Experimental study on mitigating the cathode flooding at low temperature by adding hydrogen to the cathode reactant gas in PEM fuel cell

Severe flooding can be critical in a fuel cell vehicle operating at a high current density, and in a fuel cell vehicle at the initial stage of start up. It is often difficult to remove the condensed water from the cathode gas diffusion layer (GDL) of the fuel cell because of the surface tension between the water and the GDL. In this research, in order to remove the condensed water from the cathode GDL, a small amount of hydrogen was injected into the cathode reactant gases. The results showed that the hydrogen addition method successfully removed the liquid water from the cathode GDL. Water removal was verified for various hydrogen flow rates and hydrogen addition durations. Furthermore, the dew point temperature of the outlet gas at the cathode was observed to determine the amount of water removed from the cathode GDL. In addition, the water droplet in the cathode gas flow channel was visualized by using a transparent cell. Furthermore, degradation tests are also performed. Considering the degradation test, the hydrogen addition method is expected to be effective in mitigating cathode flooding.     

Effects of cathode catalyst layer design parameters on cold start behavior of polymer electrolyte fuel cells (PEFCs)

We present a theoretical study on the effects of key catalyst layer (CL) design parameters on the cold start behavior of a polymer electrolyte fuel cell (PEFC) using a three-dimensional transient cold start model developed in a previous study 1 and 2. Among several CL design parameters, we adopt the ionomer fraction (?_I) and weight ratio of Pt to carbon support (wt%_Pt–C) in the cathode CL as CL design variables for this study. Therefore, other design parameters such as CL thickness and the oxygen reduction reaction (ORR) kinetic parameter are accordingly adjusted due to changes in ?_I and wt%_Pt–C for cold start simulations. The calculated results confirm that these two design parameters provide control of the ice storage capacity and water absorption potential of the cathode CL, and consequently have a substantial influence on the cold start behavior of a PEFC. We provide a guideline to design and optimize a cathode CL and membrane electrode assembly (MEA) for improved PEFC cold start capability.     

PEM fuel cell electrodes

The design of electrodes for polymer electrolyte membrane fuel cells (PEMFC) is a delicate balancing of transport media. Conductance of gas, electrons, and protons must be optimized to provide efficient transport to and from the electrochemical reactions. This is accomplished through careful consideration of the volume of conducting media required by each phase and the distribution of the respective conducting network. In addition, the issue of electrode flooding cannot be neglected in the electrode design process. This review is a survey of recent literature with the objective to identify common components, designs and assembly methods for PEMFC electrodes. We provide an overview of fabrication methods that have been shown to produce effective electrodes and those that we have deemed to have high future potential. The relative performances of the electrodes are characterized to facilitate comparison between design methodologies.     

Droplet pinning by PEM fuel cell GDL surfaces

In this work, side view images of liquid–gas–solid interfaces are observed during the evaporation of liquid water droplets on various commercially available untreated gas diffusion layers (GDLs). The change in contact diameter as a function of evaporative volume loss is measured to quantify the unpinning rates of micro-sized droplets. This contact diameter pinning behaviour during evaporation is correlated to the material topography, which is quantified through profilometry measurements. The carbon fibre paper with the smallest average roughness (15 μm) exhibits the strongest degree of pinning (unpinning at a rate of 0.13 mm/μL). Higher average surface roughnesses for felt (30 μm) and cloth yarn (32 μm) result in higher unpinning rates, 0.21 mm/μL and 0.19 mm/μL, respectively. These results indicate that common GDL materials exhibit Cassie–Baxter wetting behaviour, and reduced GDL roughness promotes droplet pinning. The material-specific droplet contact diameter progression should be considered during GDL selection for polymer electrolyte membrane (PEM) fuel cells. This work provides insight into the effect of GDL material properties on gas channel water management, as water droplets are expected to experience similar pinning to that observed in this work within the cathode gas channels of a PEM fuel cell.     

Two-phase modeling of gas purge in a polymer electrolyte fuel cell

Gas purge intended to minimize residual water in a polymer electrolyte fuel cell (PEFC) is critical for successful shutdown and sub-zero startup. In the present work, we present a two-phase transient model describing water removal from PEFC under gas purge conditions. The role of back diffusion from the cathode to anode along with liquid water transport in the gas diffusion layers behind the drying front and vapor diffusion ahead of the drying front is highlighted. The underlying ineffectiveness of cathode-only purge is outlined. The model predictions are compared with experimental results under various purge conditions. A good match with experiments is obtained at higher purge temperatures whereas some differences in the HFR profile is observed at lower temperatures. The role of drying front morphology in addressing the observed differences between numerical and experimental results is hypothesized.     

The effects of the cathode flooding on the transient responses of a PEM fuel cell

This study investigates the effects of the flooding of the gas diffusion layer (GDL), as a result of liquid water accumulation, on the performance of a proton exchange membrane fuel cell (PEMFC). The transient profiles of the current generated by the cell are obtained using the numerical resolution of the transport equation for the oxygen molar concentration in the unsteady state. The dynamics of the system are captured through the reduction of the effective porosity of the GDL by the liquid water which accumulates in the void space of the GDL. The effects of the GDL porosity, GDL thickness and mass transfer at the GDL–gas channel interface on the evolution with time of the averaged current density are reported. The effects of the current collector rib on the evolution of the molar concentration of oxygen are also examined in detail.     

Studies on the cathode humidification by exhaust gas recirculation for PEM fuel cell

For high efficiency and long durability of proton exchange membrane fuel cells (PEMFCs), polymer electrolyte membranes should be kept wet. Reactant gases should be humidified on this account. For the humidification, the PEMFC system uses an external or internal humidifier as a part of balance of plants (BOPs). However, external humidifiers have many disadvantages such as parasitic power loss, system complexity, high cost and bulky volume. As such, efforts have been made to remove the external humidifier or replace it with an advanced humidifier. In this work, to remove a humidifier, humidification by exhaust gas recirculation is investigated by theoretical analysis and experiments with 5-cell stack of an active area 250 cm². In the theoretical analysis, species conservation equations and energy conservation equation are solved to obtain the O₂ concentration, stoichiometric ratio, humidity ratio, temperature, amount of condensed water and so on. With the theoretical results, experiments with 5-cell, 250 cm² fuel cell stack were carried out in order to analyze the stack performance at the theoretical conditions of the cathode process stream of exhaust gas recirculation.     

Hydrogen purge and reactant feeding strategies in self-humidified PEM fuel cell systems

A 6 kW proton exchange membrane fuel cell system, operating in self-humidified conditions, was characterized in two anode operative modes: dead-end and flow through with exhaust recirculation. The anode sub-system was specifically designed in order to adjust the level of recycled anodic stream. The role of anode purge frequency, anode recirculation level and air stoichiometric ratio was analysed in the power range 1–5 kW. The aim of this study was to define management strategies to assure efficient and reliable cell performance during steady-state, warm-up and load variation phases. The results evidenced the combined effect of hydrogen purge, air flow rate impulse, and recycled anodic stream on individual cell performance recovery when unstable working conditions were detected during system start-up and load variations.     

Effects of various operating and initial conditions on cold start performance of polymer electrolyte membrane fuel cells

Cold start is a challenging and important issue that hinders the commercialization of polymer electrolyte membrane fuel cell (PEMFC). In this study, a three-dimensional multiphase model has been developed to simulate the cold start processes in a PEMFC. Numerical simulations have been conducted for a single PEMFC starting at various operating and initial conditions, which are cell voltages, initial water contents and distributions, anode inlet relative humidity (RH), surrounding heat transfer coefficients, and cell temperatures. It is found that the heating-up time can be significantly reduced by decreasing the cell voltage and effective purge is critical for PEMFC cold start. The largest heating source at high cell voltages is the activational heat, and it becomes the ohmic heat at low cell voltages. The water freezing in the membrane is not observed when the cell is producing current due to the heat generation and the slow water diffusion into the membrane at subzero temperatures, and it is only observed after the cold start is failed, further confirming the importance of purge. Humidification of the supplied hydrogen has negligible effect on the cold start performance since only small amounts of water vapour can be taken by the gas streams at subzero temperatures. The surrounding heat transfer coefficients have significant influence on the heating-up time, indicating the importance of cell insulation or heating. The rate of cell heating up is reduced when the startup temperature is lowered due to the more sluggish electrochemical reaction kinetics.     

Characteristics of proton exchange membrane fuel cells cold start with silica in cathode catalyst layers

In this study, a novel strategy is reported to improve the cold start performance of proton exchange membrane (PEM) fuel cells at subzero temperatures. Hydrophilic nano-oxide such as SiO₂ is added into the catalyst layer (CL) of the cathode to increase its water storing capacity. To investigate the effect of nanosized SiO₂ addition, the catalyst coated membranes (CCMs) with 5 wt.% and without nanosized SiO₂ are fabricated. Although at normal operation conditions the cell performance with nanosized SiO₂ was not so good as that without SiO₂, cold start experiments at −8 °C showed that the former could start and run even at 100 mA cm⁻² for about 25 min and latter failed very shortly. Even at −10 °C, the addition of SiO₂ dramatically increased the running time before the cell voltage dropped to zero. These results further experimentally proved the cold start process was strongly related with the cathode water storage capacity. Also, the performance degradation during 8 cold start cycles was evaluated through polarization curves, cyclic voltammetry (CV) and electrochemical impedance spetra (EIS). Compared with the cell without SiO₂ addition, the cell with 5 wt.% SiO₂ indicated no obvious degradation on cell performance, electrochemical active surface area and charge transfer resistance after experiencing cold start cycles at −8 °C.     

Effects of porosity distribution variation on the liquid water flux through gas diffusion layers of PEM fuel cells

Flooding of the membrane electrode assembly (MEA) and dehydrating of the polymer electrolyte membrane have been the key problems to be solved for polymer electrolyte membrane fuel cells (PEMFCs). So far, almost no papers published have focused on studies of the liquid water flux through differently structured gas diffusion layers (GDLs). For gas diffusion layers including structures of uniform porosity, changes in porosity (GDL with microporous layer (MPL)) and gradient change porosity, using a one-dimensional model, the liquid saturation distribution is analyzed based on the assumption of a fixed liquid water flux through the GDL. And then the liquid water flux through the GDL is calculated based on the assumption of a fixed liquid saturation difference between the interfaces of the catalyst layer/GDL and the GDL/gas channel. Our results show that under steady-state conditions, the liquid water flux through the GDL increases as contact angle and porosity increase and as the GDL thickness decreases. When a MPL is placed between the catalyst layer and the GDL, the liquid saturation is redistributed across the MPL and GDL. This improves the liquid water draining performance. The liquid water flux through the GDL increases as the MPL porosity increases and the MPL thickness decreases. When the total thickness of the GDL and MPL is kept constant and when the MPL is thinned to 3 μm, the liquid water flux increases considerably, i.e. flooding of MEA is difficult. A GDL with a gradient of porosity is more favorable for liquid water discharge from catalyst layer into the gas channel; for the GDLs with the same equivalent porosity, the larger the gradient is, the more easily the liquid water is discharged. Of the computed cases, a GDL with a linear porosity 0.4x + 0.4 is the best.     

Effects of gas diffusion layer structure on the open circuit voltage and hydrogen crossover of polymer electrolyte membrane fuel cells

The clamping pressure of polymer electrolyte membrane fuel cells for vehicle applications should be typically high enough to minimize contact resistance. However, an excessive compression pressure may cause a durability problem. In this study, the effects of gas diffusion layer (GDL) structure on the open circuit voltage (OCV) and hydrogen crossover have been closely examined. Results show that the performances of fuel cells with GDL-1 (a carbon fiber felt substrate with MPL having rough surface) and GDL-3 (a carbon fiber paper substrate with MPL having smooth surface) are higher than that with GDL-2 (a carbon fiber felt substrate with MPL having smooth surface) under low clamping torque conditions, whereas when clamping torque is high, the GDL-1 sample shows the largest decrease in cell performance. Hydrogen crossover for all GDL samples increases with the increase of clamping torque, especially the degree of increase of GDL-1 is much greater than that of the other two GDL samples. The OCV reduction of GDL-1 is much greater than that of GDL-2 and GDL-3. It is concluded that the GDL-3 is better than the other two GDLs in terms of fuel cell durability, because the GDL-3 shows the minimum OCV reduction.     

Internal behavior of segmented fuel cell during cold start

In order to improve cold start capability and survivability of proton exchange membrane fuel cell (PEMFC), a fundamental understanding of its internal behavior is required. In this study, the cold start processes of a PEMFC with different operating conditions have been investigated, and the characteristics of current density and temperature distributions are studied through in-situ experiments with a printed circuit board (PCB). It is found that the start ability of PEMFC is strong at −3 and −5 °C, but weak at −7 and −10 °C. Also the self-start ability can be enhanced by decreasing the initial current load. Polarization curves show almost no degradation after successful cold start at −3 and −5 °C, while the PEMFC degrades a lot after failed cold start at low temperature like −10 °C. Also electrochemical impedance spectroscopy (EIS) shows a big degradation after galvanostatic mode cold starts. Local current density of segmented cell results shows that the highest current density is initially near the inlet region and then quickly moves downstream, reaching to the region near the middle eventually during the successful cold start process. However, during the failed cold start process, the highest current density is initially near the inlet region of the flow channels and quickly moves down stream, reaching the upper left corner region (A1) before shut down eventually. For both successful and failed cold starts, the highest temperature can be observed near the middle of the cell after the reaching of the highest current density.     

Effects of operating and design parameters on PEFC cold start

During startup from subzero temperatures the water produced in a polymer electrolyte fuel cell (PEFC) forms ice/frost in the cathode catalyst layer (CL), blocking the oxygen transport and causing cell shutdown once all CL pores are plugged with ice. This paper describes an experimental study on the effects of operating and design parameters on PEFC cold-start capability. The amount of total product water in mg cm⁻² during startup is used as an index to quantify the cold-start capability. The newly developed isothermal cold-start protocol is used to explore the basic physics of cold start, and the effects of purge methods prior to cold start, startup temperature and current density, and the membrane thickness are shown. The experimental data also confirm the current density effect predicted earlier by a multiphase model of PEFC cold start.     

Experimental study of the effect of dissolution on the gas diffusion layer in polymer electrolyte membrane fuel cells

The gas diffusion layer (GDL) is important for maintaining the performance of polymer electrolyte membrane (PEM) fuel cells, as its main function is to provide the cells with a path for fuel and water. In this study, the mechanical degradation process of the GDL was investigated using a leaching test to observe the effect of water dissolution. The amount of GDL degradation was measured using various methods, such as static contact angle measurements and scanning electron microscopy. After 2000 h of testing, the GDL showed structural damage and a loss of hydrophobicity. The carbon-paper-type GDL showed weaker characteristics than the carbon-felt-type GDL after dissolution because of the structural differences, and the fuel cell performance of the leached GDL showed a greater voltage drop than that of the fresh GDL. Contrary to what is generally believed, the hydrophobicity loss of GDL was not caused by the decomposition of polytetrafluoroethylene (PTFE).     

Effect of capillary pressure on liquid water removal in the cathode gas diffusion layer of a polymer electrolyte fuel cell

In order to investigate the effect of capillary pressure on the transport of liquid water in the cathode gas diffusion layer (GDL) of a polymer electrolyte fuel cell, a one-dimensional steady-state mathematical model was developed, including the effect of temperature on the capillary pressure. Numerical results indicate that the liquid water saturation significantly increases with increases in the operating temperature of the fuel cell. An elevated operating temperature has an undesirable influence on the removal of liquid water inside the GDL. A reported peculiar phenomenon in which the flooding of the fuel cell under a high operating temperature and an over-saturated environment is more serious in a GDL combined with a micro-porous layer (MPL) than in a GDL without an MPL [Lim and Wang, Electrochim. Acta 49 (2004), 4149–4156] is explained based on the present analysis.     

Reversible performance loss induced by sequential failed cold start of PEM fuel cells

This study correlates the post start cell performance and impedance with the cold start process in the subzero environment. The sequential failed cold starts are deliberately conducted as well as the start at small current density. Here the failed cold start means the cell voltage drops to or below zero within very short time during the start process. It is found that there are reversible performance losses for the sequential failed cold starts, while not obvious degradation and no recovery happen for the start at small current density. Using the thin film and agglomerate model, it is confirmed that this is due to the water blocking effect. Comparing the results from different start processes, a model with respect to the shifting of reactive region within the catalyst layer is applied to explain that the reversible performance loss is associated with the amount of the generated water or ice and the water location or distribution during cold start. The relationship of the cold start performance at high current density and the pore volume in the catalyst layer is also discussed.     

Parametric study of the porous cathode in the PEM fuel cell

The electrochemical behavior and the reactant transport in the porous gas diffusion layer (GDL) and catalyst layer (CL) are controlled by a large number of parameters such as porosity, permeability, conductivity, catalyst loading, and average pore size, etc. A three‐dimensional polymer electrolyte membrane fuel cell model is developed. The model accounts for the mass, fluid, and thermal transport processes as well as the electrochemical reaction. Using this model, the effects of the various porous electrode design parameters including porosity, solid electronic conductivity, and thermal conductivity of cathode GDL, and the catalyst loading, average pore size of cathode CL are investigated through parametric study. The model is shown to agree well with the experimental data of some porous electrode specifications. In addition, the model shows promise as a tool for optimizing the design of fuel cells. Copyright © 2008 John Wiley & Sons, Ltd.     

A dynamic model for hydrogen consumption of fuel cell stacks considering the effects of hydrogen purge operation

The actual hydrogen consumption of a fuel cell stack varies with a fixed time delay under the step load change. For each individual stack, the delay time in the step-up load stage is generally shorter than in the step-down stage. Due to the hydrogen purge operation, transient overshoots take place intermittently after the actual hydrogen consumption reaches the steady state, and the duration and peak value of such overshoots are distributed approximately within a fixed range. Based on the performance investigation mentioned above, an improved dynamic model for hydrogen consumption of a fuel cell stack considering the effects of hydrogen purge operation is introduced in this paper. Compared with the previous model, the suggested model indicates a better agreement between test and simulation, especially in the working condition of hydrogen purge operation.     

Experimental study on enhancing the fuel efficiency of an anodic dead-end mode polymer electrolyte membrane fuel cell by oscillating the hydrogen

This paper investigates how to improve the fuel efficiency of an anodic dead-end mode fuel cell for portable power generation. Generally, a periodic purge process in anodic dead-end operation is required to avoid anode flooding caused by back diffusive water from the cathode. However, during the purge process, small amounts of the hydrogen are discharged with the water, lowering the fuel utilization efficiency. Therefore, hydrogen pulsations are introduced and experimental attempt to minimize the purge frequency is conducted in this study. The experimental results indicate that pulsation reduces partial pressure of the water vapor in the anode channel, increasing the interval between purges by approximately three times, thus improving overall efficiency.