Experimental investigation of liquid water formation and transport in a transparent single-serpentine PEM fuel cell

Liquid water formation and transport were investigated by direct experimental visualization in an operational transparent single-serpentine PEM fuel cell. We examined the effectiveness of various gas diffusion layer (GDL) materials in removing water away from the cathode and through the flow field over a range of operating conditions. Complete polarization curves as well as time evolution studies after step changes in current draw were obtained with simultaneous liquid water visualization within the transparent cell. The level of cathode flow field flooding, under the same operating conditions and cell current, was recognized as a criterion for the water removal capacity of the GDL materials. When compared at the same current density (i.e. water production rate), higher amount of liquid water in the cathode channel indicated that water had been efficiently removed from the catalyst layer.

Visualization of the anode channel was used to investigate the influence of the microporous layer (MPL) on water transport. No liquid water was observed in the anode flow field unless cathode GDLs had an MPL. MPL on the cathode side creates a pressure barrier for water produced at the catalyst layer. Water is pushed across the membrane to the anode side, resulting in anode flow field flooding close to the H₂ exit.     

Numerical simulation of two-phase flow in gas diffusion layer and gas channel of proton exchange membrane fuel cells

Liquid water within the cathode Gas Diffusion Layer (GDL) and Gas Channel (GC) of Proton Exchange Membrane Fuel Cells (PEMFCs) is strongly coupled to gas transport properties, thereby affecting the electrochemical conversion rates. In this study, the GDL and GC regions are utilized as the simulation domain, which differs from previous studies that only focused on any one of them. A Volume of Fluid (VOF) method is adopted to numerically investigate the two-phase flow (gas and liquid) behavior, e.g., water transport pattern evolution, water coverage ratio as well as local and total water saturation. To obtain GDL geometries, an in-house geometry-based method is developed for GDL reconstruction. Furthermore, to study the effect of GDL carbon fiber diameter, the same procedure is used to reconstruct three GDL structures by varying the carbon fiber diameter but keeping the porosity and geometric dimensions constant. The wall wettability is introduced with static contact angles at carbon fiber surfaces and channel walls. The results show that the GDL fiber microstructure has a significant impact on the two-phase flow patterns in the cathode field. Different stages of two-phase flow pattern evolution in both cathode domains are observed. The liquid water in the GDL experiences water invasion, spreading, and rising, followed by the droplet breakthrough in the GDL/GC interface. In the GC, the water droplets randomly experience accumulation, combination, attachment, and detachment. Due to the difference in surface wettability, the water coverage of the GDL/GC interface is smaller than that of the channel side and top walls. It is also found that the water saturation inside the GDL stabilizes after the water breakthrough, while local water saturation at the interface keeps irregular oscillations. Last but not the least, a water saturation balance requirement between the GDL and GC is observed. In terms of varying fiber diameter, a larger fiber diameter would result in less water saturation in the GDL but more water in the GC, in addition to faster water movement throughout the total domain.     

Water distribution measurement for a PEMFC through neutron radiography

Neutron radiography has been used for in situ and non-destructive visualization and measurement technique for liquid water in a working proton exchange membrane fuel cell (PEMFC). In an attempt to differentiate water distribution in the anode side from that in the cathode side, a specially designed cell was machined and used for the experiment. The major difference between our design and traditional flow field design is the fact the anode channels and cathode channels were shifted by a channel width, so that the anode and cathode channels do not overlap in the majority of the active areas.

The neutron radiography experiments were performed at selected relative humidities, and stoichiometry values of cathode inlet. At each operating condition, the water distribution in anode/cathode gas diffusion layers (GDLs) was obtained. Image processing with four different spatial masks was applied to those images to differentiate liquid water in four different types of areas. Results indicate that the reactant gas relative humidity and stoichiometry significantly influence current density distribution and water distribution.     

Two-phase flow and maldistribution in gas channels of a polymer electrolyte fuel cell

Liquid water transport in a polymer electrolyte fuel cell (PEFC) is a major issue for automotive applications. Mist flow with tiny droplets suspended in gas has been commonly assumed for channel flow while two-phase flow has been modeled in other cell components. However, experimental studies have found that two-phase flow in the channels has a profound effect on PEFC performance, stability and durability. Therefore, a complete two-phase flow model is developed in this work for PEFC including two-phase flow in both anode and cathode channels. The model is validated against experimental data of the wetted area ratio and pressure drop in the cathode side. Due to the intrusion of soft gas diffusion layer (GDL) material in the channels, flow resistance is higher in some channels than in others. The resulting flow maldistribution among PEFC channels is of great concern because non-uniform distributions of fuel and oxidizer result in non-uniform reaction rates and thus adversely affect PEFC performance and durability. The two-phase flow maldistribution among the parallel channels in an operating PEFC is explored in detail.     

Water management studies in PEM fuel cells,Part I: Fuel cell design and in situ water distributions

A proton exchange membrane fuel cell (PEMFC) must maintain a balance between the hydration level required for efficient proton transfer and excess liquid water that can impede the flow of gases to the electrodes where the reactions take place. Therefore, it is critically important to understand the two-phase flow of liquid water combined with either the hydrogen (anode) or air (cathode) streams. In this paper, we describe the design of an in situ test apparatus that enables investigation of two-phase channel flow within PEMFCs, including the flow of water from the porous gas diffusion layer (GDL) into the channel gas flows; the flow of water within the bipolar plate channels themselves; and the dynamics of flow through multiple channels connected to common manifolds which maintain a uniform pressure differential across all possible flow paths. These two-phase flow effects have been studied at relatively low operating temperatures under steady-state conditions and during transient air purging sequences.     

Disclosure of the internal transport phenomena in an air-cooled proton exchange membrane fuel cell—Part I: Model development and base case study

In this study, the internal transport phenomena and mechanism inside an air-cooled proton exchange membrane fuel cell (PEMFC) are investigated. It helps to understand the factors that affect the performance of an air-cooled PEMFC and optimize the design of Membrane Electrode Assembly (MEA) and the flow field. This series article contains two parts. In this paper, i.e., Part Ⅰ of this series, a three-dimensional, two-phase flow, non-isothermal, steady-state Computational Fluid Dynamics (CFD) model is established to investigate the liquid water generation mechanism and the species distributions inside an air-cooled PEMFC single cell with a Base Case flow field design. Dry hydrogen and ambient air (the relative humidity and the stoichiometry are 60% and 150 separately) are considered for the simulation and validation. It is found that the liquid water appears mostly inside the cathode electrode underneath the cathode rib. Inside the anode gas diffusion layer (GDL), the mass fraction of H₂ underneath the cathode ribs is lower than that underneath the cathode channels, while the mass fraction of H₂O shows the opposite. The distributions of O₂ mass fraction and H₂O mass fraction inside the cathode GDL have the same trend as those of H₂ mass fraction and H₂O mass fraction inside the anode GDL. The membrane water content is periodically distributed from channel to channel and its value underneath the cathode rib is much larger than that underneath the cathode channel. The current density distribution is affected by the distribution of water content, i.e., the part underneath the cathode rib shows a larger current density than that underneath the cathode channel.     

Parameter sensitivity examination for a complete three-dimensional,two-phase,non-isothermal model of polymer electrolyte membrane fuel cell

Parameter sensitivity analysis is carried out for a complete three-dimensional, two-phase, non-isothermal model of polymer electrolyte membrane (PEM) fuel cell with a parallel flow field design. The model couples the two-phase flow of the multi-component reactants and liquid water, species transport, electrochemical reactions, proton and electron transport, and the electro-osmosis transport, back diffusion of water in the membrane, and energy transport. Twenty nine parameters, which are classified into the structural or transport parameters of porous layers (tortuosity, porosity, permeability, proton conductivity, electron conductivity, and thermal conductivity) as well as the electrochemical parameters (anodic and cathodic exchange current densities, anodic and cathodic transfer coefficients for anode and cathode reactions), are used to implement individual parameter investigation. The results show the parameters can be divided in to strongly sensitive, conditional sensitive and weak sensitive parameters according to its effect on the cell polarization curve. The optimization of parameters of cathode gas diffusion layer (GDL) and catalyst layer (CL) is more important to improve cell performance than that of anode GDL and CL because liquid water transport and removal affect significantly membrane hydration and reactant transport. Electrochemical parameters determine the activation potential and the slope of ohmic polarization hence these parameters can be used to fit experimental polarization curve more effectively than the other parameters.     

Two-phase flow in GDL and reactant channels of a proton exchange membrane fuel cell

Understanding the effect of two-phase flow in the components of proton exchange membrane fuel cells (PEMFCs) is crucial to water management and subsequently to their performance. The local water saturation in the gas diffusion layer (GDL) and reactant channels influences the hydration of the membrane which has a direct effect on the PEMFC performance. Mass transport resistance includes contributions from both the GDL and reactant channels, as well as the interface between the aforementioned components. Droplet–channel wall interaction, water area coverage ratio on the GDL, oxygen transport resistance at the GDL–channel interface, and two-phase pressure drop in the channels are interlinked. This study explores each factor individually and presents a comprehensive perspective on our current understanding of the two-phase transport characteristics in the PEMFC reactant channels.     

Numerical study of the effect of the GDL structure on water crossover in a direct methanol fuel cell

A two-dimensional two-phase non-isothermal mass transport model is developed to numerically investigate the behavior of water transport through the membrane electrode assembly (MEA) of a direct methanol fuel cell. The model enables the visualization of the distribution of the liquid saturation through the MEA and the analysis of the distinct effects of the three water transport mechanisms: diffusion, convection and electro-osmotic drag, on the water-crossover flux through the membrane. A parametric study is then performed to examine the effects of the structure design of the gas diffusion layer (GDL) on water crossover. The results indicate that the flow-channel rib coverage on the GDL surface and the deformation of the GDL can cause an uneven distribution of the water-crossover flux along the in-plane direction, especially at higher current densities. It is also found that both the contact angle and the permeability of the cathode GDL can significantly influence the water-crossover flux. The water-crossover flux can be reduced by improving the hydrophobicity of the cathode GDL.     

Impact of channel geometry on two-phase flow in fuel cell microchannels

An important function of the gas delivery channels in PEM fuel cells is the evacuation of water at the cathode. The resulting two-phase flow impedes reactant transport and causes parasitic losses. There is a need for research on two-phase flow in channels in which the phase fraction varies along the flow direction as in operating fuel cells. This work studies two-phase flow in 60 cm long channels with distributed water injection through a porous GDL wall to examine the physics of flows relevant to fuel cells. Flow regime maps based on local gas and liquid flow rates are constructed for experimental conditions corresponding to current densities between 0.5 and 2 A cm⁻² and stoichiometric coefficients from 1 to 4. Flow structures transition along the length of the channel. Stratified flow occurs at high liquid flow rates, while intermittent slug flow occurs at low liquid flow rates. The prevalence of stratified flow in these serpentine channels is discussed in relation to water removal mechanisms in the cathode channels of PEM fuel cells. Corners facilitate formation of liquid films in the channel, but may reduce the water-evacuation capability. This analysis informs design guidelines for gas delivery microchannels for fuel cells.     

Visualization of flooding in a single cell and stacks by using a newly-designed transparent PEMFC

The flow phenomenon in the fuel-cell channels is difficult to understand and predict because of the two-phase flow. Proton exchange membrane fuel cells (PEMFCs) with transparent windows are widely used for visualizing the two-phase flow in the channels. In this paper, the visualization of the two-phase flow in the channels was accomplished under various current-density conditions using a transparent cell. The visualization of the single serpentine flow field clearly reveals that anode flooding is more severe than cathode flooding. The main cause for anode flooding is a low gas-flow rate in the channel because of the absence of the carrier gas. In addition, flooding is more significant under a low current-density condition than under a high current-density condition; under the latter condition, there is significantly more reaction heat that prevents flooding. The flow phenomena in the PEMFC stack were also visualized by electrically connecting three transparent cells in series and supplying fuel to each cell from a manifold. Sudden voltage drops and overshoots were detected, and the voltage fluctuations were found to be strongly related to flooding.     

Water flooding and two-phase flow in cathode channels of proton exchange membrane fuel cells

The water flooding and two-phase flow of reactants and products in cathode flow channels (0.8 mm in width, 1.0 mm in depth) were studied by means of transparent proton exchange membrane fuel cells. Three transparent proton exchange membrane fuel cells with different flow fields including parallel flow field, interdigitated flow field and cascade flow field were used. The effects of flow field, cell temperature, cathode gas flow rate and operation time on water build-up and cell performance were studied, respectively. Experimental results indicate that the liquid water columns accumulating in the cathode flow channels can reduce the effective electrochemical reaction area; it makes mass transfer limitation resulting in the cell performance loss. The water in flow channels at high temperature is much less than that at low temperature. When the water flooding appears, increasing cathode flow rate can remove excess water and lead to good cell performance. The water and gas transfer can be enhanced and the water removal is easier in the interdigitated channels and cascade channels than in the parallel channels. The cell performances of the fuel cells that installed interdigitated flow field or cascade flow field are better than that installed with parallel flow field. The images of liquid water in the cathode channels at different operating time were recorded. The evolution of liquid water removing out of channels was also recorded by high-speed video.     

Validity of two-phase polymer electrolyte membrane fuel cell models with respect to the gas diffusion layer

A dynamic two-phase model of a proton exchange membrane fuel cell with respect to the gas diffusion layer (GDL) is presented and compared with chronoamperometric experiments. Very good agreement between experiment and simulation is achieved for potential step voltammetry (PSV) and sine wave testing (SWT). Homogenized two-phase models can be categorized in unsaturated flow theory (UFT) and multiphase mixture (M²) models. Both model approaches use the continuum hypothesis as fundamental assumption. Cyclic voltammetry experiments show that there is a deterministic and a stochastic liquid transport mode depending on the fraction of hydrophilic pores of the GDL. ESEM imaging is used to investigate the morphology of the liquid water accumulation in the pores of two different media (unteflonated Toray-TGP-H-090 and hydrophobic Freudenberg H2315 I3). The morphology of the liquid water accumulation are related with the cell behavior. The results show that UFT and M² two-phase models are a valid approach for diffusion media with large fraction of hydrophilic pores such as unteflonated Toray-TGP-H paper. However, the use of the homgenized UFT and M² models appears to be invalid for GDLs with large fraction of hydrophobic pores that corresponds to a high average contact angle of the GDL.     

Determination of effective water vapor diffusion coefficient in pemfc gas diffusion layers

The primary removal of product water in proton exchange membrane (PEM) fuel cells is through the cathode gas diffusion layer (GDL) which necessitates the understanding of vapor and liquid transport of water through porous media. In this investigation, the effect of microporous layer (MPL) coatings, GDL thickness, and polytetrafluorethylene (PTFE) loading on the effective water vapor diffusion coefficient is studied. MRC Grafil, SGL Sigracet, and Toray TGP-H GDL samples are tested experimentally with and without MPL coatings and varying PTFE loadings. A dynamic diffusion test cell is developed to produce a water vapor concentration gradient across the GDL and induce diffusion mass transfer. Tests are conducted at ambient temperature and flow rates of 500, 625, and 750 sccm. MPL coatings and increasing levels of PTFE content introduce significant resistance to diffusion while thickness has negligible effects.     

An approach for determining the liquid water distribution in a liquid-feed direct methanol fuel cell

In determining the liquid water distribution in the anode (or the cathode) diffusion medium of a liquid-feed direct methanol fuel cell (DMFC) with a conventional two-phase mass transport model, a current-independent liquid saturation boundary condition at the interface between the anode flow channel and diffusion layer (DL) (or at the interface between the cathode flow channel and cathode DL) needs to be assumed. The numerical results resulting from such a boundary condition cannot realistically reveal the liquid distribution in the porous region, as the liquid saturation at the interface between the flow channel and DL varies with current density. In this work, we propose a simple theoretical approach that is combined with the in situ measured water-crossover flux in the DMFC to determine the liquid saturation in the anode catalyst layer (CL) and in the cathode CL. The determined liquid saturation in the anode CL (or in the cathode CL) can then be used as a known boundary condition to determine the water distribution in the anode DL (or in the cathode DL) with a two-phase mass transport model. The numerical results show that the water distribution becomes much more realistic than those predicted with the assumed boundary condition at the interface between the flow channel and DL.     

Characteristics of water transport through the membrane in direct methanol fuel cells operating with neat methanol

The water required for the methanol oxidation reaction in a direct methanol fuel cell (DMFC) operating with neat methanol can be supplied by diffusion from the cathode to the anode through the membrane. In this work, we present a method that allows the water transport rate through the membrane to be in-situ determined. With this method, the effects of the design parameters of the membrane electrode assembly (MEA) and operating conditions on the water transport through the membrane are investigated. The experimental data show that the water flux by diffusion from the cathode to the anode is higher than the opposite flow flux of water due to electro-osmotic drag (EOD) at a given current density, resulting in a net water transport from the cathode to the anode. The results also show that thinning the anode gas diffusion layer (GDL) and the membrane as well as thickening the cathode GDL can enhance the water transport flux from the cathode to the anode. However, a too thin anode GDL or a too thick cathode GDL will lower the cell performance due to the increases in the water concentration loss at the anode catalyst layer (CL) and the oxygen concentration loss at the cathode CL, respectively.     

Effects of hydrophilic/hydrophobic properties of gas flow channels on liquid water transport in a serpentine polymer electrolyte membrane fuel cell

The droplet dynamics in the serpentine flow channel of a hydrogen fuel cell has been numerically investigated to obtain ideas for designing a serpentine channel with the aim of effectively preventing flooding. Three-dimensional two-phase flow simulations employing the volume of fluid (VOF) method have been performed. Liquid droplets emerging from four adjacent pores at the hydrophobic bottom wall are subjected to airflow in the bulk of the serpentine flow channel. The effects of contact angle variation of the channel walls on liquid water removal have been tested in terms of liquid water saturation and coverage of liquid water on the gas diffusion layer (GDL) surface. The numerical results show that the hybrid case, which consists of hydrophilic channel walls at the straight part and hydrophobic walls at the turning part of the serpentine flow channels, enhances water removal compared with two other cases in which the channel wall is homogeneously hydrophilic or hydrophobic. The three-dimensional visualization of liquid water droplets reveals that the hydrophobic wall at the turning part reduces the water saturation in the channel and the hydrophilic wall at the straight part prevents the liquid water from covering the GDL surface.     

A three-dimensional two-phase flow model for a liquid-fed direct methanol fuel cell

A three-dimensional, two-phase, multi-component model has been developed for a liquid-fed DMFC. The modeling domain consists of the membrane, two catalyst layers, two diffusion layers, and two channels. Both liquid and gas phases are considered in the entire anode, including the channel, the diffusion layer and the catalyst layer; while at the cathode, two phases are considered in the gas diffusion layer and the catalyst layer but only single gas phase is considered in the channels. For electrochemical kinetics, the Tafel equation incorporating the effects of two phases is used at both the cathode and anode sides. At the anode side the presence of gas phase reduces the active catalyst areas, while at the cathode side the presence of liquid water reduces the active catalyst areas. The mixed potential effects due to methanol crossover are also included in the model. The results from the two-phase flow mode fit the experimental results better than those from the single-phase model. The modeling results show that the single-phase models over-predict methanol crossover. The modeling results also show that the porosity of the anode diffusion layer plays an important role in the DMFC performance. With low diffusion layer porosity, the produced carbon dioxide cannot be removed effectively from the catalyst layer, thus reducing the active catalyst area as well as blocking methanol from reaching the reaction zone. A similar effect exits in the cathode for the liquid water.     

Reactants flow behavior and water management for different current densities in PEMFC

Computational fluid dynamics analysis was carried out to investigate the reactants flow behavior and water management for proton exchange membrane fuel cell (PEMFC). A complete three-dimensional model was chosen for single straight channel geometry considering both anode and cathode humidification. Phase transformation was included in the model to predict the water vapor and liquid water distributions and the overall performance of the cell for different current densities. The simulated results showed that for fully humidified conditions hydrogen mole fraction increases along the anode channel with increasing current density, however, at higher current densities it decreases monotonically. Different anode and cathode humidified conditions results showed that the cell performance was sufficiently influenced by anode humidification. The reactants and water distribution and membrane conductivity in the cell depended on anode humidification and the related water management. The cathode channel–GDL (Gas Diffusion Layer) interface experiences higher temperature and reduces the liquid water formation at the cathode channel. Indeed, at higher current densities the water accumulated in the shoulder area and exposed higher local current density than the channel area. Higher anode with lower cathode humidified combination showed that the cell had best performance based on water and thermal management and caused higher velocity in the cathode channel. The model was validated through the available literature.     

Liquid water transport between graphite paper and a solid surface

We studied the interaction of a water droplet with a solid wall on a hydrophobic gas diffusion layer (GDL). Of particular interest is the stability of the droplet as a function of plate wetting properties and the potential for liquid entrapment in the GDL/land contact area. Such transport is of relevance to breakthrough dynamics and convective liquid droplet transport in polymer electrolyte membrane (PEM) fuel cell cathode gas channels. While a variety of complex coupled transport phenomena are present in the PEM fuel cell gas channel, we utilize a very simplified experimental model of the system where a droplet originally placed on a hydrophobic GDL is translated quasistatically across the GDL surface by a solid surface. Transport and entrapment are imaged using fluorescence microscopy. This work provides new insights into droplet behaviour at the GDL/land interface in a PEM fuel cell and suggests that hydrophobic land areas are preferable for mitigating the accumulation of liquid water under the land area of the gas flow channels.