1. Introduction Fuel cells are regarded as the technology of choice to maximize the potential benefits of hydrogen in terms of efficiency. Today’s fuel cell plants exhibit efficiencies in the range of 40 to 55% LHV, almost independently of their size, while hybrid fuel cell – gas turbine cycles overcome 70% LHV (Figure 1). They are electrochemical devices that convert the chemical energy of a fuel directly to electricity, bypassing the thermodynamic limitations of conventional thermal engines. Their physical structure consists of an (solid or solidified) electrolyte in contact with two porous electrodes on either side. All types of fuel cells combine hydrogen and oxygen to produce dc electricity, water and heat. On the other hand, their sensitivity and endurance over time and above all their high initial costs are the main obstacles for broad commercialization [1-3].
Figure 1 Comparative efficiency (% LHV) of power generation systems [1,2]
Based on the type of the electrolyte the most common classification of fuel cells includes:
proton exchanging membranes (PEM) fuel cells (or polymer electrolyte fuel cells – PEFCs), with proton conducting polymeric membranes, transports hydrogen (fuel) cations, generated at the anode, to an ambient air exposed cathode, where they are electro-oxidized to water at low temperatures
solid oxide fuel cells (SOFC), which use oxygen conducting ceramic membranes to electo-combust H2, at the anode, by O2- anions provided by the cathodic reduction of ambient oxygen at high temperatures
molten carbonate fuel cells (MCFC), with alkali carbonate (in LiAlO2 matrixes) electrolyte, conduct CO32- anions, generated at an O2/CO2 exposed cathode to electroxidize H2 (fuel) at the anode and at high temperatures
alkaline fuel cells (AFC), with concentrated KOH (in asbestos matrixes) electrolyte, conduct OH- anions, generated at an O2/H2O exposed cathode to electroxidize H2 (fuel) at the anode and at moderate temperatures, and
phosphoric acid fuel cells (PAFC) with concentrated H3PO4 (in silicon carbide matrixes) electrolyte, which transports H+ cations, generated at the anode, to an ambient air exposed cathode, where they are electro-oxidized to water at moderate temperatures
as shown in Table 1. Regardless the specific type of fuel cell, gaseous fuels (usually hydrogen) and oxidants (usually ambient air) are continuously fed to the anode and the cathode, respectively. The gas streams of the reactants do not mix, since they are separated by the gas tight phase of the electrolyte. The electrochemical combustion of hydrogen, and the electrochemical reduction of oxygen, takes place at the surface of the electrodes, the porosities of which are to provide an extensive area for these reactions to be catalysed, as well as to facilitate the mass transport of the reactants/products to/from the electrolyte from/to the gas phase. Under closed circuit, the electrochemical reactions involve a number of sequential steps, including adsorption/desorption, surface diffusion of reactants or products, and the charge transfer to or from the electrode. Charge transfer is restricted to a narrow (almost one-dimensional) three-phase-boundary (tpb) among the gaseous reactants, the electrolyte, and the electrode-catalyst.
Besides their catalytic role, electrodes collect (anode) or supply (cathode) the electrons involved in the electrochemical reactions, and should be made of materials with good electrical conductivity. Continuous electrons supply (or removal) is necessary for the electrochemical reactions to proceed, resulting a constant electron flow from the anode to the cathode. At the same time, the electrolyte, by transporting reactants in the form of ionic species, completes the cell circuit. The electro-combustion of hydrogen sustains a difference in the chemical potentials of the electro-active species (conducting ions) between the two electrodes, which is the driving force for the ionic flux through the electrolyte, expressed as the open circuit potential of the cell or its electromotive force (emf).
Figure 2 Visualization of the physical structure of a planar fuel cell stack
Completing the physical structure of a fuel cell, a current collector, in closed contact to the porous electrodes, facilitates the electrons transport. In actual fuel cell devices, conductive interconnects are used to combine unit cells, in order to upgrade voltage, as illustrated in Figure 2 for the classical planar cell stuck assembly (gaseous fuels and oxidants flow through the current collector formatted channels, in a cross-flow pattern). These interconnects also serve as separator plates between the fuel and the oxidant gaseous streams of successive unit cells, so they must be impermeable to gases. Furthermore, they form the structures for distributing the reactant gases across the electrode surface.
2. Operation and Performance The maximum electrical work (Wel) of a fuel cell is given by the change in the free energy:
Wel = ΔG = − nFE 1.
of the overall (combined anodic and cathodic) electrochemical reaction:
aA + bB → cC + dD 2.
where n is the number of electrons participating the reaction, F is Faraday's constant (96,487 cb/mole), and E is the reversible potential of the cell (the open circuit voltage – emf). The difference between ΔG and ΔH is proportional to the change in entropy (ΔS):
ΔG = ΔH – TΔS 3.
where ΔH is the total thermal content of the feed and TΔS is the amount of heat produced by a fuel cell operating reversibly. The reversible potential of a fuel cell at temperature T is calculated from the ΔG of the cell reaction, at that temperature:
so that the reversible potential, becomes:
the general form of the Nernst equation, where ΔGo and Eo refer to 298 K. The ideal performance of a fuel cell is defined by its Nernst potential. Nernst equations, which quantify the relationship between the ideal standard potential (E°) and the ideal equilibrium potential (E), for the electrochemical reactions of the various types of fuel cells, along the typical values of Nernst potentials, at their operation temperatures are presented in Table 2 .
Table 2 Nernst equations and ideal voltages for the various types of fuel cells
overall (anode + cathode) reaction
H2 + 1/2O2 + H2O → 2H2O
H2 + 1/2O2 → H2O
H2 + 1/2O2 + CO2 → H2O + CO2
As noticed from Table 2, and because the entropy change of hydrogen combustion is negative, the reversible potential decreases with temperature by a factor of 0.84 mV/°C (assuming reaction product is liquid water). Furthermore, the actual cell voltage is smaller than its reversible one because of irreversible potential losses (or overpotentials), which originate either from the potential requirements to activate the electrochemical reactions (activation overpotential – ηact), the ohmic losses (ohmic overpotential – ηohm), or the losses due to the mass transport (gas and electrode’s surface diffusion) of the species participating the electrochemical reactions (concentration overpotential – ηconc). Activation overpotential is the primary source of voltage losses at low current densities, expressing, in a sense, the activation energy of the electrochemical reactions to occur, and is described by the Butler-Volmer equation, or its high field approximation, known as the Tafel equation:
(α, io are the charge transfer coefficient and the exchange current density, expressing the effectiveness of the electrode/electrolyte interface under the specific fuel (anode) or oxidant (cathode) conditions). Ohmic overpotential:
ηohm = iRohm 7.
increases linearly with current (since the resistance of the cell is essentially constant) becomes gradually predominant, as the current density increases. Finally, concentration (mass transport) losses:
(iL is the limiting or maximum obtainable current due to mass transport) are present over the entire range of current densities, but become prominent at high currents, where it becomes difficult for homogenous or surface diffusion to provide enough electro-active species to the electrode’s (or the tpb) reaction sites.
The operational cell voltage is the difference between the potentials of the cathode and the anode (as these potentials are altered due to the corresponding activation and concentration losses of each electrode) minus the ohmic losses, of the various stack components:
Current flow in a fuel cell results in a decrease of cell voltage, revealing the goal to minimize polarization, since the product of Vcell with the corresponding current density (at each point of the I – Vcell curve of Figure 2a) gives the specific (per unit of apparent electrode area – power density) electrical power output of the cell. This product tends to be minimized for low and high current densities (when current and operating voltages approach zero, respectively) and exhibits a maximum in between, as shown in Figure 2b.
maximum power density
Figure 3 Ideal and operational cell voltage (a), and the dependence of power density on cell voltage (b)
The thermal efficiency of fuel cells is defined as the amount of useful energy produced over the consumption of the chemical energy of the fuel (the amount of energy released in the form of heat, during the total combustion of the fuel, known as higher heating value of the fuel). Ideally the electrical work produced in a fuel cell should be equal to the change in Gibbs free energy, ΔG, of the overall reaction, and the ideal efficiency for reversible operation at standard conditions, will be:
The thermal efficiency of an actual fuel cell, operating irreversibly at temperature T, reduces to:
Thus, the efficiency of an actual fuel cell can be expressed in terms of the ratio of the operating cell voltage to the standard cell voltage.
A fuel cell can be operated at different current densities. It seems reasonable to operate the fuel cell at its maximum power density. However, decreasing current density below this value, brings the cell voltage closer to its reversible one, and increases the efficiency. On the other hand, the active cell area must also be increased in order to obtain a given power output, which means that, high efficiencies increase the capital cost, for a certain power level, although it decreases fuel requirements. Balancing between the above, it is usual practice to operate fuel cells to the left side of its power peak and at a point that yields as a compromise between low operating cost and low capital cost. It is interesting to observe that this situation provides fuel cells with the unique, among other energy conversion technologies, benefit of increased efficiency at part load conditions .
For a given fuel cell, it is possible to improve performance by adjusting temperature, pressure, gas composition, reactant utilizations, current density and/or other parameters which influence the ideal cell potential and the magnitude of the voltage losses. The selection of these parameters starts with defining the power level requirements for a specific fuel cell application. Flowingly, the voltage, and current requirements of the fuel cell stack and individual cells need to be determined, at certain operating temperature and (in some cases) pressure. Starting with temperature, its effect on the operational voltage is quite different from its effect on the reversible potential. The latest drops with temperature, while the operating voltage increases, due to the decrease of polarization losses (at higher temperatures the reaction and mass transfer are accelerated, and, in most cases, the ionic conductivity of the electrolyte – the main source of IR losses – increase), resulting in an overall improvement of the performance of the cell. Furthermore, the increase of operating pressure increases the partial pressures of reactants and consequently the reaction and mass transfer rates, improving performance and efficiency. However, pressure increases power needs to compress reactants, and capital costs.
Reactants utilization and gas composition also affect the fuel cell efficiency. Utilization factor (Uf) refers to the fraction of the total fuel or oxidant supply, that it is electrochemically consumed. In low-temperature fuel cells (PEFCs, AFCs and PAFCs), Uf is directly connected to H2 consumption, which is the only reactant involved in the electrochemical reaction:
where H2,in and H2,out are the molar flow rates of H2 at the inlet and the outlet of the fuel cell. Similar is the calculation for the oxidant utilization:
for PEMFs, SOFCs and PAFCs. The oxidant utilization in MCFCs and AFCs, where two reactant gases (O2 and CO2 in the first and O2 and H2O in the latest) are utilized in the electrochemical reaction, the oxidant utilization is based on the limiting reactant. Frequently O2, which is readily available from ambient air, is present in excess, and CO2 or H2O are the limiting reactants. A significant advantage of high-temperature fuel cells such as SOFCs and MCFCs is their ability to use CO as a fuel. The anodic oxidation of CO is slow compared to that of H2. However, the water gas shift reaction:
CO + H2O ↔ CO2 + H2 15.
rapidly reaches equilibrium at temperatures over 650 oC, and the equilibrium is shifted to the right because H2O and CO2are produced, over the anode. Thus, H2 utilization in high temperature fuel cells is defined as:
where the H2 consumed originates both from the H2 feed and the water gas shift reaction.
3. Types of Fuel Cells – Technology Status A determinative factor concerning the choice of fuel cell type, is hydrogen purity. Low temperature fuel cells require pure hydrogen, because the catalyst exhibits almost no tolerance to sulphur compounds and carbon monoxide, arising problems for hydrogen produced from natural gas. In contrast, SOFCs and MCFCs are more tolerant to impurities, due to their high operating temperatures, and CO, a contaminant for PEM fuel cells, can be used as a fuel. Polymer electrolyte fuel cells (PEFCs), solid oxide fuel cells (SOFCs) and molten carbonate fuel cells (MCFCs) are considered as the most promising candidates for stationary applications. PEM fuel cells generate power densities up to 4 A/cm2, at high efficiencies, while most technologies can hardly reach 1 A/cm2. This characteristic, in combination to weight, volume, and cost advantages, makes this type of fuel cells the most attractive for both mobile and stationary applications. SOFCs and MCFCs appear to have similar prospects to gain a part of the market in the coming decades [1,2].
3.1 Polymer Electrolyte Fuel Cells
A typical PEM fuel cell assembly includes the polymeric proton exchange membrane, on the opposite sides of which two porous electrocatalytic layers (electrodes) are suppressed. Two conductive and porous collectors are layered over the electrodes in closed conduct to the hard-plate interconnects, which form the reactants and products flow channels [1,2]. The proton exchange membrane consists of perfluorosulfonic acid polymers. These materials are gas-tight electrical insulators, in which the ionic transport is highly dependent on the bound and free water in the polymer structure. Nafion of the perfluorosulfonic acid family is the most commonly used material. Its Teflon-like structure is bonded to perfluorinated side chains with terminal sulfonic acid groups (SO3-). These groups are hydrophyllic, and the degree of their hydration is determinative for the ionic conductivity, the gas permeability, and the elasticity . Nafion membranes exhibit high thermal stability and chemical durability against Cl2, H2, and O2 attacks at temperatures up to 125°C [1,4,5], and their operational lifetime has been proved for over 50,000 hours. Apart Nafion, research has also focused on polybenzimidizole (PBI) electrolytes [2,6], which can operate at temperatures over 160°C, and annihilate CO poisoning [7,8].
The electrode-catalyst layer, for both the anode and the cathode, is in intimate contact with the membrane and consists of micro-dispersed platinum in a binder. The degree of intimacy between its particles and the membrane is crucial for the optimal proton mobility. The binder stabilizes the catalytic particles within the electrode structure and could be either hydrophobic (usually polytetrafluoroethylene) or hydrophyllic (usually perfluorosulfonic acid). Platinum loading, an important cost factor for PEFCs, has decreased to 1.0 mg Pt/ cm2 of membrane (total on anode and cathode) – from 2.0 – 4.0 mg Pt/cm2 .
The Nafion membrane is sandwiched between two porous and conductive carbon-based cloths, which support the membrane, diffuse the gaseous reactants and products and collects or supplies the electrons. This layer incorporates a hydrophobic material (usually polytetrafluoroethylene) to prevent withholding water within its pores, and to facilitate the removal of product water in the cathode. The current collecting cloth is in closed contact (the membrane-electrode-cloth assembly is suppressed between) with interconnecting carbon composite plates, for current collection, gas tightness, gas distribution, and thermal management. Flow paths for reactants, products and/or the cooler are printed on either side of these plates. In most PEFCs cooling is accomplished by circulating water that is pumped through integrated coolers within the stack, so that the temperature gradient across the cell is kept to less than 10°C.
Because of Nafion membranes, PEFCs operate at temperatures typically not higher than 60 – 80°C. At these temperatures CO is strongly chemisorbed on platinum, poisoning its catalytic activity and reducing the performance of the cell. The effect is reversible for only up to 50 ppm CO, while reformed and shifted hydrocarbons contain over 10,000 ppm CO. Although electrolysis H2 is favored for PEFC applications, in combined reformer – PEFC systems these concentrations can be eliminated by preferential oxidation (a process that selectively oxidizes CO in rich H2 streams, over precious metal catalyst). Recently PEFC research has focused on upgrading temperatures over 160°C using polybenzimidizole (PBI) electrolytes [1,2,6]. At these temperatures not only CO poisoning is eliminated, but also and because PBI requires lower water content to operate, water management is simplified [7,8].
As a proton is conducted through the membrane, it drags 1 – 2.5 water molecules with it [1,9], thus it is critical to maintain the water content of the electrolyte(the conductivity of the electrolyte is maximised when the membrane is fully saturated). On the other hand, and due to operation at less than 100°C at atmospheric pressure, liquid water is produced at the cathode. Thus, in case the anode is drier than the cathode, the back-diffusion of water from the cathode to the anode inhibits the protonic flow [9,10]. Furthermore, in case of water excess, the electrodes will be flooded blocking the pathways of reactants (products) to (from) the electrode/electrolyte interface, while in case of water shortage, the membrane will be dehydrated and destroyed . Despite the complexity, effective forms of water management have been developed based on continuous flow fields and appropriate operating adjustments focusing on controlled humidification of the anode gas in case of water shortage and temperature rise in case of flooding [12,13].
With operation voltages of 0.7 – 0.75 V, the maximum efficiency of PEFCs can be as high as 64%. In today’s applications, certain losses and ancillary equipment lower the efficiency, resulting in a situation in which PEFCs are more efficient than internal combustion engines only for operation at partial loads [2,14]. Current – Voltage performance characteristics of up to 5 kWe PEMFC are in the range of 0.5 – 0.76 V/cell for current densities of 0.55 – 1 A/cm2 and power densities of 0.22 – 0.57 We/cm2 . Operating temperature has a significant influence on performance (decrease of the ohmic resistance of the electrolyte and mass transport limitations), resulting a voltage gain of 1.1 - 2.5 mV / °C [1,16]. Improving the cell performance through temperature, however, is limited by water management issues. The goal for stationary PEFC operating life is 40,000 – 60.000 hours or 5 – 8 years . This life depends to a large extent on the operating conditions, such as the external temperature at start-up, excessive or insufficient humidification, and fuel purity. The principal areas of development concern improved cell membranes and electrode designs, targeting to improve performance and reduce cost [1,18,19,20].