GDL Surface Research Articles

A semi-analytical model for droplet dynamics on the GDL surface of a PEFC electrode

Water management is one of the key factors in Proton Exchange Fuel Cell (PEFC) performance. The water produced within the fuel cell is evacuated through the gas channels, but at high current densities water can block the channel, thus limiting the current density generated in the fuel cell. A semi-analytical model of a water droplet emerging from a gas diffusion layer pore in a PEFC channel is developed. The transient model contains a detailed adhesion and drag force estimation model. Results show that the predicted values for both drag and surface tension force are higher than the results found in literature. The results for the detachment force are consistent with the experimental data available. Higher air velocity values lead to more deformation of the droplet and oscillation with lower frequency but higher amplitude. Similar effects have been identified when the liquid mass flow is increased, leading to faster detachment of the droplet.

Effect of Chloride on PEFCs in Presence of Various Cations

The performance and durability of Polymer electrolyte fuel cells (PEFCs) were investigated by separately introducing HCl and five different chloride salts (AlCl3, FeCl3, CrCl3, NiCl2, and MgCl2) into the air stream of an operating fuel cell. Under the same operating conditions and at a fixed chloride (Cl−) concentration of 50 ppm in air stream, it appeared that HCl affected the cell performance faster than the other chloride salts, therefore the presence of the cation in the contaminant solutions slowed and/or reduced the decrease in the cell performance. The overall cell performance degradation was ranked as HCl, AlCl3 > FeCl3 > CrCl3 > NiCl2, MgCl2. The performance degradation is believed to be due to the adsorption of chloride on the Pt surface reducing active surface area. At lower relative humidity, the contaminant concentration increased in liquid phase and salt precipitated eventually blocking some of the flow channels and the GDL surface and resulted in even lower cell performance.

Development of Numerical Simulation and Visualization System for the Designing of Polymer Electrolyte Fuel Cell Stack

Overview Polymer Electrolyte Fuel Cell (PEFC) is expected as a new power source in automobiles with high power density and low emission. Although PEFC has many advantages compared with internal combustion engines, the cost reduction and the improvement of performance and durability are needed for its practical application. For improvement of PEFC, it is important to understand the mechanism inside it under various operating and structural conditions and to feedback into the design of membrane electrode assembly (MEA) and bipolar plate constituting fuel cell stack.In order to visualize inside PEFC, we have been developed a numerical simulation tool calculating the cell and stack performance and physical distributions and a nondestructive, real-time/space visualization technique for oxygen partial pressure in the cathode flow channel and the GDL using oxygen-sensitive luminescent dye film [1].In this study, as for the first step of our numerical modeling development, the effects of operating condition, for example, cathode flow rate on the test cell as well as commercial-sized cell and stack were investigated. Numerical Modeling Inside the fuel cell, multi-physical and multi-scale of phenomena such as multi-component gas, liquid water and heat transport inside the flow channel and the GDL, electrochemical reactions and proton/water transport in the MEA are nonlinearly coupled [2]. These phenomena require too much calculation cost to be solved in high resolution by fully 3-dimensional equations based on computed fluid dynamics (CFD), so it is necessary to apply simplified models. In our modeling, macroscopic models are applied for microscopic phenomena in the catalyst layer and electrolyte such as electrochemical reactions (HOR/ORR) and proton/water transport. And these models are coupled with multi-dimensional heat, gas and liquid water transport equations in which effective parameters such as gas diffusivity and permeability in the GDL and equivalent hydraulic diameter in the flow field of bipolar plate and the manifold of stack are included [3]. Results First of all, the performance of 25cm2test cell with single serpentine flow channel was carried out. Hydrogen and air flow rate is 300ml/min, relative humidity is 60%RH, cell temperature is 353.15K and back pressure is 1atm, respectively.The distribution of oxygen partial pressure in cathode flow channel and on the GDL surface at cathode utilization 80% (0.58 A/cm2) are shown in Figure 1 compared to the measured data by a nondestructive visualization system. The results are good agreement with the measured data and indicate that the numerical prediction of current density distribution is valid in our model.Applying the validated model, the effect of cathode flow rate on the performance of short stack with 20 cells (max. 5kW) was carried out. As for the other operating conditions, relative humidity is 50%RH, cell temperature is 343.15K and back pressure is 1.5atm, respectively. Figure 2 shows the effect of cathode utilization on the distribution of normalized in-flow rate and ORR gain for each cell at 0.5A/cm2. As cathode utilization increases, in-flow rate decreases and ORR gain increases remarkably near the dead end of stack. This is because liquid water accumulates in the flow channel and the GDL.In addition, the performance of full-scale PEFC stack for automobiles under various operating conditions will also be discussed in this meeting.

Cross-Sectional Visualization of Liquid Water Behavior and Evaluation of Wettability Effect in Cathode Channel of an Operating PEFC

During operation of polymer electrolyte fuel cells (PEFCs), excessive liquid water clogs cathode channel and causes oxygen transport limitation. This phenomenon known as “water plugging” is recognized as a critical barrier to high performance operation. To alleviate this problem, it is necessary to deeply understand liquid water transport within cathode channel and design the optimum channel structure. It is also effective to control the surface wettability of cathode channel for promoting water removal. In the previous works, many experimental efforts to visualize liquid water behavior in PEFCs have been conducted by neutron radiography and optical technique using transparent cells. However, most of the conventional visualization studies investigated only the water transport in the in-plane direction. The through-plane behaviors of liquid droplets inside a cathode channel, including the emergence, growth and combination of droplets on the GDL surface and the attachment to the channel sidewall were not grasped in detail. To directly observe the detailed behaviors of water droplets in cathode channel, "cross-sectional visualization cell" was recently developed by several researchers. This visualization technique is very useful to evaluate the growth and attachment behaviors of liquid droplets and the blow-off of water plugs in a gas channel. However, in the past experiments using cross-sectional cell, liquid water was injected into a simplified channel model simulating actual PEFCs by syringe pump, and ex-situ observation of liquid water behavior was only carried out under non-operating condition. The cross-sectional imaging in an operating PEFC was not attempted, and the correlation between water transport in cathode channel and cell performance was not yet discussed. In this study, we fabricate a cross-sectional visualization cell, which enables the internal observation of each gas channel, and attempt to visualize the dynamic behaviors of liquid droplets in the cathode channel of PEFC under operating conditions. In the experiment, the attachment behavior of liquid droplets to the channel sidewall and the removal of water plugs in the cathode channel are investigated, and the effect of wettability of channel wall on the liquid water transport is also evaluated. Furthermore, the cell voltage and differential pressure are simultaneously measured, and the correlation between the water removal and the cell performance is discussed.Fig. 1 presents the schematic diagram of the cross-sectional visualization cell developed in this study. The membrane electrode assembly (MEA) is sandwiched between two stainless steel current collectors (11 mm X 38 mm) with a two-pass serpentine flow channel. The wide, depth, and length of flow channel are 1.0 mm, 1.0 mm, and 88.5 mm, respectively. Two separators placed outside the current collectors are held together by four bolts. To observe the water droplet behavior inside the cathode channel, quartz windows were attached to the sidewalls of each current collector. In addition, to evaluate the effect of channel surface wettability, the cathode flow channel was treated with hydrophilic coating. The contact angles of uncoated and hydrophilized channels are 73 and 18o, respectively.After starting the operation, liquid droplets emerging from the GDL first adhere to the channel sidewall. When the channel surface of cathode is uncoated, water droplets attached to the channel sidewall gradually grow up and reach the upper wall, resulting in channel clogging. On the other hand, when liquid droplets are in contact with the hydrophilic channel sidewall, thin liquid films are immediately formed on the channel surface. And then, liquid films reach the corner between the upper and side walls and spread along the flow direction. Therefore, it is not easy to form water plugs in the hydrophilic cathode channel.

Distributed PEFC Performance during Cationic Contamination

In our previous studies (1-3), severe performance degradation of PEFCs was observed due to cationic impurities. In this study, we investigate how cations affect the spatial performance distribution in a PEFC using a segmented cell.The segmented cell hardware consists of an aluminum anode flow field, a segmented aluminum cathode flow field, two aluminum end plates and one cupper current collector. All the aluminum parts were gold plated. Cathode flow field was divided into eight segments. Eight pieces of equal size aluminum current collectors were placed inside a PTFE block and straight flow channels were machined through both the PTFE and the aluminum segments (Figure 1). Total active area of the cell was 5 cm2. 2.85 mM CaSO4solution was introduced in the air stream through a nebulizer (1-3) at a rate of 110 uL/min, corresponding to 21.2 ppm based on dry air flow rate.Figure 2 shows the voltage and current density distribution profile in each segment. The cell was operated at a constant total current of 1 A. During the baseline test of 24 hours, pure air was injected through cathode. The current in each segment remained constant through this period. Following the baseline test, the contaminant was injected in the air stream. The segments closest to the gas inlet were first affected by the contaminant. The current density in the segments closest to the gas inlet decreased while the segments closest to the outlet increased. Since it was a constant current test, when the current in the segments closest to the gas inlet decrease, other segments picked up that current to maintain total current constant.During the baseline test, voltage remained constant in all segments. After contaminant injection, the cell voltage started to decrease and after 50 hours, voltages in each segments fluctuated. The voltage fluctuation was found to be mainly due to mass transport loss. The contaminant solution evaporated and deposited on the flow field and the GDL surface (Figure 3), and entered into the GDL which caused the mass transport losses. Moreover, the aluminum flow fields corroded and corrosion byproducts also deposited on the flow field and GDL surface. Throughout the test, there was no significant variation in voltage between segments. Acknowledgements The authors gratefully acknowledge financial support from NSF (CBET-0748063), DOE-EERE through University of Hawaii-Hawaii Natural Energy Institute (DE-EE0000467).

Numerical study of two-phase flow patterns in the gas channel of PEM fuel cells with tapered flow field design

In this study the air–water two-phase flow in a tapered channel of a PEMFC was numerically simulated using the volume of fluid (VOF) method. In particular, a 3D mathematical model of the fuel cell flow channel was used to obtain a reliable evaluation of the fuel cell performance for different taper angles and different temperatures and to calculate the total amount of water produced. This information was then used as boundary conditions to simulate the two-phase flow in the cell channel through a 2D VOF model. Typical operating conditions were assigned and the numerical mesh was constructed to represent the real fuel cell configuration. The results show that tapering the channel downstream enhances the water removal due to increased airflow velocity. In the rectangular channel no film formation is noted with a marked predominance of slug flow. In contrast, as the taper angle is increased the predominant two-phase flow pattern is film flow. Finally many contact angles have been used to simulate the effect of the hydrophobicity of a GDL surface on the motion of the water. As the hydrophobicity of a GDL surface is decreased the presence of film is more evident even for less tapered channels.

Effect of channel materials and trapezoidal corner angles on emerging droplet behavior in Proton Exchange Membrane Fuel Cell gas channels

Ex situ experiments were conducted to study water droplet dynamics in Proton Exchange Membrane Fuel Cell (PEMFC) gas channels with different trapezoid channel open angles and GDL materials under controlled air velocities (0.48–7.23 m s−1). High speed videos revealed that the droplets interacted with the channel walls at lower air velocities (0.48–4 m s−1) and led to channel corner filling depending on the channel open angle. However, at higher air velocities (4.1–7.23 m s−1), the droplets did not contact the channel walls and slid off from the GDL surface. For higher air velocities, a correlation was obtained for the minimum velocity required to remove the droplet from the channel without interacting with the channel walls. For lower air velocities, the minimum pressure drop required for removing the droplet from the gas channel for both corner filling and non-filling scenarios was obtained. The study clearly establishes the important role played by the channel sidewalls and the channel angle on the water droplet transport characteristics.

The Performance Degradation of PEMFCs Fabricated with Different GDLs During Exposure to Simultaneous Sulfur Impurity Poisoning Condition

This paper reveals the performance decrease and recovery of PEMFC when the contaminated fuel gas and air source with sulfur impurities such as hydrogen sulfide and sulfur dioxide were simultaneously introduced to anode and cathode, respectively. Three different GDLs were fabricated with different carbon black and activated carbon to prevent an introduction of sulfur compound impurities into MEA. components. The severity of SO2 and H2S poisoning was depended on concentrations(3 ppm 10 ppm) of sulfur impurities. Especially, cell performance degradation rate was rapid when MEA fabricated with CN-2 GDL because it had little porosity on GDL surface. Moreover, the cell performance can be recovered up to 90%-95% only with neat hydrogen and fresh air feeding.. Conclusively, MEA fabricated with porous CN-1 GDL showed the best cell performance and recovery efficiency during exposure to poisoning condition by simultaneous sulfur impurities.

J056024 Mass Conversation Improvement in LBM and Its Application in PEFC Simulation

The emergence and transportation process of liquid water from a micro-pore on the gas diffusion layer surface under the action of gas flow in a polymer electrolyte membrane fuel cell (PEFC) is simulated by a modified free energy based lattice Boltzmann method (LBM) with large density ratio. A so called space averaged index function method is used to improve the mass conservation in the LBM scheme. And the effect of the GDL surface wettability on the liquid water removal efficiency is investigated. The simulation results show that the mass conservation of the LBM for the 3D simulation has been greatly improved and satisfied. The highly hydrophobic surface of GDL has a better performance in liquid water removal and less surface coverage area. It is shown that this LBM method can be an effective tool in simulating the two phase flow in the PEFC.

A Cross-Sectional Observation of Water Behavior in the Flow Channel in PEFC

Liquid water growth pattern in the gas flow channel in PEFC was visualized by using an ex-situ optical cross-sectional observation system developed. We investigated the impact of surface hydrophobicity of the GDL and channel on the liquid water growth pattern. The liquid water, which was emerged at 5 mm back from the gas outlet, was observed with spatial resolution of 20 μm by optimizing the optical system including CCD camera. Pressure drop through the channel was measured to evaluate the impact of water droplet on it. As a result, it is obtained that adhesion force formed in combination of hydrophobic GDL and hydrophilic channel sucked the water from GDL to channel corner, and suppressed the water contact with the GDL surface, and accelerated the water drainage, leading to prevent the water droplet growing in the channel.

Analysis of liquid water discharge performance in a channel of PEM fuel cells using lattice Boltzmann method

Polymer electrolyte membrane fuel cells (PEMFCs) have been given much attention in the last decade. Understanding the liquid water behavior in PEMFC is of considerable practical significance. We applied numerical lattice Boltzmann method (LBM) to simulate the liquid water droplet dynamics, its subsequent movement on the GDL surface under the action of gas. The effect of the inlet air flow velocity, droplet characteristics (size, initial position), gas channel height on the pump work, needed to remove the liquid water droplet, are investigated. The optimum design will be investigated for best performance.

D203 多孔体セパレータを有する固体高分子形燃料電池の性能特性に関する研究(OS4 水素・燃料電池),動力エネルギーシステム部門20周年,次の20年への新展開)

This study conducted perfomance evaluation of single cell with porous cathode separataor. The direct observation of the porous separator cross-section and the impedance spectra were applied. The observation showed that the porous separator is superior in the drainage performance from the GDL surface, and the hydrophily of separator fiber is important for the performance. The results of impedance spectra indicated that the porous separator is superior to the serpentine separator in the supply of reactant gas.

Drops, slugs, and flooding in polymer electrolyte membrane fuel cells

AbstractThe process of flooding has been examined with a single‐channel fuel cell that permits direct observation of liquid water motion and local current density. As product water flows through the largest pores in the hydrophobic GDL, drops detach from the surface, aggregate, and form slugs. Flooding in polymer electrolyte membrane (PEM) fuel cells occurs when liquid water slugs accumulate in the gas flow channel, inhibiting reactant transport. Because of the importance of gravity, we observe different characteristics with different orientations of the flow channels. Liquid water may fall away from the GDL and be pushed out with minimal effect on the local current density, accumulate on the GDL surface and cause local fluctuations, or become a pulsating flow of liquid slugs and cause periodic oscillations. We show that flooding in PEM fuel cells is gravity‐dependent and the local current densities depend on dynamics of liquid slugs moving through the flow channels. © 2008 American Institute of Chemical Engineers AIChE J, 2008

Predicting Water-Droplet Detachment from GDL/Channel Interfaces in PEM Fuel Cells

An analytical model is presented for predicting the critical air-flow velocity at the onset of water-droplet detachment from the GDL/channel interfaces in PEM fuel cells. Our model is based on the force balance between pressure drag that tends to detach the droplet and surface tension that tends to hold the droplet in place. In the present work, we consider the flow regime in which pressure drag, which arises from inertia effects, dominates over viscous shear - this is the flow regime of interest in real-world PEM fuel cell applications, both automotive and stationary. Our analytical model predicts that the critical air-flow velocity varies inversely (to the 2/3 power) with water-droplet size. It further predicts that making the GDL surface more hydrophobic, decreasing contact-angle hysteresis, and shrinking channel height reduce the critical air-flow velocity. Model predictions are compared with experimental data available from the literature and reasonably good agreements are obtained.

Nanostructured Gas Diffusion and Catalyst Layers for Proton Exchange Membrane Fuel Cells

Nanostructured components are introduced in membrane electrodes assembly MEA in proton exchange membrane fuel cell as asolution to improve the performance. Single-walled carbon nanotubes and multiwalled carbon nanotubes supported platinum areused to fabricate the gas diffusion layer GDL and the catalyst layers in the MEAs, respectively. The physicochemical andelectrochemical characterizations of these nanotube-based components demonstrate excellent GDL surface morphology and uni-form distribution of the platinum catalyst over the carbon nanotube support. The fuel cell testing using these nanostructuredcomponents exhibits promising fuel cell performance using hydrogen-air and hydrogen-oxygen at ambient pressure.© 2006 The Electrochemical Society. DOI: 10.1149/1.2422751 All rights reserved.Manuscript submitted September 19, 2006; revised manuscript received October 16, 2006.Available electronically December 26, 2006.

A model for a vertical planar air breathing PEM fuel cell

This paper presents a two-dimensional, steady state, single phase, non-isothermal and complete model for a vertical, fully planar, Air Breathing Proton Exchange Membrane Fuel Cell (AB-PEMFC) with hydrogen fuel supplied by forced convection at the anode. It is applied to a cell with an active area of 6 cm 2 operating at ambient atmospheric conditions of 23 °C and 20% relative humidity. The transport characteristics in terms of the velocity, and heat and mass transfer coefficients in the various components of the fuel cell are reported for several values of the operating current density. The effect of cathode height (1–5 cm) and operating atmospheric conditions (10–40 °C and 20–80% relative humidity), on the cell performance is also reported. Further, the applicability of the model to a non-planar AB-PEMFC is examined by comparison with available experimental data. The average mass transfer coefficients for oxygen and water vapor at the cathode GDL surface are found to be of the order of 10 −3 m s −1. The operating current density is seen to substantially affect the variation of the local current density distribution and the cathode surface temperature along the height of the fuel cell as also the temperature variation across the MEA thickness. The maximum power density and the corresponding current density, herein defined as the optimum current density, are found to increase with decreasing height of the fuel cell, decreasing ambient temperature and increasing ambient relative humidity. However, the local cell temperature at high current densities is found to increase beyond the safe operating limits for short fuel cells. Comparison of the model predictions with available experimental data points to its applicability in the ohmic polarization zone of a non-planar cell. The cell performance at high current densities deteriorates due to mass transport limitation and electrolyte dehydration.