PEM fuel cell current regulation by fuel feed control

We demonstrate that the power output from a PEM fuel cell can be directly regulated by limiting the hydrogen feed to the fuel cell. Regulation is accomplished by varying the internal resistance of the membrane-electrode assembly in a self-draining fuel cell with the effluents connected to water reservoirs. The fuel cell functionally operates as a dead-end design where no gas flows out of the cell and water is permitted to flow in and out of the gas flow channel. The variable water level in the flow channel regulates the internal resistance of the fuel cell. The hydrogen and oxygen (or air) feeds are set directly to stoichiometrically match the current, which then control the water level internal to the fuel cell. Standard PID feedback control of the reactant feeds has been incorporated to speed up the system response to changes in load. With dry feeds of hydrogen and oxygen, 100% hydrogen utilization is achieved with 130% stoichiometric feed on the oxygen. When air was substituted for oxygen, 100% hydrogen utilization was achieved with stoichiometric air feed. Current regulation is limited by the size of the fuel cell (which sets a minimum internal impedance), and the dynamic range of the mass flow controllers. This type of regulation could be beneficial for small fuel cell systems where recycling unreacted hydrogen may be impractical.     

Reforming methanol to electricity in a high temperature PEM fuel cell

In this paper we demonstrate for the first time a compact power unit, where a methanol reforming catalyst is incorporated into the anode of a PEMFC. The proposed internal reforming methanol fuel cell (IRMFC) mainly comprises: (i) a H₃PO₄-imbibed polymer electrolyte based on aromatic polyethers bearing pyridine units, able to operate at 200 °C and (ii) a 200 °C active and with zero CO emissions Cu–Mn–O methanol reforming catalyst supported on copper foam. Methanol is being reformed inside the anode compartment of the fuel cell at 200 °C producing H₂, which is readily oxidized at the anode to produce electricity. The IRMFC showed promising electrochemical behavior and no signs of performance degradation for more than 72 h.     

Ignition waves in a stirred PEM fuel cell

A mathematical model of slow transient behavior in an autohumidified stirred tank reactor (STR) polymer electrolyte membrane (PEM) fuel cell is developed. The key feature of the model is the positive feedback between current, water production, and membrane resistance which leads to two stable “ignited” states, corresponding to either a uniform current distribution or a partially ignited cell with localized current production. The switching between the two regimes is accompanied by hysteresis and transient behavior on the order of 2-4 h in a small cell. We compare the numerical results to experimental data gathered by [Benziger et al. 2005. Chemical Engineering Science 60 (6), 1743-1759] and show that the lateral diffusion of water within the ionomer membrane is a possible mechanism behind the hysteresis and slow transient behavior they observed.     

Ethanol crossover and direct ethanol PEM fuel cell performance modeling and experimental validation

In the present work, a one-dimension, steady-state and single phase model is developed with the purpose of describing the mass transport within a PtRu/Nafion^®-115/Pt membrane-electrode assembly and the performance of a direct ethanol proton exchange membrane fuel cell (DE-PEMFC). The effect of the most important cell operating parameters on the ethanol crossover rate and the fuel cell performance is investigated. According to the results, in the case of low current density values and high concentrations of ethanol aqueous solutions, ethanol crossover could pose serious problems to the DEFC operation. Moreover, it was pointed out that the ethanol crossover rate dependence on the ethanol feed concentration is an almost linear function presenting a maximum at about . A further increase of the ethanol feed concentration leads to a steep decrease of ethanol crossover rate. This behavior could be attributed to the membrane swelling which is responsible for the membrane volume fraction decrement. It was also found that by the aid of the same model the performance of a direct ethanol PEM fuel cell over three different anode catalysts can be predicted. A relatively good agreement between theory and experimental results related to both ethanol crossover rates and direct ethanol fuel cell performance was found.     

Modeling, simulation and experimental validation of a PEM fuel cell system

The aim of this work is the development and experimental validation of a detailed dynamic fuel cell model using the gPROMS modeling environment. The model is oriented towards optimization and control and it relies on material and energy balances as well as electrochemical equations including semi-empirical equations. For the experimental validation of the model a fully automated and integrated hydrogen fuel cell testing unit was used. The predictive power of the model has been compared with the data obtained during load change experiments. A sensitivity analysis has been employed to reveal the most critical empirical model parameters that should be estimated using a systematic estimation procedure. Model predictions are in good agreement with experimental data under a wide range of operating conditions.     

General relations for power generated and lost in a fuel cell stack

We derive the general electrodynamic relations for useful power generated by individual cell in a fuel cell stack and for power loss in a bipolar plate. The practical calculations with these relations are illustrated with the specially constructed one-dimensional case. The results are valid for stack of the cells of any type.     

Development and operation of a 150 W air-feed direct methanol fuel cell stack

A five-cell 150 W air-feed direct methanol fuel cell (DMFC) stack was demonstrated. The DMFC cells employed Nafion 117^® as a solid polymer electrolyte membrane and high surface area carbon supported Pt-Ru and Pt catalysts for methanol electrooxidation and oxygen reduction, respectively. Stainless steel-based stack housing and bipolar plates were utilized. Electrodes with a 225 cm² geometrical area were manufactured by a doctor-blade technique. An average power density of about 140 mW cm^–2 was obtained at 110 °C in the presence of 1 M methanol and 3 atm air feed. A small area graphite single cell (5 cm²) based on the same membrane electrode assembly (MEA) gave a power density of 180 mW cm^–2 under similar operating conditions. This difference is ascribed to the larger internal resistance of the stack and to non-homogeneous reactant distribution. A small loss of performance was observed at high current densities after one month of discontinuous stack operation.     

Three-dimensional numerical simulation of straight channel PEM fuel cells

The need to model three-dimensional flow in polymer electrolyte membrane (PEM) fuel cells is discussed by developing an integrated flow and current density model to predict current density distributions in two dimensions on the membrane in a straight channel PEM fuel cell. The geometrical model includes diffusion layers on both the anode and cathode sides and the numerical model solves the same primary flow related variables in the main flow channel and the diffusion layer. A control volume approach is used and source terms for transport equations are presented to facilitate their incorporation in commercial flow solvers. Predictions reveal that the inclusion of a diffusion layer creates a lower and more uniform current density compared to cases without diffusion layers. The results also show that the membrane thickness and cell voltage have a significant effect on the axial distribution of the current density and net rate of water transport. The predictions of the water transport between cathode and anode across the width of the flow channel show the delicate balance of diffusion and electroosmosis and their effect on the current distribution along channel.     

A distributed dynamic model for chronoamperometry, chronopotentiometry and gas starvation studies in PEM fuel cell cathode

Electrochemical systems differ significantly from conventional chemical systems. The response of voltage to changes in current and that of current to changes in voltage is much faster compared to typical transients observed in transport variables. In this work, the transient characteristics of various transport and electrochemical phenomena are studied in the PEM fuel cell cathode using a dynamic model. Model-based chronoamperometry and chronopotentiometry studies are performed to investigate the interactions among the various phenomena and the limiting mechanisms under various operating modes. The dynamic response of current to changes in voltage under chronoamperometry and that of voltage to changes in current under chronopotentiometry are found to be significantly different. Moreover, it is also observed through simulations that the dynamics in the output variables are strongly influenced by the operating cell voltage. Results from chronoamperometry studies are used to highlight the problem of oxygen starvation, which is also reflected by the magnitude of oxygen excess ratio or stoichiometric ratio. Results from step tests in chronopotentiometry studies are used to study nonlinearities in the response of voltage to changes in inputs such as, current and air flow rate.     

Quantifying desorption and rearrangement rates of carbon monoxide on a PEM fuel cell electrode

A simple procedure to quantify the rates of carbon monoxide (CO) desorption from, and simultaneous rearrangement on, supported platinum fuel cell electrode (Pt on Vulcan XC-72R) is reported. The surface coverage of CO on Pt electrode in equilibrium with bulk CO was measured from the anodic peaks in the CO stripping voltammogram. The decline in these surface coverages due to desorption and rearrangement, once CO was replaced by N₂ in the gas phase was recorded and used in conjunction with a kinetic model to quantify the respective rates. Two distinct CO oxidation peaks observed in the voltammogram due to the oxidation of two distinct ad-species, namely weakly and strongly adsorbed CO ( and ), were baseline corrected and deconvoluted using a bimodal Gaussian distribution. Saturation surface coverage of decreased with increasing temperature, while the opposite was true for . Rearrangement from to was faster than the desorption rate of either CO species. The desorption rate of was at least an order of magnitude lower than that of molecules at all temperatures studied. The activation energies for desorption of and were estimated to be 24.08 and 27.99 kJ/mol, respectively. The activation energy for rearrangement from to was 35.23 kJ/mol and that from to was 27.55 kJ/mol.     

Two-dimensional analysis of PEM fuel cells

This study reports a two-dimensional numerical simulation of a steady, isothermal, fully humidified polymer electrolyte membrane (PEM) fuel cell, with particular attention to phenomena occurring in the catalyst layers. Conservation equations are developed for reactant species, electrons and protons, and the rate of electrochemical reactions is determined from the Butler–Volmer equation. Finite volume method is used along with the alternating direction implicit algorithm and tridiagonal solver. The results show that the cathode catalyst layer exhibits more pronounced changes in potential, reaction rate and current density generation than the anode catalyst layer counterparts, due to the large cathode activation overpotential and the relatively low diffusion coefficient of oxygen. It is shown that the catalyst layers are two-dimensional in nature, particularly in areas of low reactant concentrations. The two-dimensional distribution of the reactant concentration, current density distribution, and overpotential is determined, which suggests that multi-dimensional simulation is necessary to understand the transport and reaction processes occurring in a PEM fuel cell.     

Modelling of performance of PEM fuel cells with conventional and interdigitated flow fields

A simple mathematical model is developed to investigate the superiority of the interdigitated flow field design over the conventional one, especially in terms of maximum power density. Darcy's equation for porous media and the standard diffusion equation with effective diffusivity are used in the gas diffuser, and a coupled boundary condition given by the Butler–Volmer equation is used at the catalyst layer interface. The performance of PEM fuel cells with a conventional flow field and an interdigitated flow field is studied with other appropriate boundary conditions. The theoretical results show that the limiting current density of a fuel cell with an interdigitated flow field is about three times the current density of a fuel cell with a conventional flow field. The results also demonstrate that the interdigitated flow field design can double the maximum power density of a PEM fuel cell. The modelling results compared well with experimental data in the literature.     

Measurement of in-plane effective thermal conductivity in PEM fuel cell diffusion media

Gas diffusion media used in polymer electrolyte membrane (PEM) fuel cells are highly anisotropic with significantly different transport property values in the through- and in-plane directions. In this study, experimental measurements of the in-plane effective thermal conductivity k for gas diffusion media used in PEM fuel cells have been carried out using a parallel thermal conductance technique. Conductivity values are measured at a mean sample temperature of 70 °C for six different material types and two different orientations in order to quantify the effect of PTFE content on thermal conductivity and to reveal any anisotropic behavior. The results vary from a minimum of k = 3.54 W/(m °C) to a maximum value of 15.1 W/(m °C) for various samples and configurations tested in this study, with an uncertainty between 1% and 2% for all the cases investigated.     

The role of mechanically induced defects in carbon nanotubes to modify the properties of electrodes for PEM fuel cell

The characteristics and reactivity of two anodes based on Pt supported on carbon nanotubes (CNTs) without or with defects induced by ball-milling are studied by SEM, TEM, cyclic voltammetry (CV) and single-cell measurements using a flow of pure H₂ or containing 50 ppm CO. It is evidenced that the presence of defects influences several properties and not only the dispersion of Pt particles. Therefore, the performances cannot be correlated neither with the geometrical surface area of Pt particles, neither with the electrochemical active surface area determined from CV tests. The presence of defects, enhancing the amount of surface functional groups on CNT, influences various aspects: (i) the efficiency of three-phase boundary and thus the transport of protons to or from the active metal particles, (ii) the resistance of electron transfer and (iii) the tolerance of the catalyst to CO poisoning. The latter is attributed to carbon functional groups in close contact with very small Pt particles favoring the reactivation of Pt sites poisoned by CO.     

High-surface-area CoTMPP/C synthesized by ultrasonic spray pyrolysis for PEM fuel cell electrocatalysts

Ultrasonic spray pyrolysis (USP) was used to synthesize a high-surface-area CoTMPP/C catalyst for oxygen reduction reaction (ORR). SEM micrographs showed that the USP-derived CoTMPP/C consists of spherical, porous and uniform particles with a diameter of 2-5 μm, which is superior to that with a random morphology and large particle sizes (up to 100 μm) synthesized by the conventional heat-treatment method. BET results revealed that the USP-derived catalyst had a higher specific surface area (834 m² g⁻¹) than the conventional one. Cyclic voltammetric, rotating ring-disk electrode (RRDE) and H₂-air PEM fuel cell testing were employed to evaluate the USP-derived CoTMPP/C. The kinetic current density of the USP-derived catalyst at 0.7 V versus NHE was two times higher than that of the conventional catalyst. Compared to Pt/C catalyst, the USP-derived CoTMPP/C catalyst showed a strong methanol tolerance and a higher ORR activity in the presence of methanol. In a H₂-air PEM fuel cell with USP-derived CoTMPP/C as the cathode catalyst, the cell performance was much higher than that with conventional heat-treated CoTMMP/C as the catalyst.     

Novel layer wise anode structure with improved CO-tolerance capability for PEM fuel cell

This study proposes a novel layer wise anode structure to improve the CO-tolerance ability and utilization efficiency of catalyst. The layer wise structure consists of an outer and an inner catalyst layer. The outer catalyst layer acting as a CO barrier is composed of two nano-Ru layers (0.06 mg cm⁻²) by magnetron sputtering deposition method and a Pt₅₀-Ru₅₀ layer (0.10 mg cm⁻²) by screen-printing method on the GDL. The inner catalyst layer providing the hydrogen oxidation reaction is a pure Pt layer (0.07 mg cm⁻²) prepared by direct-printing method on PEM. The roles of the outer and inner catalyst layer relating to the improvement of CO-tolerance ability and utilization efficiency of catalyst for the proposed catalyst layer structure are investigated in this paper. SEM, X-ray, EDS and EPMA analysis were used to characterize microstructures, phases, chemical composition and distributions for the obtained electrocatalyst layers. The hydrogen fuel containing 50 ppm CO/hydrogen fuel containing 50 ppm CO + 2% O₂ is continuously fed to the anode side to investigate the dependence of CO-tolerance ability over time for the MEAs, respectively. The results demonstrate that this proposed anode catalyst layer structure presents a superior CO-tolerance ability and performance to those of conventional and Huag's structures in both oxygen free and oxygen present CO containing hydrogen fuels as well as pure hydrogen fuel. The filtering effect of the outer catalyst layer causes the improved CO-tolerance capability.     

Electrochemical promotion of CO conversion to CO2 in PEM fuel cell PROX reactor

The possibility of electrochemically promoting the water–gas-shift reaction and the CO oxidation reaction in a PEM fuel cell reactor supplied with a methanol reformate mixture was investigated in PEM fuel cells with Pt or Au state-of-the-art E-TEK anodes, in order to explore the use of PEMFC units as preferential oxidation of CO (PROX) reactors. The electropromotion of CO removal was investigated both with air or H₂ fed to the cathode side and also by O₂ bleeding to the anode during normal PEMFC operation. It was found that the catalytic activity of the anode for CO conversion to CO₂ can be modified significantly by varying the catalyst potential. The magnitude of the electrochemical promotion depends strongly on the anodic electrocatalyst (Pt or Au), on the CO concentration of the fuel mixture, on the operating temperature and on the presence of oxygen. The electropromotion effect and the Faradaic efficiency were found to be much higher in CO-rich anode environments.     

The corrosion of PEM fuel cell catalyst supports and its implications for developing durable catalysts

Studying the corrosion behavior of catalyst support materials is significant for understanding the degradation of polymer electrolyte membrane (PEM) fuel cell performance and developing durable electrocatalysts. The oxidation of Vulcan carbon black (the most widely used catalyst support for PEM fuel cells) was investigated using various electrochemical stressing methods (fixed-potential holding vs. potential-step cycling), among which the potential-step cycling was considered to mimic more closely the real drive-cycle operation of vehicle PEM fuel cells. The oxidation of carbon was accelerated under potential-step conditions as compared with the fixed-potential holding condition. Increasing the potential-step frequency or decreasing the lower potential limit in the potential-step can further accelerate the corrosion of carbon. The accelerated corrosion of carbon black was tentatively attributed to the cycle of consumption/regeneration of some easily oxidized species. These findings are being employed to develop a test protocol for fast-screening durable catalyst support.     

Effect of hydrophobicity and pore geometry in cathode GDL on PEM fuel cell performance

The effect of hydrophobic and structural properties of a single/dual-layer cathode gas diffusion layer on mass transport in PEM fuel cells was studied using an analytical expression. The simulations indicated that liquid water transport at the cathode is controlled by the fraction of hydrophilic surface and the average pore diameter in the cathode gas diffusion layer. Deposition of a hydrophobic microporous layer reduces the average pore diameter in the macroporous substrate. It also increases the hydrophobic surface, which improves the mass transport of the reactant. The optimized hydrophobicity and pore geometry in a dual-layer cathode GDL leads to an effective water management, and enhances the oxygen diffusion kinetics.