improved GDL configurations for improved performance. Previously, studies have been conducted to characterize the effect of operating temperature on liquid water pathways in GDLs by using 3D imaging techniques on operando PEM fuel cells [5]. High-speed, high-resolution 3D imaging techniques, such as computed tomography (CT), enable the visualization of dynamic pore-scale effects and are advantageous in elucidating transport mechanisms in the GDL.In this work, the effect of operating relative humidity on the formation of liquid water pathways in the GDL will be explored by imaging an operando PEM fuel cell with synchrotron X-ray CT. To date, a custom PEM fuel cell has been developed in-house which features a novel design with two rotary unions on either side of the cell. The rotary unions enable continuous rotation of the cell during imaging, and thus are central to achieving high temporal resolution for visualizing the development of preferential water pathways. High spatial resolution (pixel resolution of 1.44 microns per pixel [6]) is also obtained with the synchrotron beam, which allows for the visualization of liquid water in individual pores at the microscale in the GDL. Visualizing water in the micropores of the GDL can help us further understand water transport mechanisms within the complex structure of the GDL to ultimately develop methods to expel excess water efficiently. In the next steps of this research, we will use the results from electrochemical testing and the reconstructed images of the fuel cell to draw important conclusions about the relationship between liquid water distributions in the GDL and fuel cell performance. As well, the effect that operating relative humidity has on this relationship will be determined. The aim of this work is to inform optimal design for GDL materials for improved PEM fuel cell performance, which will ultimately accelerate their utilization on a global scale as a reliable and sustainable energy source.[1] P. Shrestha, CH. Lee, K. F. Fahy, M. Balakrishnan, N. Ge, and A. Bazylak, “Formation of Liquid Water Pathways in PEM Fuel Cells: A 3-D Pore-Scale Perspective,” Journal of The Electrochemical Society, vol. 167, no. 5, p. 054516, Jan. 2020, doi: 10.1149/1945-7111/ab7a0b.[2] S. Park, J. W. Lee, and B. N. Popov, “A review of gas diffusion layer in PEM fuel cells: Materials and designs,” International Journal of Hydrogen Energy, vol. 37, no. 7, pp. 5850– 5865, Apr. 2012, doi: 10.1016/J.IJHYDENE.2011.12.148.[3] Y. Nagai et al., “Improving water management in fuel cells through microporous layer modifications: Fast operando tomographic imaging of liquid water,” Journal of Power Sources, vol. 435, p. 226809, Sep. 2019, doi: 10.1016/J.JPOWSOUR.2019.226809.[4] U. U. Ince et al., “3D classification of polymer electrolyte membrane fuel cell materials from in situ X-ray tomographic datasets,” International Journal of Hydrogen Energy, vol. 45, no. 21, pp. 12161–12169, Apr. 2020, doi: 10.1016/J.IJHYDENE.2020.02.136.[5] Hong Xu, Shinya Nagashima, Hai P. Nguyen, Keisuke Kishita, Federica Marone, Felix N. Büchi, Jens Eller, Temperature dependent water transport mechanism in gas diffusion layers revealed by subsecond operando X-ray tomographic microscopy, Journal of Power Sources, Volume 490, 2021, 229492, ISSN 0378-7753, https://doi.org/10.1016/j.jpowsour.2021.229492.[6] Canadian Light Source, “Detectors,” BMIT. [Online]. Available: https://bmit.lightsource.ca/tech-info/detectors/. [Accessed: 05-Apr-2022].

Deformation Mechanics of Fuel Cell Gas Diffusion Layer: Cyclic Response and Constitutive Model

The deformation mechanics of a typical gas diffusion layer using experimental and advanced modelling technique is reported. The experimental cyclic response is observed similar to pseudo-elastic materials with highly nonlinear loading/unloading. The cyclic compressive mechanical response of gas diffusion layer (GDL) is modelled to be the outcome of cumulative changes in deformation kinematics of matrix and fiber fractions. The individual mechanisms necessitating the energy dissipation, residual strain, and stress softening during cyclic mechanical response are related to nonlinear hyperelastic matrix with the damage function and inelastic activation function at the interface of constituents. The model predicts highly nonlinear elastic loading, residual strain, hysteresis, and damage quotient associated with stress softening as a function of several cycles. The significant takeaway from this study is in terms of quantifying strength, inelastic nature of individual constituents. The proposed model is simulated for low-level altering stresses of up to twenty cycles. The results show the build-up of residual strains and hysteresis as a function of fuel cell clamping pressure.

A 3D convolutional neural network accurately predicts the permeability of gas diffusion layer materials directly from image data

Three dimensional convolutional neural networks (3D CNNs) were used to accurately predict the permeability of fuel cell gas diffusion layer (GDL) materials directly from 3D binary image data. A straightforward 3D CNN architecture was trained and evaluated on a dataset of 7500 numerically generated GDL materials using five-fold cross validation. The permeability of the generated GDL materials was calculated using pore network modelling and comported with literature reported values. The 3D CNN converged to acceptable prediction error quickly and with minimal hand tuning of hyperparameters. The trained 3D CNN had low prediction error when compared to contemporary models performing similar tasks, with an average mean absolute error (MAE) of 4.660E-13 m2 and an R2 of 0.9885, when evaluated on test fold materials. Lastly, a learning curve analysis was also performed.

Effect of In-Pore Wettability on Mass Transfer Performance of Fuel Cell Gas Diffusion Layer

The gas diffusion layer (GDL), as the main mass transfer carrier in a hydrogen fuel cell, transports fuel and discharges water, the only by-product of the electrochemical reaction. The dispersion process of water in the pore will hinder the diffusion of gas, thus reducing the concentration of fuel gas at the catalytic site, resulting in the decrease of the electrochemical reaction rate. In this paper, the effect of wettability in the GDL hole on the water transport process is studied. When the pore wall is hydrophilic, the liquid phase is affected by the gas phase eddy current velocity field, and the particles at the center advance to the edge, forming a liquid phase interface with a thin center and thick edge. With the increase of the wall contact angle, the curvature of the three-phase interface increases, the wall adhesion decreases, and the liquid phase is more likely to be discharged. When the contact angle is 130°, the liquid phase almost does not shift in the hole with a radius of 5 μm. With the increase of the radius or inlet pressure difference, the liquid phase is discharged gradually, and the discharge rate of the liquid phase is only related to the wettability of the wall.

A phase-dependent constitutive model to predict cyclic electrical conductivity in fuel cell gas diffusion media

Structure-property relation in fuel cell gas diffusion layer (GDL) is a dependent function of its constituents. The bulk electrical conductivity of these layers is known to be relative density function varying due to external force or cell operating conditions. To locally predict the changes due to complex working conditions, an accurate model that predicts the nonlinearity of GDLs is highly desirable. To this end, this article proposes a material model that is phenomenologically derived to address the cyclic electrical conductivity of GDLs. Functional variables are taken to operate on porosity variation, fiber contact density, and fiber dislocation parameters. In the presence of these parameters, the results illustrate nonlinear conductivity variation with the magnitude of applied cyclic compressive load. Through successive loading-unloading, the porous structure is modeled to reach a steady-state reflecting stable conductivity-stress behavior for the constant stress limit. An interesting behavior of GDL can be captured where conductivity reduces as compressive load exceeds a threshold limit called break stress due to fiber breakages or dislocations. A greater applicability of this model may lie in mapping localized in-situ response of GDLs under cyclic operations.

Effect of graphitization degree of fuel cell gas diffusion layers on their heat management: Modeling and experiments

Effect of graphitization degree of fuel cell gas diffusion layers on their heat management: Modeling and experiments

Non-Solvent Induced Phase Separation: A Versatile Synthetic Method for High Performance Redox Flow Battery Electrodes

Carbonaceous porous electrodes are ubiquitous to advanced electrochemical systems where they are responsible for multiple critical functions in the cell related to thermodynamics, kinetics, and transport, including providing surfaces for electrochemical reactions, conducting electrons and heat, and distributing fluids. Thus, their design governs the performance, durability, and consequently, the cost of these systems. However, there is limited knowledge about how to deterministically design these electrode materials which forces the repurposing of materials that have not been tailored for the specific application (e.g. the use of polymer electrolyte fuel cell gas diffusion layer as redox flow battery porous electrode1,2). Our current arsenal of materials is limited to fibrous electrodes which are manufactured using various mechanical methods (e.g. paper making, weaving, hydro-entangling) resulting in idiosyncratic structures such as papers, cloths, and felts. These fabrication methods involve multiple complex subprocessing steps impacting the final manufacturing cost and offering limited versatility to control the electrode microstructure and surface composition, which ultimately limits the performance of the electrochemical cell. Thus, there is a need to develop novel material sets with precise control over microstructure and composition while employing synthetic methods that are compatible with large scale manufacturing. In this work, we focus our efforts on designing tailored electrodes for redox flow batteries (RFBs), which are promising for grid-scale energy storage3 if their costs can be significantly reduced4.Here, we introduce the non-solvent induced phase separation (NIPS) as a simple and versatile fabrication method for carbonaceous porous electrodes for redox flow batteries5. Drawing inspiration from membrane technology, the NIPS method has been leveraged to synthesize morphologically-diverse microstructures (e.g., isoporous, macrovoids, porosity gradient) which are appealing to electrode manufacturing6. A polymer solution, containing polyacrylonitrile (PAN, carbon-containing) and polyvinylpyrrolidone (PVP, pore-forming agent) dissolved in N,N-dimethylformamide (solvent) was casted in a mold and subsequently immersed in water (non-solvent). Finally, the polymeric scaffold is carbonized under inert conditions to form a conductive network. Easily adjustable parameters, such as solvent type, polymer concentration and temperature enable control of the final electrode microstructure. To elucidate material synthesis-property-performance relationships, we vary the polymer content, polymer ratio, and solvent type and perform microscopic, spectroscopic, and electrochemical characterization techniques over the synthetized electrodes. Microstructural characterization revealed a multimodal pore size distribution composed of fine, interconnected microvoids (pore diameter 2-15μm) coupled with through plane, finger-like macrovoid channels (throat diameter > 50 μm) forming honeycomb networks. The unique microstructure, not attainable with traditional carbon-fiber manufacturing techniques, enables large surface area at the membrane-electrode interface and fast electrolyte replenishing which reduces mass transfer resistance within the electrode. We performed flow cell tests with an aqueous iron single electrolyte which revealed a considerable reduction of the charge transfer and mass transfer overpotentials of the novel electrodes compared to the commercial baseline at the expense of a slight increase in pressure drop. In the final part, we demonstrate the use of NIPS-electrodes in a full all-vanadium RFB which results in a significant improvement in power density compared to the baseline material, which can be attributed to reductions in the charge transfer and mass transport overpotentials. Although nascent, NIPS emerges as a promising platform to engineer porous electrodes for RFBs and other convection-enhanced electrochemical systems. References Mathias, M. F., Roth, J., Fleming, J. & Lehnert, W. Diffusion media materials and characterisation. Handb. Fuel Cells 3, 517–537 (2010).Forner-Cuenca, A. & Brushett, F. R. Engineering porous electrodes for next-generation redox flow batteries: recent progress and opportunities. Curr. Opin. Electrochem. 18, (2019).Weber, A. Z., Mench, M. M., Meyers, J. P., Ross, P. N., Gostick, J. T. & Liu, Q. Redox flow batteries: A review. J. Appl. Electrochem. 41, 1137–1164 (2011).Darling, R. M., Gallagher, K. G., Kowalski, J. A., Ha, S. & Brushett, F. R. Pathways to low-cost electrochemical energy storage: a comparison of aqueous and nonaqueous flow batteries. Energy Environ. Sci. 7, 3459–3477 (2014).Wan, C. T. C., Jacquemond, R. R., Chiang, Y. M., Nijmeijer, K., Brushett, F. R. & Forner-Cuenca, A. Non-Solvent Induced Phase Separation Enables Designer Redox Flow Battery Electrodes. Adv. Mater. 2006716, (2021).Young, T. H. & Chen, L. W. Pore formation mechanism of membranes from phase inversion process. Desalination 103, 233–247 (1995).

Influence of the surface microstructure of the fuel cell gas diffusion layer on the removal of liquid water

The effective removal of water on the gas diffusion layer (GDL) surface and low flow channel resistance are essential for the water management of proton exchange membrane fuel cells (PEMFCs). In this paper, a 3D two-phase volume of fluid (VOF) model is used to compare and analyze the influence of different GDL surface microstructures on the liquid hydrodynamic behavior and optimize the design of the sine wave microstructure. The results show that the surface microstructure of the GDL has a more significant impact on the water removal and flow resistance coefficient in the flow channel, and the sine wave microstructure has substantial advantages. The sine wave peak and period significantly influence the water removal and flow resistance coefficient. As the peak increases, the average relative change rate and the flow resistance coefficient also increase; the influence of the period is opposite to the peak, and the continuous decrease of the period will accelerate the water removal in the flow channel. The sine wave's height and width have little impact on water removal and the flow resistance coefficient. When the sine wave A = 75 μm, T = 1.5, H = 15 μm, and L = 25 μm, good flow channel water removal and low resistance are achieved. This work has particular guiding significance for removing liquid water on the GDL surface and obtaining low flow channel resistance.

Operando Liquid Pressure Determination in Polymer Electrolyte Fuel Cells.

Extending the operating range of fuel cells to higher current densities is limited by the ability of the cell to remove the water produced by the electrochemical reaction, avoiding flooding of the gas diffusion layers. It is therefore of great interest to understand the complex and dynamic mechanisms of water cluster formation in an operando fuel cell setting as this can elucidate necessary changes to the gas diffusion layer properties with the goal of minimizing the number, size, and instability of the water clusters formed. In this study, we investigate the cluster formation process using X-ray tomographic microscopy at 1 Hz frequency combined with interfacial curvature analysis and volume-of-fluid simulations to assess the pressure evolution in the water phase. This made it possible to observe the increase in capillary pressure when the advancing water front had to overcome a throat between two neighboring pores and the nuanced interactions of volume and pressure evolution during the droplet formation and its feeding path instability. A 2 kPa higher breakthrough pressure compared to static ex situ capillary pressure versus saturation evaluations was observed, which suggests a rethinking of the dynamic liquid water invasion process in polymer electrolyte fuel cell gas diffusion layers.

Electrochemical Residence Time Distribution As a Diagnostic Tool for Electrodes in Redox Flow Batteries

Redox flow batteries (RFBs) are a burgeoning technology for use in grid-scale energy storage; however, broad adoption is stymied by large capital costs.1,2 Housed within the reactor, porous electrodes are critical components to system performance and cost, as they provide surfaces for redox reactions, void space for electrolyte passage, and conductive solids for electron transport.3 Carbon papers, cloths, and felts are commonly used in RFBs, despite their tailored design and function towards the governing physics as fuel cell gas diffusion layers. Consequently, when coupled with geometrically varying flow field structures, these carbon substrates underperform in RFBs with forced liquid advection, resulting in uneven flow and current distributions. These inhomogeneities lead to inefficient operation, motivating the investigation of electrode morphologies and flow conditions that enable more uniform reactions throughout the reactor.One promising approach to assessing electrode performance is the measurement of residence time distribution, a traditional chemical reactor efficacy metric for quantifying flow pathways and hydraulic short-circuiting.4 Currently, most residence time estimations in flow cells assess the fluid dynamic spreading of a known solution leaving a reactor, using both experiments and computational fluid dynamics.5 However, in the absence of complementary electrochemical performance measurements, hydrodynamic calculations can only serve as a proxy for possible surface accessibility. Additionally, sole measurements of the reactor effluent are susceptible to added variance, which can obscure experimental results6 as opposed to spatial quantities within the reactor.Herein, we present an electrochemical method to measure spatially averaged electrochemical residence times in a small-scale flow cell. Using two disparate flow fields each with carbon cloth and paper electrodes, we compare electrochemical residence time responses to conventional steady-state polarization measurements to capture performance scaling relationships. Further, we present complementary modeling approaches, depicting multi-scale residence time calculations to contextualize the experimental values. Together, these approaches have the potential to elucidate the effects of electrode microstructure and flow field selection on cell-level fluid dynamics and electrochemical performance, which, in turn, may be used to advance RFB reactor design.AcknowledgmentsThis work was funded by the Joint Center for Energy Storage Research, an Energy Innovation Hub of the U.S. Department of Energy, Office of Science, Basic Energy (De-AC02-06CH11357). K.M.T. recognizes additional support from the U.S. NSF Graduate Research Fellowship (1122374).References A. Z. Weber et al., J. Appl. Electrochem., 41, 1137 (2011).C. Minke, U. Kunz, and T. Turek, J. Power Sources, 361, 105–114 (2017).K. J. Kim et al., J. Mater. Chem. A, 3, 16913–16933 (2015).A. E. Rodrigues, Chem. Eng. Sci., 230, 116188 (2021).G. Aparicio-Mauricio, F. A. Rodríguez, J. J. H. Pijpers, M. R. Cruz-Díaz, and E. P. Rivero, J. Energy Storage, 29, 101337 (2020).A. Bérard, B. Blais, and G. S. Patience, Can. J. Chem. Eng., 98, 848–867 (2020).

Predicting through-Plane Porosity Profiles of Fibrous Porous Media from 2D Images with a Single-Input Multi-Output Convolutional Neural Network

While significant advances have been made in polymer electrolyte membrane fuel cells (PEMFC), water management in the gas diffusion layer (GDL) continues to be an important avenue for increasing overall cell efficiency (1). Efforts have been made to better understand how the structure of the GDL impacts water management through advanced microstructure characterization techniques such as synchrotron x-ray and laboratory-based computed tomography imaging, as well as neutron imaging (2-4).Deep learning tools like convolutional neural networks (CNNs) are openly available resources and have been employed for image classification and object detection since the late 1990s (5). While CNNs are less often used for real number regression tasks, they present a unique opportunity to gain meaningful insight into the GDL microstructure from lower quality data sources. Recently deep learning has been implemented in geological porous media applications to predict morphological, hydraulic, and mechanical properties with good success, and should be further investigated for use in the fuel cell community (6).In this study, a database containing over 2200 3D fibrous porous materials was created. The materials exhibited porosities ranging from 40% to 95% and were designed to represent GDLs in a PEMFC. The materials were used to create 2D images for training CNNs to predict average porosities and through plane porosity profiles. The CNNs, based on the popular ResNet50 and Xception network architectures, were adapted for real number regression rather than for traditional classification tasks. Both architectures accurately predicted average porosities with an R2 of 0.98 (ResNet50) and 0.99 (Xception). Xception was then further adapted into a single-input multi-output convolutional neural network (SiMo CNN) and trained to predict through-plane porosity profiles using only 2D images. We achieved good results with the SiMo CNN for predicting porosity profiles with an R2 of 0.91 and a mean absolute error of 1.7%. This study illustrates the usefulness of CNNs in image analysis of fibrous porous materials like the GDL and highlights the potential for CNNs to be further applied in the design and characterization of materials for electrochemical energy conversion.References Ijaodola OS, El-Hassan Z, Ogungbemi E, Khatib FN, Wilberforce T, Thompson J, et al. Energy efficiency improvements by investigating the water flooding management on proton exchange membrane fuel cell (PEMFC). Energy. 2019;179:246-67.Ince UU, Markötter H, George MG, Liu H, Ge N, Lee J, et al. Effects of compression on water distribution in gas diffusion layer materials of PEMFC in a point injection device by means of synchrotron X-ray imaging. Int J Hydrogen Energy. 2018;43(1):391-406.Battrell L, Patel V, Zhu N, Zhang L, Anderson R. Imaging of the desaturation of gas diffusion layers by synchrotron computed tomography. J Power Sources. 2019;416:155-62.Siegwart M, Harti RP, Manzi-Orezzoli V, Valsecchi J, Strobl M, Grünzweig C, et al. Selective visualization of water in fuel cell gas diffusion layers with neutron dark-field imaging. J Electrochem Soc. 2019;166(2):F149.LeCun Y, Bottou L, Bengio Y, Haffner P. Gradient-based learning applied to document recognition. Proc IEEE. 1998;86(11):2278-324.Rabbani A, Babaei M, Shams R, Da Wang Y, Chung T. DeePore: a deep learning workflow for rapid and comprehensive characterization of porous materials. Adv Water Resour. 2020;146:103787.

Numerical investigation on the sensitivity of endplate design and gas diffusion material models in quantifying localized interface and bulk electrical resistance

A localized non-intuitive relationship between electrical interface contact resistance and bulk properties such as bulk electrical resistance and permeability in the fuel cell gas diffusion layer (GDL) is reported. A numerical method is adopted to investigate contact pressure and hence the interface contact resistance at the interfaces of bipolar plate (BPP)|GDL and GDL|Polymer electrolyte membrane (PEM). The results are observed to be sensitive to GDL material models as well as endplate designs. This means, endplates designed to improve the electrical contact resistance or contact pressure at the BPP|GDL interface may not necessarily assure an improvement in bulk properties, in fact, it is observed in this study that these properties are inversely related. Further, a differential deformation in GDL along with consolidation effect is predicted with compressible version of hyperelastic material model. More importantly, it is revealed that the selection of material models plays a significant role in the deformation behaviour of the GDLs irrespective of the clamping design adopted.

Study of the anisotropic permeability of proton exchange membrane fuel cell gas diffusion layer by lattice Boltzmann method

In proton exchange membrane fuel cells (PEMFCs), the gas diffusion layers (GDLs) are located between the flow channel and the catalytic layer, which not only provides a transport channel for the reaction gas and the output water, but also its fibrous skeleton is the key structure for the electron load transfer, and its transport performance is an important factor that affects the performance of the cell. In this paper, GDLs are stochastically reconstructed and the Lattice Boltzmann method (LBM) is adopted to study permeability. The changes of through-plane permeability and in-plane permeability are studied by varying the porosity, fiber diameter, and pressure drop. It is found that the porosity plays a dominant role in the variation of permeability, and the through-plane permeability at a porosity of 0.85 is 24.78 times greater than that at 0.5. The influence of fiber diameter can not be ignored. When the diameter is larger than 8 μm, the influence on permeability is more obvious, and the permeability is no longer sensitive to changes in diameter after a certain degree of porosity has been reached. The pressure drop only changes the magnitude of the average cross-sectional velocity of the gas diffusion layer, and the effect on the permeability is negligible. The in-plane permeability is larger than the through-plane permeability. In the scope of the paper research, the ratio of the two values ranges from 1.3 to 1.7.