Fuel Cell explained

All the material available in this section has been produced by Catherine Lepiller ,phD in electrochemistry and chief scientist at Pragma Industries.

It can be used freely. You can copy/paste the text if you wish. We just ask you to refer to us as information source.
You can download the full document in pdf format here.

see the article below :


Article Index

  1. Origins and principle

1. Back to the origins: from the gas battery to the fuel cell

The seminal work of William Grove on fuel cells in 1839 is well known nowadays but at the time of his invention, the scientist rather called the device a “gas voltaic battery”. Therefore, while year 1839 is unambiguously considered at the birth date of fuel cells, the first official appearance of the term “fuel cell” will only be found in a publication of the Transactions of the Faraday Society almost one century later, in 1922*.

Grove’s first invention, a battery called “Grove cell” was used by the American Telegraph Company due to its high current output until 1860. After having started as a lawyer Grove will turn to a professor of physics, and after then switch between a legal and a scientific carrier several times. With probably some help from the German chemist Christian Schönbein, a friend and colleague with whom he is exchanging fruitful ideas; he successfully reverses the electrolysis of water that has been discovered in the early 1800s by English chemists William Nicholson, Anthony Carlisle and experimented by Humphry Davy. Grove constructs a cell consisting of two separate sealed compartments, each one having a porous platinum foil electrode dipped in aqueous sulfuric acid and being fed by hydrogen gas and by oxygen gas, respectively. The experiment however, does not generate enough electricity to do useful work. Therefore, Grove combines in series several sets of electrodes and obtains the actual “gas battery”. He shows this way that a constant current can be drawn between electrodes and observes that water and heat are produced as byproducts. Yet, he is unable to quantify the reaction products and study in much more detail the system he has created. The reason is that these questions could not possibly be answered due to the lack of a comprehensive theory and adequate equipment in the 1840s. He is also conscious that the chief issue is to increase the “surface of action” between the components and comes close to the idea of gas electrodes as used in current fuel cells. It really seems that the man was too much ahead of his time.

Unlike batteries such as the Grove cell, the gas battery is left as a scientific curiosity during a large part of the XIXth century. Times were not up to, should we say. Despite continued technological advancements, applications in the real world have not come to the mind of engineers and inventors. Nevertheless, during the same period the debate is lively among scientists trying hard to clarify the basic principles of electrochemical phenomena. Grove’s gas battery becomes a perfect practical illustration of these theoretical discussions. In order to explain the origin of current flow between certain materials, a “contact” theory involving mere physical contact and a “chemical” theory involving a chemical reaction are opposed. Schönbein and Grove are in favour of the chemical theory. The truth is actually in-between, and after a long controversy it is eventually established that in a gas battery the reaction will only occur in the contact zone between reactant, electrode and electrolyte.

None but the main founder of modern physical chemistry, the Russan-German Wilhelm Ostwald eventually brings a decisive contribution to the theoretical and experimental understanding of fuel cell reactions in the 1880s. By skillfully associating measurements of physical properties and chemical analysis, and thanks to his work on metal catalysis, he succeeds to explain how fuel cells operate: not only the function of the different components (electrodes, electrolyte and gas reactants) but also their thermodynamically- and kinetically-driven interactions at the interface between the electrode and the electrolyte under the presence of a fuel. This breakthrough in the scientific knowledge about gas batteries is the necessary open door for practical attemps to make them become a reality. At the turning point between the XIXth and XXth centuries, the first systems begins to come out from European and US laboratories as researchers are examining the possibility of converting coal or coal gas directly into electricity. Coal being a fuel, Grove’s gas battery changes its name into “fuel battery” and finally “fuel cell” as we know today.

In agreement with Grove’s request for a higher contact area between fuel cell components, early developments (50 years after Grove’s experiments!) are devoted to experimental improvements in the design, e.g., introduction of more efficient platinum black powder as a catalyst coated on the bulk platinum electrodes and addition of a porous diaphragm in order to impregnate the liquid electrolyte. In 1889 German chemist Ludwig Mond and his assistant Charles Langer built a device running on air and coal gas known as the “Mond gas”. At the same time other systems are setup by US and French teams. Typical problems experienced by researchers are due to materials chosen in the different parts or gas leakages between compartments, which prevent from reaching high voltages upon series combination of the unit cells and cause limited durability. It is also observed that only high-cost precious metals such as platinum can speed up the reaction with a valuable efficiency. This is obviously a serious drawback for practical applications of the process.

The end of the XIXth century is also time for a prospective debate about the possible direct production of electricity from inexpensive coal and combustible gasses: such a perspective is asserted by some authors nothing less than a revolution while strongly tempered by others, (e.g., controversy in the Electrical World Journal in 1895). Finally, consensus is made on the fact that gas batteries are complex and costly systems are unable to compete with simpler batteries. In the following period, batteries undergo continued development for several applications including cars, whereas “gas” batteries are put away back into labs for a few more decades. Remind that in the first quarter of the XXth century, one third of all automobiles on the roads were battery-powered electric vehicles. What visionary mind could have predicted at this time that it would require hundred more years or so before gas batteries, now called fuel cells, could come close to being “ready for market”?

It is therefore at the laboratory level that the next improvements are obtained on fuel cells. The importance of kinetics in the electrochemical reactions is discovered. At the beginning of the XXth century, new electrolyte materials performing at higher temperatures than aqueous solutions are explored and this will progressively lead to the various types of modern fuel cells: melted carbonates, solid oxides, phosphoric acid. Further historic information about practical developments of a specific fuel cell technology will be found in the corresponding pages of this website. For example, Francis Bacon’s pioneering work on acid phosphoric fuel cells is reported in the AFC page.

* According to an alternative thesis, the first fuel-cell system builders Lang, Mond and Jacques were also the inventors of the term “fuel cell” around 1890, but the exact birth date remains elusive and unsolved.

2. The basic principle of a fuel cell

Basically, a fuel cell is a device that converts directly the chemical energy stored in gaseous molecules of fuel and oxidant into electrical energy. When the fuel is hydrogen the only by-products are pure water and heat. The overall process is the reverse of water electrolysis. In electrolysis, an electric current applied to water produces hydrogen and oxygen; by reversing the process, hydrogen and oxygen are combined to produce electricity and water (and heat).

A fuel cell can be seen with profit as a “chemical factory” that continuously transforms fuel energy into electricity as long as fuel is supplied. However, unlike internal combustion engines that can be regarded as factories as well, fuel cells rely on an electrochemical reaction involving the fuel, and not on its combustion.

During combustion, molecular hydrogen and oxygen bonds are broken and electrons reconfigure into molecular water bonds at a picosecond length scale. There is no possible way to “catch up” these free electrons and the net energy difference between molecular bonds in products vs. reactants can only be recovered in the most degraded form of energy, i.e. heat. A Carnot cycle involving the transformation of heat into mechanical and electrical energy is then involved in conventional methods for generating electricity: these successive steps of transformation of energy severely limit the overall efficiency of the process (which is by definition the product of the efficiency of the different steps).

In a fuel cell, the direct conversion of the chemical energy of covalent bonds into electrical energy is made possible through the spatial separation of the hydrogen and oxygen reactants by the electrolyte also called the “separator”. The electron transfer necessary to complete the bonding reconfiguration into water molecules occurs over a much longer length scale. This allows direct collection of electrons as a current in fuel cells and leads to fuel efficiencies two to three times higher than in internal combustion engines (depending on the fuel cell technology).

Unlike batteries, there is no chemical transformation of any component of the fuel cell device during operation and it can generate power without recharging, as long as it is being fed with fuel.

The unit fuel cell structure called the membrane electrode assembly (MEA) typically consists of an electrolyte in contact on its both sides with two electrodes, one negative electrode (anode) and one positive electrode (cathode). Fuel is continuously fed to the anode side and oxidant is continuously fed to the cathode side.

Fuel cell reactants are classified as fuels and oxidants on the basis on their electron donor and electron acceptor properties. Fuels include pure hydrogen and hydrogen-containing gases, e.g. methanol, ethanol, hydrazine, natural gas, gasoline, biogas, diesel, etc. Oxidants mainly include pure oxygen and oxygen-containing gases e.g. air, or halogens e.g. chlorine.

In the most straightforward case, i.e. the hydrogen fuel cell the combustion of hydrogen into water is split into two electrochemical reactions occurring at the anode and cathode, respectively, which are termed as the two half-cell reactions:

Hydrogen oxidation reaction (HOR) at the anode H2 = 2H+ + 2e

Oxygen reduction reaction (ORR) at the cathode ½ O2 + 2H+ + 2e = H2O

Combination of the two half-cell reactions gives the overall combustion reaction:

H2 + ½O2 ® H2O

In any fuel cell configuration, the role of the electrolyte is crucial because it must insulate the two half-cell reactions electrically in a strict sense, while allowing the ionic passage of protons produced at the anode to the cathode side where they will combine and form a molecule of water. Hence, electrolytes are both proton conductors and electric insulators.

The third requirement for electrolytes is impermeability to gases in order to separate the anodic and the cathodic compartments, and thus prevent parasitic reactions due to gas crossover. Finally, the electrolyte has to be chemically resistant to any reactant or product during the process.

As passage of electrons is hindered through the electrolyte, they are forced to flow another way. To this purpose, electrodes are connected to an external electrical circuit and instead to follow protons the electrons take this second pathway. This allows direct collection of electricity. Depending on the type of fuel cell, the most suitable electrode materials are of various natures: metals or oxides, catalyzed or not. They are described in the section relative to the members of the fuel cell family. The common feature of fuel cell electrodes is a high surface area in order to maximize each half-cell reaction zone; therefore they are relatively porous compounds.

Every type of fuel cell is characterized by its own particular geometry, dimensions, and materials; yet, the core of the device remains the same: it consists of an electrolyte, two electrodes, and two gas backing layers and most often, bipolar plates separating unit cells.

For the gas backings five main requirements must be fulfilled:

1. Good electronic conductivity to transport the electrons from the electrochemical oxidation of hydrogen most efficiently;

2. High gas permeability to allow easy access of the gas reactants from the feeding source to the reaction zone;

3. High porosity to optimize product water management in the system;

4. Good resistance strength to give a mechanical support to the MEA;

5. High corrosion resistance to the acidic environment in the fuel cell.

The bipolar plates are the interconnecting components that collect the electrons and drive them to the external circuit. They are grooved with channels for gas flow input and output and must manage water as well as possible. The design and the geometric dimensions of the channels (in the order of 1 mm) are crucial for obtaining a homogeneous transport of gases on the whole surface of electrodes, evacuate liquid water droplets formed by the fuel cell reaction, thus achieving stable continuous operation. As every component in a fuel cell, they must be corrosion-resistant; but unlike gas backings, the bipolar plates must be gastight.

  1. Benefits of the fuel cell technology

1. Stacks and systems

Now moving from the single fuel cell unit to real-world systems, what do we have to add to get them all setup and why?

Similar to all electrical devices the output power of a fuel cell is equal to the current multiplied by the voltage. While the current may be in theory indefinitely increased by increasing the reaction area between hydrogen- and oxygen-containing reactants, the voltage, i.e. the potential difference between the anode and cathode, is thermodynamically limited to a little more than 1 V (1.23 V under the normal conditions of temperature and pressure) by the nature of the two half-cell reactions in a fuel cell: hydrogen oxidation reaction (HOR) at the anode and oxygen reduction reaction (ORR) at the cathode (cf. The basic principle of a fuel cell section).

Moreover, potential losses inevitably occur in a fuel cell due to slow kinetics of the electrode reactions, especially at the cathode where the reaction rate is about 100 times slower than at the anode (activation overpotential); intrinsic resistances of the different components and contact resistances between interfaces (ohmic overpotential), and transport resistances of the reactants (concentration overpotential). Therefore, under operational load the actual voltage of a single fuel cell is in the 0.6-0.7 V range.

Useful voltages are generally achieved by interconnecting multiple unit fuel cells in series. This is the concept of “stacking”. The stack’s final output voltage will depend on the number of cells and the available current will be proportional of the total surface area of the cells. In this configuration, the conductive interconnecting element is in contact with both the anode of one cell and with the cathode of the adjacent cell, hence the name “bipolar plate”. Flow channels are grooved on each side for gas distribution and water removal. Bipolar plate materials are highly impermeable to gases in order to avoid harmful fuel and oxidant mixtures: these materials are mainly graphite, polymer-graphite composites and metals such as stainless steel or aluminum (most often coated with a corrosion-resistant alloy).

Bipolar stacking has been up to now the most simple and the most conventional configuration in most types of fuel cell systems, particularly low-temperature systems. For high-temperature systems such as SOFCs however, sealing issues due to large temperature gradients during operation have driven research toward alternative arrangements, leading to the development of a tubular design.

In tubular stacking, the elements of the fuel cell assembly (anode/electrolyte/cathode) are arranged concentrically forming a hollow cylinder. Fuel is fed on the anode side, either through the inside or along the outside of the cylinder, and oxidant is fed on the cathode side. Series connection is accomplished by vertical addition of the cells (in the height direction) while parallel connection is accomplished by horizontal addition of the cells (in the same plan). The tubular design is well suited for high-temperature applications since it minimizes the number of seals in the fuel cell system thus alleviating problems due to unmatching expansion coefficients.

Planar stacking is a second alternative to the bipolar arrangement, in which cells are connected laterally rather than vertically. Several planar designs have been explored, mostly for small-scale systems: the banded-membrane design, in which the anode of one cell is connected to the cathode of the adjacent cell across the band; and the flip-flop design, in which there is interconnection of unit cells on the same side of the band thanks to alternate anodes and cathodes. The main advantage of this third arrangement is a better volumetric packaging, yet at the expense of increased resistance losses.

Besides the fuel cell stack, referred to as the fuel cell subsystem, the other subsystems that are needed to keep the whole system running can be classified into three categories:

1. The thermal management (cooling) system

2. The fuel delivery/processing system

3. The power electronics (and safety) system for power regulation and monitoring

The components that draw electrical power from the fuel cell causing parasitic power losses are called ancillaries. For example, an actively cooled fuel cell system will employ an ancillary device like a fan, a blower or a pump for cooling fluid circulation. Ancillaries include thermal, water and air management systems.

1. As fuel cells are usually about 30-60% electrically efficient (depending of the type of fuel cell), the balance of energy is released in the form of heat and this has to be managed by the system in order to maintain the thermal gradients inside the stack at the lowest possible level (within a few °C) and ensure stable operation. A cooling system is required for fuel cells that cannot benefit from natural heat regulation by the ambient, i.e. all systems except small PEMFCs (output power < 100 W). The cooling fluid can be either a gas (air), or a liquid (distilled water or aqueous glycol-based solution) depending on the heat dissipation capacity needs and the other characteristics of the fuel cell system. Given that the heat capacity of liquids is much greater than that of gases; consequently, small liquid-cooled devices will generally be far more efficient than their massive gas-cooled equivalents.

In advanced fuel cell systems, the heat released by the stack can be purposely recovered for internal (1,2) and/or external (3) heating. Examples follow:

(1) Heat can be used for conditioning reactant gases = pre-heating and humidification;

(2) Heat can be used for providing energy to the endothermic reforming reaction of the fuel (see below);

(3) Heat can be used for providing space and/or water heating in a house, passenger compartment warming in a car, etc.

Cogeneration by heat recovery is a powerful means to increase the overall efficiency of fuel cells systems up to 80-85%. It is very advantageous in high-temperature fuel cell systems, mainly PAFCs and SOFCs.

2. Given that almost all practical fuel cells today use hydrogen or compounds containing hydrogen as a fuel, there are two primary options to feed a fuel cell: (1) in a direct way by pure hydrogen or (2) by integrated upstream processing of a “hydrogen carrier”in a reformer unit.

(1) In the first case, hydrogen is produced outside the fuel cell system in an industrial process (steam reforming for example), and is ready for direct use. The fuel management subsystem will include a hydrogen reservoir related to the physical state of hydrogen stored: high-pressure gas cylinder (up to 700 bars) for compressed gas, double-walled insulator under cryogenic conditions (22 K) for liquid hydrogen in extreme situations where mass storage capacity is especially important, e.g. space conditions, or low-pressure container for metal hydrid compounds ground into extremely fine powders. The advantages of direct hydrogen feeding include high performance, simplicity, and the elimination of impurity concerns. But the current storage options, mainly in the form of compressed gas or reversible metal hydride, are not optimal yet.

(2) In the second case, the system is more complex. Since hydrogen is not available as is, it must be derived from hydrogen-containing fuels called “hydrogen carriers” that are widely available in the industry, like methane, methanol, diesel or gasoline. Except a few hydrogen carriers that are directly usable in fuel cells systems including methanol in DMFCs and methane in SOFCs of MCFCs, a vast majority of them must be processed before they enter the fuel cell. This is possibly achieved in two different ways:

1. By direct electro-oxidation

2. By chemical reforming

A further distinction must be made between external reforming where the reaction takes place in a reformer separated from the fuel cell, and internal reforming where the reaction takes place at the catalyst surface inside the fuel cell.

1. Direct electro-oxidation of the carrier fuel into hydrogen is attractive because it avoids the extra step of reforming it prior to the fuel cell reaction and all chemical reactors associated with it. Direct methanol fuel cells are based on this principle, and other simple compounds like ethanol and formic acid can also be employed. Unfortunately, the overall electrical efficiency of this category of fuel cells is significantly reduced due to the complexity of the reactions. As a result, the energy density gained by the absence of a reformer or a fuel reservoir can be largely offset by the low fuel efficiency and the need for larger stacks. Direct electro-oxidation is best applied in portable applications, where simple systems, minimal ancillaries, and low power are needed.

2. External reformers are composed of several devices for successively treating chemically or physically the gas reactant (hydrogen carrier) and the products (including hydrogen). Several ways are possible, and the exact conditions will vary with the process and the hydrogen carrier. The most used process is steam reforming: fuel molecules are burned over a catalyst (nickel-, copper oxide- or zinc-based) under the presence of water steam at a few hundred degrees Celsius (³ 700°C), according to the reversible reaction:

CxHy + xH2O(g) « xCO + (1/2y + x)H2 Þ H2,CO,CO2,H2O

This is an endothermic process, whose yield can be increased in the presence of excess water vapor via the water-gas shift reaction involving the product CO:

CO + H2O(g) « CO2 + H2

Prior to the reforming reaction, the fuel may have to be desulfurized (i.e., decontaminated from its sulfur compounds) and heated. After reaction, a hydrogen-containing gas mixture is obtained that in certain cases has to be purified in multi-processing steps. This additional post-treatment is required to remove poisons for electrodes and feed the fuel cell with a pure hydrogen gas, which is an important requirement for low-temperature systems.

Internal fuel reforming is possible for high-temperature fuel cells with certain fuels. In this case, the fuel is mixed with steam prior entering the fuel cell anode where it is both reformed into hydrogen and the usual co-products CO and CO2 and then split into protons in the fuel cell reaction. Under these high-temperature conditions, the presence of carbon monoxide is not an issue anymore since it serves as fuel. It is further processed in situ like hydrogen thus contributing to the fuel cell net efficiency. Although the different interplaying parameters are difficult to optimize, internal reforming is a promising solution because it gives an elegant (and economically winning) answer to a complex question.

Fuel reforming is best applied in stationary applications, where fuel flexibility is important and the excess heat can be managed inside or outside the system. However, fuel reforming technology is not a current choice of authorities for transportation applications since the existing technologies do not meet the technical or economic targets, and only marginal improvement is expected in efficiency and emissions between a hybrid vehicle and an FC vehicle equipped with on-board reforming.

For any of the fuel delivery/processing systems considered previously, gas pumps are used to feed the gas reactants in the fuel cell, and a water purge system must also be integrated.

3. Last but not least, it is necessary to manage the direct power output of the fuel cell into usable power. The power electronics subsystem consists of:

(1) Power regulation;

(2) Power inversion;

(3) Power monitoring and control;

(4) Power supply management.

Power conditioning corresponds to regulation and inversion of the fuel cell power output.

(1) Regulation allows delivery of power at a voltage level that is stable over time from a fuel cell output power that most often is not. Fuel cell power is generally regulated by DC/DC converters, which transform the fluctuating direct current (DC) voltage input into a fixed, stable DC voltage output. A DC/DC converter is generally 85-98% efficient.

(2) Inversion means converting the DC power provided by the fuel cell into alternative current (AC) power consumed by most electronic devices, electric motors, and the electrical grid. This task is performed by a DC/AC converter. Similar to DC/DC converters, DC/AC converters are 85-97% efficient. Consequently, these devices taken together yield a 5-20% decrease in fuel cell electrical efficiency, which is far from negligible. Selections of optimal options for a given fuel cell stack technology, geometry and size in a given environment is therefore an essential point.

(3) Power monitoring and control includes system monitoring by gauges, sensors, etc. that measure the conditions of the fuel cell, system actuation by valves, pumps, switches, fans, etc. that regulate them, and a central control unit that mediates the interaction between sensors and actuators. Most control systems are based on feedback algorithms to maintain stable fuel cell operation, i.e. different sets of feedback loops are implemented between stack monitoring elements and ensuing corrective actions by actuators.

(4) Power supply management is the part of the power electronics subsystem that adapts the electrical power output of the fuel cell to the load requirements. Depending on the application, the demand may be driven be toward short-time and/or large-scale load changes that the fuel cell alone is not able to answer due to lag times in system ancillary components such as compressors and pumps. The dynamic response of the fuel cell can be enhanced by energy storage buffers like batteries or capacitors. The response time will be reduced from seconds or even hundreds of seconds to milliseconds. In the case of stationary fuel cell applications, the power supply management subsystems must also incorporate a special control device for interacting with the local grid, allowing for example shutdown or disconnection during a power outage.

The target application ultimately dictates the fuel cell system design and the choice of fuel delivery. In portable systems for instance, there is a strong incentive to minimize the size of components and avoid the use of ancillaries. Direct or reformed methanol fuel systems may provide energy density improvements compared to direct hydrogen storage solutions. A delicate trade-off is necessary between the size of the fuel cell unit and the size of the fuel reservoir. In utility-scale stationary power generation, durability and efficiency are of prime importance. Reformed natural gas and biogas are the leading fuel solutions due to their wide availability and low cost.

2. Advantages


Fuel cells combine many of the advantages of both internal combustion engines (ICE) and batteries. Thanks to the direct conversion of chemical energy into electrical energy, fuel cells are 2-3 times as efficient as ICEs for vehicle propulsion: the net electrical efficiency of a PEMFC ranges between 40 and above 50% in a driving schedule, which is favorably compared to the 21-24% efficiency range of ICEs “from well to wheel”, i.e. accounting for the type of fuel used and its entire life cycle. Now if we add in the calculations the reforming process of gasoline and methanol or the use of compressed hydrogen in the calculations the efficiencies are 33, 38, and about 50%, respectively.

Interestingly, the fuel cell efficiency does not drop for small systems because it does not depend on its size: unlike gas turbines for example that suffer from scale effects, small fuel cell devices are quite as efficient as larger ones. Accounting for energy losses in ancillaries, the efficiency is somewhat lowered but in any case is higher than conventional systems.

In cogeneration mode with simultaneous use of electricity and heat, a global efficiency has to be considered. This explains that stationary systems like fuel cell power plants can attain energy efficiencies of 85%. Thanks to the thermal yield the global efficiency is roughly doubled with respect to the use of electrical yield only.

Reduced emissions

Because fuel cells are electrochemical systems and do not rely on combustion, they can be considered the cleanest fuel-consuming energy technology, with near-zero smog-causing emissions. They produce benefits in all applications: power generation, industrial equipment, transportation, military power and consumer electronics.

The emissions produced by a fuel cell system strongly depend on the fuel used and its origin. For example, a FC vehicle produces only water if it is fed by compressed hydrogen, some CO, CO2 and CH4 if it is fed with ethanol, and additional SO2 if it is fed with gasoline. Under fuel cell operation, undesirable products such as carbon monoxide CO, sulfur oxides SOx or nitrogen oxides NOx, ashes and carbon particulate emissions are virtually zero.

Best results are achieved with a fuel cell system running on pure hydrogen, the hydrogen being produced by water electrolysis from renewable electricity. Emissions of pollutants are of course increased for electricity coming from the grid, i.e. a mixture of thermic, nuclear and renewable sources.

Reliability, low maintenance and quietness

Fuel cells can help provide stability and continuity to the electric grid since they can maintain a continuous base power in parallel with or independent of the grid. Fuel cells provide high quality power without any risk of power outage. They have more predictable performance over wider operating temperature ranges than lead acid batteries.

Fuel cells can be recharged everywhere within a few minutes by refueling while batteries have to be plugged in for time-consuming recharge (and eventually disposed of when their accumulation capacity has decayed too much). They operate at constant peak performance from fuel replenishment to depletion. Therefore working time is well-known and directly proportional to the amount of fuel supplied.

Fuel cells systems have practically no rotating or even moving parts. Certain types of fuel cells (PEMFC, SOFC) are all solid state thus close to mechanically ideal. This means less noise and potentially reduced maintenance work (and related costs) besides refueling. Stationary fuel cells require only minimal maintenance (once every one to three years) compared to monthly or quaterly site visits to lead acid battery-based installations.

Fuel cells are relatively silent systems making them suitable for residential areas. The only parts that are liable to cause moderate noise are the pieces of ancillary equipment like fans, compressors and pumps. Noise levels measured on stationary systems are typically as low as 50-60 dB.


Fuel cells are powered by hydrogen, the most abundant element in the Universe. Hydrogen can be produced from a variety of sources including fossil fuels, natural gas, methanol, and various renewable energy sources: wind, photovoltaic, geothermic, waves, etc. This is a keypoint asset from the perspective of greenhouse gas reduction and follow-on process of the Kyoto Conference. For more information please refer to the Hydrogen section.

Fuel cells are essential to achieving carbon reduction goals, with CO2 reduction ranging from 40% or better using conventional fuels to nearly 100% using renewable derived hydrogen, as compared to conventional power sources. Fuel cells can contribute to the world’s end of dependence on hydrocarbons. They can greatly simplify the sequestration of CO2 from hydrocarbon fuels, enabling the use of domestically-produced fuels including coal, biomass and hydrogen.

Due to their low environmental footprint, fuel cells are a realistic option in several fields concerned by the climate change debate: automotive, residential, industrial.


Fuel cells offer higher energy density and higher storage capacity than batteries, and thus a good compactness, which is interesting especially for portable applications.

Modularity and flexibility

Fuel cells allow independent scaling between power (determined by the fuel cell size) and capacity (determined by the fuel reservoir size). The fuel cell size can be adapted by simply changing the number of elementary cells and the active area. Scale-up is therefore very easy, from the W range of a cell phone to the MW range of a power utility plant. For miniaturized systems, techniques derived from microelectronics are being developed.

Fuel cells are the ideal solution when space is limited or weight is a concern, offering clean and quiet operation in a wide range of installation conditions. For example, the reduced footprint requirements for normal rooftop loading limits, and zero-emission combined with silent operation make them highly suitable for indoor/outdoor, urban/rural applications.

In addition, they can be fueled by a variety of fuels including intermittent renewable energy like solar energy or wind turbines as the primary source, in conjunction to hydrogen storage.

3. Issues

There are three main barriers remaining to widespread adoption of the fuel cell technology:


Firstly, fuel cells and the hydrogen delivery-storage infrastructure needed to support them still cost far too much to be competitive with internal combustion engines (ICEs). In the automotive sector, the low-volume costs for PEMFCs are currently in excess of $1,800/kW vs. an automotive target price of $40/kW. The lion’s share of cost lies between bipolar plate (45-50%) and membrane-electrode assembly (35%) manufacturing. Not only is platinum catalyst a very expensive metal, but the other stack components are still driving the prices up. In the stationary sector, the threshold price is estimated somewhere around $1,500/kW, where fuel cells systems can start to compete economically with other applications, and will actually compete with industrial batteries.

The main reason is that whatever the technology considered, manufacturers haven’t made many fuel cells to date, and they have neither the capital equipment nor the experience with mass production that would be required to bring the costs down at some point. While material costs should remain stable with volume, one can forecast that production costs will massively go down.

As long as production costs are higher than the acceptable price for customers, then someone – either the producer or the government state – must make up the difference between the cost of production and the price at which it is sold. According to market studies based on this approach, it will take nearly 20 years before producers start to see payback of their investment, which means either an inacceptable 20-year period of “sunk” costs or more obviously a massive government subsidy program.

One way to overcome the cost issue is to enter the market at higher price points by focusing on niche markets that are ready to pay for the power they get or are facing issues with their current power system. This way, fuel cells can be sold at today’s commercially viable prices in relatively low volumes and profits can be realized much sooner. An example of such a target application is for industrial utility vehicles that thanks to fuel cells can be rapidly refuelled unlike their lead-acid battery counterpart. Early niche markets could provide the production volumes necessary to turn a profit and achieve further cost reductions, which would then open up new market segments with longer lifetimes, and so on.


While performance is important both for production requirements and as a key for cost reduction, lifetime is also critical. Fuel cells for transportation applications will need to perform 5,000 h of operation under aggressive cyclic conditions. Performance of fuel cells for stationary applications of up to 20,000 h has been demonstrated, but for fuel cells to displace conventional diesel generators, 40,000 h of reliable operation over the full range of external environmental conditions (-35°C to +40°C) are required.

Depending on the technology, fuel cell materials are exposed to harmful conditions during operation: extremes in electrode potential, extremes in acid/basic conditions, extreme temperatures, presence of reactive intermediates, liquid water for certain fuel cell types, etc. Core elements can be progressively fragilised either under constant operation or even more during startup/shutdown procedures that cause sudden voltage/current and internal variations.

For example, durability of the platinum catalyst is a critical problem in PEMFCs, especially at low loadings since any decay of the activity becomes proportionally more influential. The performance of the catalyst layer degrades by platinum sintering and dissolution, especially under conditions of load cycling and at high electrode potentials (on the cathode side). Furthermore, the carbon black supports that provide high electrical contact surface area to the platinum nanoparticles can also be corroded under transient conditions and partly washed up with the two-phase liquid and gas flow.

Demonstration of the lifetime required by each application implies either a deep understanding of all the possible modes of failure (mechanical, chemical, electrochemical), or a very long development and validation cycle. To this end, access to long-term performance data of fuel cells at the same scale and under the various typical operating conditions is of prime importance. A fundamental understanding of what is actually happening in the stack in all possible situations will ensure that developers can rationally trade cost versus performance and lifetime.

Hydrogen availability and distribution

The lack of a hydrogen infrastructure has long been considered the biggest obstacle in particular for the introduction of fuel cell vehicles (FCVs), although this question continues to resemble the classic “chicken-and-egg” problem (there are no FCVs because there are no hydrogen fueling stations, but there are no hydrogen fueling stations because there is no demand for hydrogen as fuel for FCVs, etc.). Establishing the necessary infrastructure for hydrogen production, transport and distribution would require significant capital investment, but there is no absolute impediment, since hundreds of hydrogen refueling stations already exist in the U.S.A., Japan, and Europe.

Obviously, legislation and standards are still missing. In a recent paper by General Motors, it is reported that the key challenge remains matching scale and timing of hydrogen investment with actual hydrogen demand. In the portable sector however, the first regulations are taking place: the U.S. Department of Transportation has published its final rule to allow the transport of fuel cells and a wide range of fuels onboard U.S. passenger aircraft as carry-on baggage in early 2009. The new rule also provides for routine cargo shipment of fuel cells and fuel cartridges by road and rail, as well as international ocean shipment. The previous rule in 2008 had been limited to only liquids and liquefied gases.

It is evident that a key barrier issue for a range of fuel cell applications, not only automotive, is hydrogen availability. A number of other early market applications are for units that run on hydrogen carriers rather than pure hydrogen (e.g. methanol, natural gas). The development of effective fuel reformers systems is therefore as critical as the development of fuel cells themselves.

We should finally add to the list the public acceptance of a daily use of hydrogen.

  1. Applications of fuel cells
  1. Automotive applications (50-250 kW)

This section is limited to the application of fuel cells for light-duty vehicle and bus propulsion. The other related application as an auxiliary power unit onboard the vehicle will be treated in the next section “Niche transport applications”.

    1. Light-duty vehicles (50 kW)

Almost all major car manufacturers have demonstrated prototype fuel cell vehicles and have announced plans for production and commercialization in the near to midterm future (5-10 years). The race to develop a viable fuel cell vehicle and bring it to market began during the 1990s and continues today. The key drivers for the development of automotive fuel cell technology are its efficiency, low or zero emissions, and fuel that could be produced from local sources rather than imported. The main obstacles for fuel cell commercialization in automobiles are the cost of components and the availability of hydrogen.

The only fuel cell technology satisfying to both temperature and time response criteria for vehicle propulsion is the PEMFC. The fast-start capability and low operating temperature combined with good durability and range makes it ideal for use in light-duty vehicle. Power range is about 50-80 kW.

PEMFCs achieve 4,000 h lifetime at the laboratory scale, but the durability target by 2010 for automotive applications is 5,000 h under cycling conditions. The effect of real-life conditions on the fuel cell system (repeated startups and shutdowns, impurities in fuel and air, low and high temperatures) has to be assessed more thoroughly. Air management for fuel cell systems is a challenge because today’s compressor technologies are not suitable for automotive applications. Startup and steady operation in extremely cold climates (-40°C) require specific water management controls, whereas the heat rejection system must be sized for hot weather conditions (+40°C). As operation at low temperatures creates a small differential with the ambient, large heat exchangers and humidifiers are necessary, thereby adding to complexity and reducing the overall system efficiency. The efficiency goal for transportation application is 50% at rated power and 60% at 25% of rated power (peak efficiency).

Four configurations are possible in a Fuel Cell Vehicle (FCV):

The fuel cell is sized to provide all the power needed. Due to the slower response of fuel processors (reformers), this configuration only applies for fast dynamic hydrogen-fed vehicles. A small battery may be present but for startup only.

In the parallel hybrid configuration, the fuel cell is sized to provide the base load, but the peak power for startup and acceleration is provided by a battery. The battery allows rapid startup without preheating of the fuel processor and recapturing of the braking energy, resulting in a more efficient system.

In the serial hybrid configuration, the fuel cell is sized to recharge the battery and the battery drives the electric motor. The relative sizes of the battery and the fuel cell are tied up: a smaller battery will have to be recharged faster and will result in a larger fuel cell.

Fuel cell serves only as an auxiliary power unit, that is, not for propulsion. This configuration is attractive for idling trucks requiring operation of air-conditioning or refrigeration systems.

A hydrogen fuel cell vehicle does not generate any pollution and is qualified as Zero Emission Vehicle (ZEV). If another fuel is used and reformed onboard, the propulsion system has some emissions generated during the reforming process, but those emissions are in general still much lower than the emissions from an internal combustion engine (ICE); therefore the fuel cell vehicles using a fuel reformer are typically qualified as Ultra Low Emission Vehicles (ULEV). Fuel cell-powered vehicles also generate significantly less greenhouses gases than the comparable gasoline-, diesel-, or methanol-powered ICEs.

Hydrogen is the only fuel that results in a zero-emission vehicle, particularly if hydrogen is produced from renewable sources. Use of hydrogen as transportation fuel could reduce dependency on imported oil. A fuel cell system that runs on pure hydrogen is relatively simple, has the best performance, runs more efficiently, and has the longest stack life. Hydrogen is nontoxic and, despite its unfair reputation, has some very safe features.

One of the biggest problems related to hydrogen use in passenger cars is onboard storage. Hydrogen can be stored as compressed gas, as cryogenic liquid, or in metal hydrides. Tanks for gaseous hydrogen are bulky, and the amount to be stored depends on the fuel efficiency and the required range (typically 300 miles or 500 km). In order to achieve a better match between the storage capacity of the tank, the fuel efficiency of the car and its range, further improvements in vehicle design, introduction of new lightweight composite materials, and compression of hydrogen at 700 bar are mandatory.

The difficulty of storing hydrogen onboard a vehicle, as well as the absence of hydrogen infrastructure has forced car manufacturers to consider other more conveniently supplied fuels. In that alternative case, the fuel cell must be integrated with a fuel processor that produces hydrogen from gasoline or methanol. However, in addition to being a nonzero-emission process, onboard reforming is not easy and raises numerous engineering issues:

Onboard reforming reduces the overall efficiency of the propulsion system, which leads to upgrade the fuel cell size;

Onboard reforming enhances complexity, size, weight, and cost of the propulsion system;

Startup time of fuel processors is too long in practice (this issue may be avoided in hybrid configurations);

Durability issues of the PEMFC due to remaining impurities in the reformate hydrogen have been evidenced.

Many car manufacturers are perfecting their proprietary PEM units for use in their vehicles, e.g. Honda with the FCX Clarity, General Motors with fuel cell-powered Chevrolet Volt and Equinox models, and Volkswagen with fuel cell-powered Touran and Tiguan models. While the first one is a purpose-built fuel cell design, all the others are derived from existing ICEs, with mere replacement of the propulsion engine by a fuel cell system. The one amongst the major automakers, BMW is developing an SOFC-based auxiliary power unit for its 7-series luxury car model.

Fuel cell vehicles, because they are still an immature technology and thus are manufactured on a prototype level, are far more expensive than current mass-produced ICEs ($25-35/kW). However, forecast studies conducted by car companies have shown that cost-competitivity could be achieved accounting for mass production manufacturing techniques. The main high-cost components in the fuel cell stack are the catalyst precious metal Pt, or Pt-based alloy, the ionomer membrane, Nafion or fluoropolymer, and the graphite bipolar plates. The cost target for fuel cell vehicles, to be competititve with ICEs, i.e. $35-$50/kW, demands large economies of scale during stack manufacturing and performance improvements in terms of Watts per unit active area. The size and weight of current fuel cell systems including the ancillary components (e.g. compressor, heat exchanger, humidifiers and sensors; cf. Stacks and systems section) that make up the balance-of-plant must be further reduced to meet the packaging requirements for automobiles.

There is a positive message for FCV commercialisation in the next decade, since major manufacturers worldwide are strongly committed to fuel cell technology as a longer term business direction alternative to plug-in battery technology. There is a market for the two of them depending on the application. The current move is from the development of concept cars to first-series manufacturing of fleets of hundreds of FCVs for leasing (e.g. Honda in California started in 2008), and the prospect of thousands of vehicles available to individual consumers after 2012 remains strong. Ultimately, the mainstream move is toward a low or zero carbon economy.

    1. Buses (250 kW)

Buses for city and regional transport are considered the most likely type of vehicles for an early market introduction of the PEMFC technology. Most of the issues discussed in the previous section, Light Duty Vehicles, also apply for the fuel cell applications in buses. The major differences are in power requirements, operating conditions and resulting lifetime, space available for hydrogen storage, and refueling sites.

Buses require typically 250 kW of power under high demanding, intermittent conditions, with frequent starts and stops. Compared to their diesel engine equivalent the efficiency gain is about 15%.

Buses are almost always operated as a fleet and refueled in a central facility. Storage of large quantities of hydrogen onboard (the roof location is very safe for a gas lighter than air) is not a concern. These two characteristics make use of hydrogen much easier.

Thanks to use of hydrogen, fuel cell buses are Zero Emission Vehicles (ZEVs), which is a big advantage over diesel buses in densely populated regions. Demonstration programs funded from local to international level have seen several fleets of fuel cell buses deployed in European cities (Clean Urban Transport for Europe program), in the U.S.A. (Sunline Transit Authority, Connecticut Transit Buses), and in large cities worlwide (United Nations Development Program, Global Environment Facility). Various tests markets are continuing in the U.S.A: as an example, a new fuel cell bus with a 250 mile range has taken service in the Burbank city network in Spring 2009.

The main obstacles for commercialization of fuel cell buses are their cost and durability. Because the production series of buses are smaller than those of passenger cars, their cost per kW is somewhat higher, as is their expected lifetime. Together with the highly demanding intermittent operating regime, this could eventually challenge the current fuel cell technology.

  1. Niche transport applications (1-10 kW)

Small mobile fuel cell systems are designed to produce 1 to 10 kW of electrical power with low to zero emissions. This application is not as demanding as passenger cars or buses. The possible applications are very diverse and include:

Utility and leisure vehicles, material handling industrial vehicles, e.g. forklifts, tow trucks, bicycles, scooters, motorbikes, golf carts; and wheelchairs for mobility assistance;

Aircraft and aerospace applications;

Marine and submarine applications;

Auxiliary power units (APUs) for on-board power supply.

An auxiliary power unit is composed of a small fuel cell system (a stack in the KW power range and a balance of plant part, with or without a reformer), which is associated to a prime driver engine i.e., internal combustion engine or electric motor, in order to supply additional power not related to the propulsion of the vehicle: air-conditioning, multimedia playing or other comfort features. The fuel cell technology allows power generation without engine operation and enhances the run time of batteries. This is especially a good point at the time where anti-idling regulations are setting place in a number of countries. Hence fuel cell-based APUs improve the power flexibility of the vehicle without a complete replacement of the existing technology, which could foster an early market uptake of these “secondary” power sources.

Moreover, the continuous increase in electrical demand for leisure vehicles and equipments is now accompanied by a desire for environmentally friendly onboard conveniences. The growth in this sector is being driven by the search for clean, quiet, efficient power with extended runtime particularly in the high-end side of the campervan and luxury boating markets. In a market somehow protected against recession, consumers are willing to pay a premium for the advantages that fuel cells have over batteries and generators. Campervan manufacturers have understood this very well and are now offering fuel cell-based APUs at least as optional extra and even standard equipment.

Fuel cells for these applications are of the PEMFC or DMFC type, with a small number of SOFC-based units essentially for APU applications. PEMFC-based units are largely dominant in the aerospace, aircraft and materials handling markets while DMFC-based units are more often found for leisure, marine and mobility assistance vehicles. DMFC-based APUs run on methanol without a reformer. SOFC- and PEM-based units usually incorporate a fuel reformer built into the unit so that the system can run on alternatives to hydrogen. In a peculiar market approach, the selection of fuel determines the type of fuel cell stack and the reformer technology. The reason for this fuel diversification is the desire to design the fuel cell APU to run on a fuel that is readily available to the end user. The choice of the same fuel for the fuel cell-based APU as the main engine is a specific requirement of this application. Today this means gasoline or diesel, and development efforts are currently geared towards efficient reformers in order to make this eco-friendly option available for use in commercial trucks shortly.

Like the portable sector, the materials handling sector is one where a real value proposition is now available to warehouse professionals because fuel cell-powered materials handling vehicles operate near silently, with no or few emissions, and offer faster refuelling (1-3 minutes) as well as longer runtime than lead-acid batteries and conventional internal combustions engines. Compared with battery-powered equipment, fuel cell systems have also the advantage of not requiring lengthy and floorspace-hungry recharges. Moreover, capital investment is less since a single fuel cell will operate continuouly, while from a logistic point of view two or even three batteries are needed per battery-powered vehicle.

Fuel cell-based two- and three-wheeled vehicles basically combine clean and efficient indoor operation with lower downtime, rapid refuelling, extended range and no operational degradation over time: since power provided by the fuel cell is constant throughout each shift, there is no performance loss of the vehicle. Altogether, lower maintenance costs are expected from fuel cell-based systems than from their equivalents. This explains why an increasing number of manufacturers are developing and selling fuel cell-powered bikes and three-wheelers. A worlwide potential market exists for example in national postal services that use thousands of bikes and delivery trolleys.

In the materials handling market, fuel cells seem to be still some way off being a serious competitor to long-used ICEs and acid-lead batteries, but some early niche markets are being actively explored and experience is currently gained in warehouse environments before large scale deployments. Ballard, Plug Power, Nuvera Fuel Cells and Hydrogenics are main players in the materials handling sector who are currently testing products on-site.

Finally, mobility assistance vehicles are an interesting niche market for light fuel cell systems, which offer an extended range but none of the inconveniences of battery recharging. SFC Smart Fuel Cells from Germany and Ajusa from Spain have shipped tens of units for impaired customers in 2008.

In the aerospace and aircraft sectors, successful test flights have been reported in recent years, as well as a continued development of fuel cells for auxiliary power units on board larger aircrafts. Due to their silentness and long runtime, unmanned aerial vehicles (UAVs) are especially attractive in the defense and aerospace fields for handling military reconnaissance, surveillance missions, or remote communications in strict secrecy. Other civilian applications are studied like remote scientific data collection under harsh conditions, disaster relief missions… It is unlikely that fuel cells will be used as a primary source of power for commercial aircrafts any time soon, but demonstration is being made that they are able to operate under extreme conditions: low temperatures and pressures, and unusual spatial orientations; hence they could provide efficient energy for onboard electrical systems in-flight or under ground operation: heating/cooling, entertainment devices, and even essential controls in the aircraft, thereby reducing fuel consumption. Here a fuel cell APU may offer better efficiency than turbine power units used today in spite of the necessary kerosene reformer. Furthermore, in-flight production of water is under investigation by several aircraft companies, e.g. Airbus.

In the marine sector, legislation is likely to act as a key driver for the adoption of fuel cells. New restricting policies requiring low or zero emission for vessels in certain rivers, lakes and inland waterways in China and Europe, as well as the growing pressure on regulating pollutant emissions in harbours, in coastal waters and on the high sea, are a favorable ground for the uptake of fuel cells as APUs onboard vessels to reduce overall emissions and also for development of fuel cells as main means of propulsion. This has already caused a doubling of unit shipments in 2008 (mainly in Europe); the trend will supposedly accelerate in the forthcoming years. Proton Motor’s Zero EMission Ship is the first boat propulsed by a fuel cell: it has carried passengers as demonstration project since July 2008 in Hamburg’s harbour. Other proof-of-concept projects of fuel cell-powered or battery hybrid systems are under development for integration in canal barges, tug or river boats. Silent operation is of utmost importance for certain applications like scientific studies of sea animals. National governements and the International Maritime Organisation are in the process of voting further reductions of pollutant and noise emissions: new laws will certainly follow. Clearly, there is a great opportunity for fuel cells at time of regulations in the marine environment.

On the naval undersea side, a number of programs are ongoing to develop fuel cell systems for ships and underwater applications. Funded at their most part by the national navies, they are split into primary propulsion for Unmanned Underwater Vehicles (UUVs) and onboard power generation for larger vessels. Like the commercial sector, the key issue is to develop fuel reforming technologies that will allow fuel cells to run on a specific range of fuels including marine diesel. Powering an UUV by a fuel cell has many advantages in terms of silent and autonomous operation, quick refueling and steady energy output.

Europe is leading the development of fuel cell APU products for recreational vehicles. There are two companies having commercial products available for this niche market, SFC Smart Fuel Cell in Germany and Voller Energy in UK. Unlike most fuel cell manufacturers, who are in still in the research and development phase or running demonstration projects, SFC has shipped in October 2008 a total of 10,000 EFOY (Energy FOr You) DMFC-based systems to industrial and private end users, and created its own fuel cartridge supply infrastructure. Delphi and Cummins Power Generation are working on projects funded by the U.S. Solid State Energy Conversion Alliance, to demonstrate SOFC technology on commercial vehicles. The EU is also funding a project to validate renewable methanol as fuel for SOFC-based commercial vessels and quantify its environmental impact in comparison with conventional systems.

The leisure sector will probably remain a leading application for fuel cell-based APUs, but improvements in battery technology could challenge the market growth. This requires continued investment in better products. To consolidate success as an early market for fuel cells, APU systems with output powers in the 1-10 kW range need to be realized at a level of cost, size and durability suitable for commercial use, and sulfur-tolerant reformers that are compatible with future fuel specifications must be developed and produced.

  1. Portable applications (0.1-100 W)

In the portable sector, industrial interest for fuel cells in the W power range is great because of recurrent issues inherent to battery technologies (Nickel-Metal Hydrure, Lithium-Ion or Lithium-Polymer). Significative improvements are possibly brought by fuel cells in this field:

Fuel cells have a higher energy density than batteries, i.e. they provide more energy per unit of weight, up to 5 times more. This allows longer run time before refueling.

Portable fuel cell systems including the fuel storage container can be designed smaller and lighter than a battery of equivalent power.

Continuous operation of fuel cells (as long as fuel is supplied) means also longer runtime (depending on the fuel reservoir volume), no time-consuming recharge and associated logistics (e.g., need for several units for battery exchange), and less degradation of the components.

Micro-power applications of fuel cells are typically the same as batteries, i.e. all electronic devices for nomad use like mobile phones, laptop computers, personal digital assistants, cameras, and music or multimedia players. Other applications are found in portable military, healthcare or in the recreational market (camping tools, etc.). In these applications fuel cells are expected to replace batteries thanks mainly to their higher storage capacity. To compete with batteries, fuel cell systems for consumer electronics need to have improved energy density by more than a factor of three. Conversion efficiencies are of less importance as long as they do not limit autonomy.

Micro fuel cells are mostly of the DMFC or the PEMFC type; they run at low temperatures on liquid methanol, formic acid, or hydrogen stored in low pressure hydride containers. The operation temperature should not exceed 50 to 60°C, which excludes the use of a reformer. Further, it is important that the fuel storage achieves a high level of security. Use of liquid high-pressure hydrogen is excluded as well. In this application, PEMFCs with chemical hydrogen storage are competing with DMFCs. While PEMFCs have a higher power density than DMFCs, chemical hydride solutions are not ready for market yet. Portable DMFC-based micro fuel cells have been demonstrated by Toshiba, Smart Fuel Cell and MTI Micro Fuel Cells. A few companies are involved in micro SOFCs: Kansai Electric Power utilizes for example nanotechnology in its low temperature SOFC devices, and Lilliputian Systems, a spin-out from the U.S. M.I.T., works on a micro SOFC and a micro reformer based on semiconductor technology.

An increasing number of companies are also developing fuel cell cartridges, either as stand-alone products (BIC, Gillette; Neah Power) or for powering their own consumer electronics portfolio (Motorola, Hitachi, Panasonic, Sony). Motorola expects its fuel cells to run about 10 times longer than today’s batteries before needing to be recharged.

The big challenge is the miniaturization of the system, which may consist of scaling down the different components of larger existing stacks or developing a specific architecture based for instance on silicone chip-supported thin films derived from microelectronics. Each of these solutions implies a specific management of the various fluxes in the fuel cell: flux of reactant gases, flux of products water and heat. The crucial point for the micro fuel cell is to handle the power surge upon switching the device from idle to active.

The current trend for portable devices is an ever growing power demand in conjunction with their increasing number of Internet functionalities. Therefore, the advantages of fuel cells of storing more power in the same volume, for longer time while being able to refuel the product quickly instead of recharging hopefully will be seen by most consumers as a new area for freedom, and lead to a decay of the battery marketshare. Furthermore, standard batteries on use today are also rather expensive, so that the barrier cost for the introduction of fuel cells to the portable market is lower in this application than in others.

  1. Stationary applications (1 kW-5 MW)

Among fuel cell applications, stationary applications are the most diversified category. This is due to a wide power range from one kW to multi MW, and many possible end users (civilian/military, industrial/utility services, private individuals…) with different objectives, specifications, and budgets. Requirements on size and weight are less critical and modularity is definitely an advantage: several types of fuel cells are eligible and a large spectrum of fuels is available.

Performance of fuel cells for stationary applications for up to 20,000 h has already been demonstrated but market acceptance will likely necessitate more than 40,000 h durability over the full range of external environmental conditions (-35°C to +40°C).

PAFCs were the first fuel cells to be tested on-site and 200 kW modules have been commercialized by UTC Fuel Cells in Japan, USA and Europe since the 1990s. Developments are limited today; nevertheless, these fuel cells are the only ones to be really competitive on the market, and have allowed gaining experience in the integration of cogeneration systems, increasing reliability, and improving management controls. Other types of fuel cells are now under late stage development in the stationary sector:

High temperature fuel cells SOFCs and MCFCs for applications from residential to industry, and from cogeneration to centralized electricity production;

Low temperature fuel cells PEMFCs at 70-80°C essentially for domestic and off-grid applications.

Every type of fuel cell has its own advantages and drawbacks, and there is room for marketplace share, depending on the specific needs of end-users and the degree of development of each technology. For example, PEMFC appears more ready for market than SOFC but in the longer term, could be supplanted for certain applications. Another big difference is the possible use of various fuels and of internal reforming in high temperature systems, whereas for low temperature fuel cells, the choice of the fuel is much more restricted and external reforming in a separate reactor is the only option. In the short term, natural gas/propane will be preferred due to an existing infrastructure and a positive feedback from people.

In most stationary applications, heat in addition to electricity is demanded. The temperature at which heat is needed will depend on the application: for residential heating purposes, temperatures below 80-90°C are more than enough and PEMFCs can fit (yet sometimes hardly), while for industrial applications where heat is usually recovered to process steam temperatures well above 200°C are required, which is a future market for SOFCs and MCFCs. For PEMFC, new technologies need to be developed that will allow higher operating temperatures and/or more effective heat recovery systems.

The small stationary fuel cell market (<10 kW) is shared between Uninterruptible Power Supply (UPS) and Combined Heat and Power (CHP) units: UPS systems provide only electrical power with more than 40% efficiency whereas CHP systems based on cogeneration provide power and heat with up to 85% conversion efficiency. PEMFC technology represents today 95% of the 4000 units shipped in 2008 with SOFC taking under 5% and AFC less than 1%. Derived from the CHP, the Combined Cooling and Power (CCP) unit integrates an absorption chiller besides the heat-recovery reactor. The low operating temperature of PEMFC limits the amount of waste heat that can be effectively used in CHP applications. Startup times need to be decreased in stationary backup power systems running on direct hydrogen.

A set of fuel cell manufacturers working on localized stationary power generation will include for instance Idatech, Hydrogenics, CellFraft, Nuvera Fuel Cells and Plug Power with UPS and CHP units on the PEMFC technology side; Ceres Power and Ceramic Fuel Cells with CHP units on the SOFC technology side. In 2008, deliveries of fuel cells for small stationary applications have reached 4000 units. The majority of the systems were sold for UPS applications, with a continuous shifting to commercial deployment, but there was progress in the use of small units for micro CHPs too: the residential sector has seen the launch of two major European demonstration projects in Germany and Denmark, and further advances in Japan’s residential programme.

The large stationary fuel cell market (>10 kW) is represented by units operating either tied to the grid or off-grid as CHPs, CCPs or electricity generators. Over the last five years MCFC and PAFC have become commercial and represented respectively 40 and 35% of the total units shipped in 2008 while SOFC technology is starting to take off after intense R&D efforts, with a percentage of 15%. PEMFC technology seems on decay with less than 10%. The current trend observed for commercial shipments is a general increase of the average unit size to the MW level and above and the continued development of key markets such as California and Connecticut in the U.S.A. where two-thirds of the (few) large stack manufacturers are located.

An important distinction must be made between localized production of small power with (residential applications) or without cogeneration (backup applications), cogeneration of medium power (remote, institutional, and backup applications), and centralized production of electricity without heat recovery. The first category is referred to as “localized stationary power” and the second one to “distributed generation”.

    1. Localized stationary power

CHP fuel cells are currently developed in sizes appropriate for use in single or multi-family residential applications (3-50 kW) in order to provide clean, quiet and efficient primary or backup power. They can either operate in parallel to the electric grid or off-grid in case of power outage. Like a boiler, they can be installed in the home basement and thanks to the reformer unit; they are able to extract hydrogen from traditional fuel sources such as natural gas and propane. The integration of a fuel cell for home power more than doubles the amount of low-impact electricity that is delivered to the grid, and the non-combustion heat eliminates the need for a boiler. Residential fuel cell cogeneration systems reduce CO2 emissions by up to 40% compared to conventional energy generation and hot water systems.

Small stationary CHP fuel cell power plants (0.1-5 MW) are a good alternative in such remote locations where grid lines are expensive and/or difficult to install: islands, mountainous areas, sparsely populated regions, etc. They offer a competitive energy solution to many communities that are currently not connected to the electric grid without the need to build a heavy infrastructure. The low maintenance and fuel-transport associated costs, and a very limited environmental footprint make fuel cell remote power plants close-to-ideal for this application. Use of propane fuel is strongly considered as an early-adopter target market for rural residential areas and small businesses in remote sites. Alternatively, hydrogen could be produced by renewable energy sources (wind, hydro or solar) and used for local transport. Clearly, there is an avenue for market growth of decentralized power by fuel cells in an increasingly sustainable world.

As compared to lead-acid battery and diesel-powered generators traditionally used by communication centers, fuel cell-powered UPS (0.5-200 kW) offer a greater reliability and more predictable performance over a wide range of operating conditions including harsher (colder/warmer) climates. They are well suited for powering networks for extended periods of time in data centers, banks, and other sensible government or commercial buildings where power interruption must be absolutely avoided. Backup power systems are being deployed across government agencies to demonstrate the increased security brought by fuel cells.

Larger stationary CHP fuel cell units (250-400 kW) can also be installed on the premises of an institutional building, e.g. school, hospital, industrial facility, and provide heat and primary power in addition to backup power. With a combined efficiency of 70-80%, cogeneration systems reduce primary energy consumption by 20 to 30%. Energy costs would be reduced especially during periods of peak demand. Moreover, along with high reliability, the voltage output delivered by fuel cells is consistently steadier than that coming from the electric grid. Voltage fluctuations, which are highly deleterious to computer systems, would be less than an issue with fuel cell powering.

Let us finally mention a new potential market for stationary fuel cells: CHP fuel cell power plants (>1 MW) could be used as CCPs in data centers and server farms, providing both power and air-conditioning to the exploding number of servers housed together for electronic data storage (over one million in Google’s server farms, for example).

As the price of residential fuel cell units today is largely prohibitive for individual consumers without proper incentives, a variety of ownership and leasing options is being progressively made available to them. First adopters will also be able to sell excess power produced by the fuel cell unit back to the electric utility at an advantageous fare and get progressive payback of their investment. This is how manufacturers expect to make the shift from technology to market in the residential sector.

Besides commercially-driven developments, the market growth of this sector is induced by a strong demand pull from governments, e.g. in Japan with the Japanese Large Residential Fuel Cell Programme, or in Germany with the Hydrogen and Fuel Cell Technology Innovation Programme including a 2020 vision to produce 72,000 residential units per year. In the U.S.A after the hurricane Katrina, a special panel order has stated in 2007 that backup power units with an extended runtime ought to be implemented in the 30,000 telecommunication base stations registered in the whole country.

Thanks to their modularity, long operation time and low maintenance costs, this is a big potential market for fuel cell systems especially in remote regions where gasoline supply chains for conventional diesel generators are at higher risk to be disrupted anytime. Over ten major telecom companies have tested fuel cell-powered UPS systems worldwide: the feedback is generally positive. This has given rise to a number of distribution deals between stack manufacturers and integrators, among which the big commercial agreement in late 2008 between Idatech and Acme Telepower for telecom backup and uninterruptible power in India. Recent national and city legislations on “greening” of new buildings, like the London Plan to cut carbon emissions by 20% and the Executive Order 111 in New York City to produce 20% of the electricity in state office buidings by renewable energies by 2010, as well as an increased awareness of the benefits of decentralised power generation also push the technology forward.

    1. Distributed generation

Fuel cell power plants in the MW range are well suited for the distributed generation of electricity at locations near demand thanks to their modularity and quietness under operation. The energy produced is of high quality, and there are fewer transport losses. This means fewer distribution lines and low environmental impact. Because the installations are smaller than typical central generation power plants, they are easier to site, permit, and finance.

Among the different types of fuel cells, only high-temperature technologies, SOFCs and MCFCs, can be used for distributed generation in conjunction with a steam and/or gas turbine to produce high efficiency electricity (> 70%) compared to fossil fuel power plants of similar size that achieve only 30 to 40% efficiency. They deliver a cleaner power for each unit of fuel used, thereby reducing power costs and CO2 emissions substantially. Like cogeneration systems, fuel cells can operate off-grid for decentralized power generation.

Ongoing developments are focused toward larger units. In the near future, however, fuel cell power plants are not intended to replace conventional power plants for primary power generation, but will more likely be used in parallel with the electric grid to increase reliability by getting rid of power blackouts during outages. Utility companies may also utilize them as additional power to relieve grid congestion during peak hours, thereby reducing overall energy costs for end-users. Finally, they could reduce the need for new central power generation, transmission and distribution, and its related high investment costs.

Recent market developments have seen an increased interest on distributed generation by legislators and business planners alike with an increased number of units being sold into office blocks and schools. The most active country in the sector has been the U.S.A since the beginning of the 2000s. Through its Self Generation Initiative Programme (SPIG), California has been subsidying the installation of fuel cells and other renewable technologies producing up to 5 MW of power, whose direct effect was the fastest growth for large stationary fuel cells in the world. Connecticut has also issued a Renewable Portfolio and is developing fuel cell-based distributed generation with FuelCell Energy, the world leading company in manufacturing of stationary fuel cell power plants.

  1. Members of the fuel cell family

All the devices that we call fuel cells can be included in a single family of technologies: each one is characterized by the type of electrolyte used and thus, by the operating temperature allowing proper proton transport. They all rely on the direct electrochemical conversion of the chemical energy contained in the fuel into electrical energy without an intermediate heat cycle. Even though the electrode half-cell reactions may differ from one type to another due to a different fuel or “hydrogen carrier”, the overall reaction is unchanged:

2H2 + O2 ® 2H2O

Plus, the basic configuration of a single fuel cell is always composed of an ionic conductor separating two electronic conductors, whatever the specific materials constituting these different parts and the exact running conditions. The fuel is always oxidized at the anode and simultaneously, the oxidant is always reduced at the cathode leading to the formation of water and heat side-product without other emissions (when hydrogen is the fuel source). The operating temperature is determined by the temperature range at which the conductivity of the electrolyte used is sufficient for the transport of protons without losses, and its mechanical resistance is optimum. Low temperature fuel cells typically operate below 200°C whereas high temperature fuel cells operate above 600°C. In the intermediate temperature range, no fuel cells systems do exist yet due to a lack of suitable electrolytes.

Let us now examine the different members of the family with their main characteristics, advantages and weaknesses, fields of applications. They have been classified from low- to high-temperature fuel cells.

1. The Proton Exchange Membrane Fuel Cell

a. How does a PEMFC work?

Proton Exchange Membrane Fuel Cells (PEMFCs) or Polymer Electrolyte Fuel Cells (PEFCs) are low-temperature fuel cells that can operate between 50 and 90°C. They are currently characterised by the use of a thin, proton-conducting polymer membrane as the electrolyte and of platinum supported on high surface-area carbon black particles as the electrodes. Within the PEM fuel cell membrane-electrode assembly, hydrogen molecules (from either pure or reformate hydrogen source) are supplied at the anode side through flow channels and split into hydrogen protons and electrons (1). Once formed, protons are hydrated by a water molecule and transported in the form of hydronium ions H3O+ from the anode to the cathode across the membrane while the electrons are pushed to an external circuit through the gas diffusion layer and bipolar plates in order to produce electricity. Oxygen (mostly from air) supplied at the cathode side through flow channels, then reacts with the protons on the catalyst to produce water (2). Overall, the reaction is simply the electrochemical combination of oxygen and hydrogen to form water molecules (3).

Anode 2H2 ? 4H+ + 4e (1)

Cathode O2 + 4H+ + 4e ? 2H2O (2)

Overall O2 + 2H2 ? 2H2O (3)

Both anodic and cathodic reactions can be catalyzed by platinum. While hydrogen oxidation reaction (HOR) over Pt is intrinsically very fast, oxidation reduction reaction (ORR) over Pt is very slow. The voltage loss at the anode under practical current densities such as 0.4 A/cm² is about 10 mV and at the cathode is over 400 mV. Due to these kinetic limitations, energy lost is released as waste heat and the reaction is exothermic: cogenerated heat can be used outside the fuel cell in combined heat and power (CHP) units.

Perfluorosulfonic ionomers like the Nafion® membrane developed by DuPont since the 1970s are the standard membranes in PEMFCs due to their high proton conductivity below 100°C and remarkable mechanical and electrochemical stability under fuel cell conditions. The molecular structure of Nafion® consists of polytetrafluoroethylene (PTFE) –(CF2)n– as the backbone and perfluorosulfonic acid –(CF2)m-SO3H as pendant side chains. Similar membranes are produced by W.C. Gore (Primea®, Gore-Select®) and Asahi (Aciplex®, Flemion®). Thickness has been progressively reduced from 175 µm (Nafion® 115) to 25 µm (Nafion® 111) and even down to 18 µm for PTFE-reinforced membranes (Gore-Select®) in order to improve water transport and as a consequence, fuel cell performance.

The coexistence of a fluorocarbon backbone and sulfonate sites provides a unique structure when the polymer is hydrated: the fluorocarbon is hydrophobic while fluorosulfonic acids are typically superacids, and thus highly ionic and hydrophilic. In this type of electrolyte, proton conduction is made possible by the presence of water inserted in the hydrophilic regions: protons can cross the membrane via these microphase pathways. This means that the membrane must be properly humidified to keep the narrow (» 1 nm) hydrophilic channels between sulfonate sites open and interconnected. The advantage of protonic conduction by water is the high conductivity obtained at low temperature. The drawback lies in the difficulty of maintaining an adequate water level inside the membrane any time. The control of system humidification near saturation point, also called “water management”, is an important consideration in PEMFC technology.

Oxygen reduction catalysts used to speed up the cathode reaction have evolved from platinum black to carbon black-supported platinum in current state-of-the-art PEMFCs. At the anode carbon-supported Pt or Pt-based alloys can be employed depending on whether pure hydrogen or reformed gas is used as the fuel. Pt/C catalysts have higher active surface area than Pt black and are available in loadings from 10% to over 50% of platinum. After the research work at the Los Alamos National Laboratory in the early 1990s, platinum loading has been decreased by more than one order of magnitude (from >> 1 mg cm-2 to 0.2-0.5 mg cm-2), with no detrimental effect on performance thanks to significant improvements in catalyst microstructure. Higher electrochemical activities are achieved by use of finely dispersed nanometric platinum particles (2-4 nm) and the introduction of about 30 wt% ionomer as binder agent in the catalyst “ink”: this allows the formation of an extended “three-phase contact zone” in the catalyst microporous layer between the protons (in the ionomer phase), the electrons (in the Pt/C phase) and the gas phase (in the pores). By this way, the catalytic reactivity is enhanced.

The catalyst layer (thickness » 15 µm) is applied to the membrane directly using various printing processes or by hot-pressing of gas diffusion electrodes formed by applying the catalyst layer to the gas diffusion layer e.g. Toray paper or SGL Sigracet® materials (thickness » 200-300 µm). The typical diffusion media in PEMFCs are carbon fiber-based products, such as non-woven papers and woven fabrics (or cloths) due to their high porosity (> 70%), good electronic and heat conductivity, and high corrosion resistance. The role of the gas diffusion medium is to transport the various species (reactants and products) with minimum voltage loss, from the channel/land structure of the flowfield to the active area of the electrode. Thus-prepared membrane-electrode assemblies (MEAs) or cells are the core of any PEMFC.

In the planar configuration, the bipolar plates perform as current collectors between cells, provide channels for reactant and coolant gas flows, and constitute the mechanical support of the stack. In current PEMFC state-of-the-art, they are mostly made of graphite/polymer composites due to graphite’s high corrosion resistance and low surface contact resistivity. However, the brittleness, gas permeability and unacceptable weight/volume ratio, as well as high manufacturing cost of graphite-based composites have spured extensive studies on metallic bipolar plates such as aluminium, stainless steel, nickel and titanium in order to replace them, especially for mobile applications where shock resistance and low weight are highly desirable. Metals exhibit higher mechanical strength, better durability to shocks and vibrations, no permeability, and much superior manufacturability when compared to carbon-based materials. Current research efforts are focused on finding a corrosion-resistant surface treatment to protect the metal surface in the harsh acidic and humid PEMFC environment, and minimize the contact surface resistance between the different components. Carbon-based including graphite or diamond-like carbon, conductive and organic self-assembled polymers, and metal-based coatings including noble metals, metal nitrides and metal carbides are also being investigated.

With the current technology, power densities of 0.5 W/cm² and more are obtained at 0.7 V for a single cell. The net efficiency of a PEMFC unit usually lies between 40 and 60%. It is essentially limited by the slow ORR kinetics at the low temperatures of PEMFCs, even on platinum catalyst, and by the resulting cell voltage decay. The inefficiency of the oxygen reduction reaction also leads to the parasitic formation of hydrogen peroxide by a side two-step reduction pathway (4,5):

Cathode O2 + 2H+ + 2e ? H2O2 (4)

and H2O2 + 2H+ + 2e ? 2H2O (5)

If reaction (5) is significantly delayed, hydrogen peroxide, which is a strong oxidant, will cause irreversible chemical degradation of the membrane by inducing radical attacks. This process has been demonstrated during PEMFC tests, in particular under transient voltage and cycling conditions. Besides the identification of potentially harmful operation conditions, a more active cathode catalyst would help reduce generation of peroxide. Therefore, research focuses on seeking for better catalysts and higher temperature systems.

Today, PEMFCs show best performance if run at temperatures of 70-80°C and fed with humidified hydrogen saturated at the stack temperature. The useful operating temperature is limited first by the glass transition of the solid electrolyte that ranges between 110°C and 140°C for perfluorosulfonic membranes. Above this threshold, their structure is weakened and the membranes become irreversibly damaged. Secondly, continuous operation above 80°C at ambient pressure would cause excessive evaporation of water and drying out of the membrane with an important loss of conductivity. Likewise, operating temperatures below 0°C are problematic due to the formation of water ice in every part of the fuel cell and consecutive degradation of the components. Protonic conduction by water molecules basically limits the temperature range of PEMFCs to that of liquid water.

These technical limitations and the high cost of Nafion® have generated extensive research into alternative membranes including other perfluorosulfonic membranes and fully organic polymers such as styrene-based systems, poly(arylene ether)s, polyimides, polyphosphazenes, etc. Challenges include stability at 120°C and ionic conductivity with minimal water uptake. Alternative ionic functional groups have also been investigated but the sulfonic acid groups exhibit the best stability and acidity, i.e. proton conductivity. Therefore, the vast majority of approaches still rely on the sulfonic acid based conduction mechanism. Difficulties generally arise due to the fact that hydrocarbon linkages to sulfonic acid groups are more prone to oxidation than fluorocarbon bonds. Dimensional stability under various water levels is also a big concern when considering stack internal compression management. Unfortunately, excessive swelling of carbon-based ionomers is often observed. Given the high stability of fluorocarbons, an organic polymer or composite membrane that would outperform them at more commercially appropriate high-temperature and low-humidity conditions is still looked for. No major breakthrough has occurred in this field at present.

Although the simple and most efficients systems obviously run on hydrogen, the still lacking hydrogen infrastructure (cf. Issues section) forces the use of other more readily available fuels also in PEMFCs both for stationary and even mobile (automotive) applications. The problem is that lots of impurities from the reforming process such as CO, H2S, NH3, NOx, SOx, etc. are brought along with the hydrogen feed stream into the fuel cell stack and cause performance degradation by poisoning the electrodes, especially on the anode side. To allow for a stable operation with reformed fuels, both an appropriate gas cleanup and improved anode catalysts are necessary. Use of binary or ternary Pt-based alloys is under investigation in order to improve the carbon monoxide tolerance of the anode catalysts. The highest performance achieved to date is by platinum-ruthenium alloys with 1:1 atomic ratio. Since platinum is less sensitive to CO contamination as temperature is increased, use of high-temperature PEMFCs is also recommended.

b. A short history of PEMFC

The development of proton exchange membrane fuel cells is stongly related to the history of the membranes. Because fuel cells of various types were known prior to PEM technology, the catalysts, fuels, and oxidants were reasonably well established. The first system was invented by General Electric in the early 1960s as part of a research programme with the U.S. Navy and Army Signal Corps. The PEMFC unit was fueled by hydrogen generated by mixing water and lithium hydride contained in disposable canisters. It was compact but very expensive.

A breakthrough occurred during the Project Gemini of NASA’s space program: General Electric was mandated with the objective to replace batteries by a PEMFC in order to provide electrical power and drinking water to the spacecraft. At the time, the cells were short-lived because of the oxidative degradation of the membrane, which was a copolymer of sulfonated polystyrene and divinylbenzene. After facing numerous technical issues due to cell contamination and leakage of oxygen through the membrane, General Electric replaced the aromatic membrane by the fluorocarbon-based ionomer Nafion®, newly commercialized by DuPont for water electrolysis, and succeeded in improving the operational durability of its power unit. The new version was eventually used for a significant number of flights until the end of the Gemini programme.

Despite the selection of more robust AFC systems by NASA for the next Space Shuttle programme during the 1970s, General Electric continued working on PEMFCs, developed a PEM-based water electrolysis system for undersea life support, and set in collaboration with the U.S. Navy for its Oxygen Generating Plant. The British Navy adopted it in the 1980s for its submarine fleet and considering this achievement, a number of groups started looking at the PEMFC technology for development and commercial applications. In the late 1980s and early 1990s, the Los Alamos National Laboratory and Texas A&M University explored new ways to reduce the amount of platinum in the electrodes, which led to a new (and lower cost) MEA architecture.

In the early 1990s, the Canadian company Ballard developed its 5 kW hydrogen/air fed Mark V fuel cell stack based on Nafion® and low platinum catalyst. In the following years, power output was increased by Ballard and DaimlerChrysler up to 75 kW (Mark 900). This power achievement was an important milestone as it demonstrated that the PEMFC technology was actually able to meet the performance targets for transportation applications. Until today, it remains the technological choice of the automotive industry.

In the 1990s, strong political and social forces further impacted the development of PEMFCs. In some U.S. Eastern states and in California, regulations were adopted that mandated auto manufacturers to provide low to zero emission vehicles. Following this, high commitment of the automotive industry has been observed and demonstration programmes are putting fuel cell vehicles (FCVs) on the roads (cf. Automotive Applications section). This boost of human and financial investment has been beneficial to the development of other PEMFC applications, and not only transportation. This has made them of more than academic interest, and has provided financial resources for more extended testing and exploration of new markets. Consequently, the use of PEMFC units for portable and small stationary applications has rapidly increased over the last few years as these sectors have begun to expand and are currently enjoying a strong period of growth (cf. Small Stationary Application and PEMFC applications and perspectives sections). This has also helped contribute to development programmes as market pull starts to match technology push.

c. PEMFC applications and perspectives

Since its first commercial development in the 1960s, the PEMFC technology has known a continuous evolution. Due to a short startup time, reduced material strain, and lower requirements on thermal insulation and security than high-temperature fuel cells, PEMFCs have found potential markets mostly in transportation including niche applications; light duty vehicles and buses, small and large stationary applications and portable sources (cf. Application of fuel cells section).

Today, about two-thirds of the installed PEMFC units are being used for portable applications, because this market is the most developed and the closest to reaching commercialization of all the sectors. In the portable sector, PEMFC technology is considered the most prevalent alternative to DMFC technology. Moreover, the interest for portable applications is likely to grow in the near future as hydrogen storage solutions will become commercially available. The second largest sector (» 15%) is for small stationary units, and altogether niche transportation accounts for 15% of the total. The outlook for PEMFCs in stationary applications is also promising. Off-grid power generation, which promotes this technology, is currently enjoying a reasonable growth.

Governments are supporting the development of the PEMFC technology worldwide:

In the U.S.A., the Energy Efficiency and Renewable Energy (EERE) Department has elaborated a plan to reduce costs to $45/kW by 2010 ($30/kW by 2015), and to increase efficiency for transportation and stationary applications: distributed generation units fuelled by natural gas/propane are fixed 40% electrical efficiency and durability of 40,000 h. The U.S. Department of Energy (DoE) is also leading a consortium involving industrial and academic research in the fuel cell and hybrid vehicle field for longer term operating, cost, weight, and efficiency targets.

The European Union has also set targets with specific reference to PEMFC systems. The goals include the development of high-temperature PEMFCs, low cost stacks and manufacturing processes, improvement of performance, and the first commercialisation of combined heat and power (CHP) units. Within the current 7e Framework Programme of the European Commission (2007-2013), a Joint Technology Initiative (JTI) on hydrogen and fuel cells has become operational in 2008. The JTI is an industry led public-private partnership, which coordinates the implementation of industrial research, demonstration projects at an appropriate scale to validate research results and provide feedback. It also leads cross-cutting and socio-economic activities including infrastructure issues in order to get a rational basis for future policies and support the future broad market introduction of fuel cell and hydrogen technologies. The JTI is updated on an annual basis.

In Japan, the government has significantly invested in a large residential fuel cell programme (predominantly PEMFC-based, with Ebara Ballard as the fuel cell stack manufacturer). A joint-agreement has been signed between Nippon Oil and 5 other Japanese firms at the beginning of 2009 in order to create a nationwide dealership network and launch the commercialization of the Ene Farm home-use fuel cell system to individual customers. Other emerging countries, e.g. India, may undertake similar subsidy programmes thereby boosting the number of installed PEMFC units worldwide.

However, whilst the automotive sector provides the most significant potential for PEMFC systems as the unchallenged technology for powertrains in future fuel cell vehicles (FCVs), it will likely be the last one to reach full commercialization, and obviously not in the short term. Development and demonstration programmes with car and bus fleets are ongoing but the actual number of FCVs on the roads is still quite low. Many of the automakers, which manufacture their own stacks, are likely to be producing PEM units. For example, Honda was originally a licensee of Ballard stack technology, but like Toyota and General Motors eventually developed its own stacks for the FCX Clarity. Other fuel cell vehicles for series production include the Chevrolet Equinox by General Motors, the BlueZERO by Mercedes. In the timetables, 2015 today sounds a realistic date for FCVs, still with caution yet!

The future of PEMFCs is really promising as the technology can be adapted to several key markets spanning growth opportunities over the short (portable), medium (stationary), and long term (transportation) timeframe. However, as with any new and changing marketplace, several challenges remain with PEMFCs due to the high cost of some of the materials used in the production of PEMFCs.

Current research shows that the main causes of short life and performance degradation are poor water management, fuel and oxidant starvation, corrosion and chemical reaction affecting cell components. Poor water management can cause dehydration and/or flooding. Operation under dehydrated conditions could damage the membrane, whereas flooding favors corrosion of the electrodes, the gas diffusion media and the membrane. Corrosion products and impurities from outside (e.g. chloride, copper, iron, and sulfur compounds) can poison the cell or accelerate the rate of peroxide generation. Thermal management is particularly important when the fuel cell is operated at sub-zero and elevated temperatures and is key at cold start-up and when subjected to freezing conditions (cf. How does a PEMFC work? section).

Further developments in catalyst design are also critical to the commercialization of PEMFCs. However, challenges remain to make them competitive with traditional power sources. Cathode losses with current materials are too high to meet the targets of the U.S. Department of Energy. Insufficient stability due to platinum dissolution and carbon corrosion is of particular concern under dynamic operating conditions and must be addressed.

The current state of the technology is such that Nafion is an acceptable material for steady state operation and low duty cycle applications. For more demanding systems such as automotive applications, important membrane and electrode issues do exist. Ultimately, given current catalysts and stack designs, Nafion® is not sufficiently stable to run PEMFCs under dynamic load conditions for long periods. Although its price should become affordable around 1 million of vehicles/year, there are more and more reasons to work further on potentially low-cost and more performing alternative membranes.

Possible strategies to improve PEMFC performance and durability include system and MEA design and engineering:

Work on PEMFC systems is aimed at minimizing the parasitic power loss while maintaining stability, especially under load cycling conditions. The water management due to high membrane humidification but no-flooding requirement is a central issue and significantly adds to the system complexity, weight, size and cost. The overall goal is towards simplification, cost decrease and high reliability.

Work on MEA components involves modifying the material and structural properties of the gas diffusion layer, cathode catalyst layer and membrane.

A combination of better cathode catalyst and better membranes may be effective. Approaches are focused on the following areas:

(i) Development of high-temperature fuel cells (120°C) that can avoid water flooding, due to the absence of liquid water at operational temperatures above 100°C;

(ii) New materials development, such as a thinner but stronger membrane that not only facilitates easy water removal through the membrane but also improves the reliability of the thin membrane; introduction of alternative alloy catalysts.

(iii) Innovative catalyst layer design: optimisation of the structure (dispersion, porosity, hydrophobicity) for increased activity and better gas transport, decreased thickness for easier water removal.

New materials that operate at higher temperature with improved stability would be a great value but are challenging to design. Vehicle system analysis indicates that 120°C high-power generation would enable the use of radiators similar to those available today in internal combustion engines. This has led to projects focused on high-temperature (120°C) and low-RH (25%) membranes. In the absence of high-temperature membranes, operation is limited to 80°C and results in heat rejection systems for FCVs that are more complex than in current automotive applications. This in turn drives for an intermediate target achievable, for example, by the development of low-RH membranes which would enable stack operation at 80°C without external humidification.

Both stationary and mobile applications can ultimately profit if the operation temperature is raised between 200 and 300°C. This would be made possible by a new membrane not relying on liquid water for proton conduction. Under such conditions, water and heat management would be greatly simplified and the poisoning of the catalyst by carbon monoxide would not be an issue anymore. Moreover, more options would become avalaible for waste heat use and internal reforming of lower alcohols.

A few promising routes are given below:

  1. Ortho-phosphoric acid-doped polybenzimidazole (PBI), first identified by the Ohio Case Western University, and ammonia polyphosphates are high-temperature membranes that allow operation up to approximately 200°C with no humidification requirement. However, issues that are barriers to use for automotive applications are instability in the presence of liquid water and inefficient cathode structures resulting in low areal power density.
  2. Recently, Australian researchers used a bilayer consisting of weatherproofing Gore-Tex® and a conductive plastic film to act both as catalyst and electrolyte.
  3. Replacement of current carbon blacks like Vulcan® XC-72 by nanostructured carbons such as nanofibers, nanotubes, etc. as electrode supports may also help realize the performance target since these materials appear much less sensitive to corrosion. Graphitization of standard carbon blacks at high temperatures (» 2000°C) also produces a material that is very resistant to oxidation and carbon corrosion.

Finally, the dependence of today’s PEMFC technology on hydrogen as fuel is one of its major drawbacks. Further research is required to enlarge the choice of suitable fuels and reforming techniques will remain a research topic in the short to medium term for stationary applications. For example, steam reforming of methane is envisioned for stationary co-generation, and the use of a compact fuel processor still seems to be feasible for automotive applications. But regarding the difficulties encountered, a clear tendency towards pure hydrogen is imposed by industry. In this context, the lack of a hydrogen infrastructure is one of the largest barriers to commercialisation of PEMFCs for vehicle powertrains. Hydrogen filling stations are not available on either a large or a smaller scale. In the future, local networks (such as the Hydrogen Highway on the Pacific West coast of North America) and cooperation with hydrogen companies will be the first step to alleviate this problem but a significant investment by either government and/or industry is mandatory. There is still some way before full commercialisation and market penetration of FCVs can be achieved.

2. The Direct Methanol Fuel Cell

a. How does a DMFC work?

The Direct Methanol Fuel Cell (DMFC) is a variant of Proton Exchange Membrane Fuel Cell (PEMFC), in the sense that it uses the same type of solid polymer electrolyte membrane as a separator between the electrodes in the unit cell. It differs from PEMFC because it uses unreformed liquid methanol fuel rather than hydrogen. A DMFC directly converts methanol and oxygen into carbon dioxide and water while producing electricity and heat. DMFCs operate at slightly higher temperatures than PEMFCs i.e., between 60 and 130°C.

Anode CH3OH + H2O ? CO2 + 6H+ + 6e (1)

Cathode O2 + 4H+ + 4e ? 2H2O (2)

Overall 2CH3OH + 3O2 ? 2CO2 + 4H2O (3)

The pure methanol is either pumped as an aqueous solution or mixed with water steam and fed directly to the cell’s anode. Here methanol is electrochemically oxidized into carbon dioxide forming six protons and six electrons per mole (1). The electrons are pushed round an external circuit whilst the (hydrated) protons cross the membrane in the form of hydronium ions H3O+, as in standard PEMFCs. At the cathode, the protons and electrons combine with oxygen (basically from air) to produce water (2). The overall reaction leads to the liberation of one molecule of CO2 per molecule of CH3OH: consequently, it is not a carbon dioxide-free or so called “zero emission” process.

Despite its known toxicity, methanol is a rather suitable fuel for fuel cell applications. It is one of the few organic compounds (including pure hydrogen, glycol, ammoniac and hydrazine for example) to be easily oxidized at relatively low temperatures. But due to the low-temperature conversion of methanol to hydrogen and carbon dioxide, the DMFC system like PEMFC, requires a noble metal catalyst. The cost associated with this catalyst is overbalanced by the ability of the unit to run without a reforming system. Liquid methanol is often considered to be easier to transport and supply to the public using current hydrocarbon fuel infrastructure than hydrogen.

The methanol oxidation reaction has a thermodynamic potential very close to 0 V vs. NHE (0.02 V) thus the theoretical open circuit voltage of a DMFC is about 1.2 V and the achievable efficiency is comparable with hydrogen-based systems. However, the process of converting methanol is not as simple as that of converting hydrogen; consequently, the average power density of a methanol-fed DMFC is lower than that of hydrogen-fed PEMFC (» 40%). In reality the anodic reaction is a multi-step process with carbon monoxide CO as an intermediate. The formation of CO causes the same problems as those originating from hydrocarbon reforming processes; therefore, the development of CO-tolerant anode catalysts is also required for DMFC applications. In particular, the use of platinum-ruthenium alloys has been quite successful. But the oxidation rate of methanol on Pt-based metals remains very low and higher catalyst loadings are required than for PEMFC electrodes: typical loadings are about 5 mg/cm-2 for metal blacks. The use of carbon-supported catalysts raises the concern of increased thicknesses that lead to unacceptably high layer resistance, hence low efficiency. A possible solution would be the use of highly conductive porous carbon supports such as carbon nanofibers and nanotubes.

Conventional DMFCs are operated with a liquid aqueous solution of methanol. In a standard setup the methanol solution is circulated with a peristaltic pump because of the very low fuel utilization of less than 10% at each passage. This is required in order to force the formed CO2 out of the cell. Typical methanol concentrations range between 1 and 2 mol/L. Although water is necessary for the electro-oxidation of methanol (1), the amount of dilution is a trade-off due of membrane issues.

As the DMFC technology is derived from PEMFC, Nafion® has been used as the membrane from the beginning, but its properties are not optimized for DMFC.

Nafion® has insufficient dimensional stability towards methanol and swells excessively in the presence of concentrated solutions.

Nafion® is not totally impermeable to methanol: the crossover rate increases drastically with concentration, thereby reducing cell voltage and poisoning the cathode catalyst.

Research efforts on membranes suitable for DMFCs have considered surface modification of the membrane to reduce crossover; additives, such as metals and inorganic salts and oxides, to enhance hygroscopy and/or dimensional stability; and formation of blends of Nafion® with other polymers. New membranes include fluorinated polymers containing silicon dioxide, fluoro- or nonfluoro-polymers with ether groups, and their composites with inorganic oxides. Some promising routes have been explored, but durability data of DMFCs under operating conditions are lacking. Although intensive work is done, improved membrane materials are still at an early stage of development.

b. A short history of DMFC

The direct electrochemical conversion of methanol as an alternative to steam reforming for the production of synthesis gas (H2 + CO,CO2) has been known since 1920, but the first report about the possiblity of using this reaction in anode fuel cell systems, published by the Austrian scientist Karl Kordesh from the Union Carbide Company, is dated back in 1951. The major developmental milestones for the DMFC technology did not come until the 1960s. At this time, methanol was converted to hydrogen by steam reforming and hydrogen was subsequently used in fuel cell systems. Researchers expected to remove the reforming step and use methanol directly to produce electricity.

The first DMFC systems were quite different than todays: in 1963, researchers at Allis-Chalmers Manufacturing Company tested a methanol fuel cell based on a liquid potassium hydroxide electrolyte. The degradation of the alkaline electrolyte by carbonate formation due to CO2 produced by the electrochemical reaction at the anode was observed as part of this work, and a process of regeneration of carbonate ions back to hydroxide ions was proposed. By 1965 oil companies like Shell and Esso were involved in the development of DMFC systems. Both worked on the replacement of the alkaline electrolyte by aqueous sulfuric acid electrolyte, and Esso presented a methanol-air fuel cell based on this new acid electrolyte. The system was further developed with the U.S. Army Electronics Laboratories for use in portable military communications equipment. At the same time, performance studies on noble metal alloy catalysts were carried out in both acid and alkaline electrolytes.

It is not before 1992 that a DMFC based on a solid Nafion membrane was first jointly developed at the Jet Propulsion Laboratory, Giner Electrochemical Systems and the University of Southern California. The technological choice of a solid electrolyte means that it became necessary to deliver methanol fuel to the anode rather than through the liquid electrolyte. This new delivery method thus began to look like the modern design of DMFCs.

Later in the 2000s, the DMFC technology has been given leading advantage in the portable electronic sector thanks to the development and investment made by companies such as SFC Smart Fuel Cell and Direct Methanol Fuel Cell Corporation manufacturing disposable methanol (and other fuel) cartridges.

c. DMFC applications and perspectives

DMFC outputs range from 1 mW to 25 kW. Applications are typically small mobile power devices such as cell phones, laptops and other consumer electronics (1-50 W), wireless tools (<1.5 kW) and some niche transport sectors such as marine and submarine vessels, scooters and motorbikes and as auxiliary power units (APUs).

Today, nearly three quarters of the DMFC units installed are being used for portable applications, thanks to the liquid fuel and the elimination of a fuel reformer. In this field, DMFCs are competing with PEMFCs fed by chemical hydrogen storage (or possibly small size pressurized storage). The advantage of PEMFCs is their higher power density. Their disadvantage is that today’s hydrogen storage devices are not satisfactory and that the missing hydrogen infrastructure prevents any easy refill (cf. the PEMFC applications and perspectives and the Issues sections). The appeals of a liquid fuel are that the fuel can be easily dispensed within the existing fueling infrastructure; consumers are more familiar with handling liquid fuels; and liquid fuels can be derived from natural gas and biomass. But these advantages come at the expense of low power density and additional ancillary water management components resulting in larger systems than their PEMFC equivalent.

In parallel with hydrogen-powered systems, there has been growing interest in direct reformation fuel cells that use organic fuels, among which DMFC is the most advanced technology. In these cells, the slow kinetics of the anode reaction is the limiting factor in the system response. The mechanistic complexity of oxidizing organic fuels like alcohols and the inadequacy of the Nafion® membrane’s properties for these fuels have limited such systems to lower power output than hydrogen-based systems. The high crossover rate of methanol through Nafion® reduces the cell voltage and the low activity of catalysts towards methanol electro-oxidation leads to high noble metal loadings inappropriate for commercialization. In DMFCs as in PEMFCs, the oxygen reduction reaction is a main source of efficiency loss. The difference is that in DMFC the parasitic oxidation of methanol crossing the membrane from the anode to the cathode adds to the overall loss. Therefore the central problem of today’s DMFCs is their low area/volume power density and better catalysts must be tracked after both for oxidation and reduction reactions.

Nevertheless, there is a strong driving force behing the commercial development of DMFCs relying on consumer pull for more effective products and high-end technology. Barriers to adoption are up to fall now, in particular for high-end leisure applications where consumers can afford to purchase and pull for the newest technological products (cf. Niche transport applications and Portable applications sections). In addition, the military sector has accounted for a significant part of past, present, and future DMFC development programmes for portable electronic products as part of a continuing commitment in fuel cell-powered equipment for soldiers. For military organisations, the objective is to develop the most efficient and cutting-edge technology and funding is not an issue.

At the current early stage of commercialization of DMFC products, the technology is being developed worldwide. Various countries including Taiwan, South Korea, India and Israel are willing to share the future market of consumer electronic goods. Many of the major electronics companies like Toshiba, Hitachi, and Sony are demonstrating miniature DMFCs for powering their equipment. Smaller fuel cell companies like Smart Fuel Cell, Neah Power and MTI Micro Fuel Cells are partnering with military and communication contractors. Toshiba has announced the commercialization of its DMFC battery charger for laptops and portable music players by 2009. Despite some recent delays, there is little doubt that the DMFC technology will get the leading market position in various mobile applications at some point in the near future.

The introduction of various standards and regulations is also expected to have a strongly positive impact on the future prospects of DMFCs for portable applications. In early 2008 methanol cartridges have been allowed onboard passenger aircrafts by the International Civil Aviation Organization (ICAO). This new regulation has removed a significant barrier to the use of DMFC-powered laptops during flights for instance. Furthermore, in 2006 the International Electrotechnical Commission published a safety recommendation about fuels like methanol, formic acid, borohydride, butane and hydrogen for fuel cells powering electronic devices. Specifications include rigorous testing and design requirements for safety use and transport. Compliance is required of devices that would be transported on aircrafts under the ICAO regulation.

On a general level, countries and organisations are on the way to setup codes and standards for DMFC systems not only for portable markets, but including them and the supporting infrastructure, for rapid widespread adoption.

Developers continue addressing membrane durability, fuel crossover and miniaturization challenges, and work hard to ensure the technology meets customer needs. To minimize the fuel storage, DMFC systems should be designed such that pure methanol can be used as the fuel. This requires that most of the water generated at the cathode is recovered and that the methanol concentration is measured online in the anode circuit. For small power devices, a simpler setup has been developed by MTI Micro Fuel Cells in which pure methanol is passively fed to the anode by capillary forces of metal foam acting both as gas diffusion media and current collector. Many improvements including miniaturization, efficiency, lifetime, environmental range of use, and cartridge design have been achieved. However, deficiencies are still related to system’s design, for instance methanol dosage into the anode circuit and carbon dioxide removal.

The biggest challenges facing micro DMFCs now are:

(i) High-performance room-temperature operation;

(ii) Miniaturization for on-chip (silicon) use;

(iii) Compatibility with existing system fabrication (CMOS, for example);

(iv) Avoidance of pumps for fuel and air which use electrical energy;

(v) Use of an efficient silicon-based proton exchange membrane and diffusion layers (novel porous layers for example);

(vi) Full integration with a microchannel architecture;

(vii) Fuel storage.

With respect to sustainable development, the DMFC suffers from the fact that methanol is almost only produced from fossil sources. But alternative “green” ways may be accessible in the future by regenerative and/or fermentation processes.

As mentioned above, direct reformation of hydrocarbon fuels, especially alcohols and other oxygenated fuels (ethanol, glycerol, ethyleneglycol, etc.) is of high interest for two main reasons: firstly, due to their thermodynamic properties oxidation is possible at relatively low temperature; secondly, they may be obtained from biomass even much easier than methanol, e.g. via fermentation. A further potential advantage is their reduced crossover rate through Nafion® compared to methanol. Yet, the major hurdle will be to find a catalyst capable to induce the break of the C-C bound at an acceptable rate at these low temperatures. Research is blooming in this field and promising results have been reported in 2009 by a team of scientists at the U.S. Brookhaven National Laboratory about the direct conversion of ethanol at room temperature. The catalyst is made of platinum and rhodium on carbon-supported SnO2 nanoparticles and is efficiently oxidizing ethanol into CO2 as the main reaction product. This represents a first milestone for research on “other-than-methanol” direct fuel cells.

3. The Alkaline Fuel Cell

a. How does an AFC work?

Modern Alkaline Fuel Cells (AFCs) are classified as low-temperature fuel cells. They operate between 60 and 90°C and use an alkaline electrolyte such as potassium hydroxide, usually in a solution of water. In an alkaline fuel cell, hydroxyl ions are produced at the cathode (2) and migrate to the anode side, where they react with hydrogen (1). Part of the water formed at the anode diffuses to the cathode, where it reacts with oxygen to form hydroxyl ions in a continuous process. The overall reaction produces water and heat as by-products and generates four electrons per mole of oxygen (3), which travel via an external circuit producing the electrical current.

Anode 2H2 + 4OH ? 4H2O + 4e (1)

Cathode O2 + 2H2O + 4e ? 4OH (2)

Overall O2 + 2H2 ? 2H2O (3)

The Oxygen Reduction Reaction (ORR) is a complex process involving several coupled proton and electron transfer steps. In acid solutions, the ORR reaction is electrocatalytic, and as pH becomes alkaline, redox processes involving superoxide and peroxide ions start to play a role and dominate in strongly alkali media. The reaction in alkaline electrolytes may stop with the formation of the relatively stable HO2 solvated ion (4), which instead of being further reduced into hydroxyl ions (5), can also be decomposed to dioxygen and hydroxyl ions (6). Although there is no consensus on the actual reaction sequence, two different pathways take place at the cathode in alkaline media:

  1. Direct 4-electron pathway

O2 + 2H2O + 4e ? 4OH (2)

  1. Peroxide or “2+2-electron” pathway

O2 + H2O + 2e ? HO2 + OH (4)

with HO2 + H2O + 2e ? 3OH (5)

or 2HO2 ? 2OH + O2 (6)

The kinetics of the ORR reaction is more facile in alkaline medium than in acid media such as H2SO4. Consequently, the use of non-precious metals like nickel is possible. Thanks to this comparatively high electrochemical rate at the cathode, AFCs units can attain overall electrical efficiencies greater than most other fuel cell types (50-70%).

Due to the faster kinetics for the ORR in alkaline media, a wide range of catalysts have been studied including noble metals, non-noble metals, perovskites, spinels, etc. The catalytic activity of the system also depends on the physico-chemical characteristics of the carbon support and the deposition method. High catalytic activity relies on a very fine and well dispersed catalyst particle. In the case of platinum, the particle size is generally in the nanometer range.

Potassium hydroxide solution (KOH) is almost exclusively used as the electrolyte because it has a higher ionic conductivity than sodium hydroxide solution, and potassium carbonate is less likely to precipitate than sodium carbonate. Concentrations typically range between 6 and 8 mol L-1 i.e., 30-40 wt% KOH in water. Concentrated aqueous potassium hydroxide is non-freezing, which allows quick start and operation at sub-zero conditions. However, it is highly corrosive and can cause sealing issues of gas compartments.

The main drawback of AFCs is their sensitivity to carbon dioxide from air. Carbon dioxide is absorbed by the electrolyte to form carbonate ions CO32-, which are less conductive than hydroxyl ions. This leads to an overall degradation of electrolyte properties. The formation of precipitated potassium carbonate K2CO3 can also lead to the blockage of the electrolyte pathways and electrode pores, and decrease cell lifetime. For effective operation it is therefore necessary to purify the gases fed to the cell. Since most methods for generating hydrogen from other fuels produce some carbon dioxide, this need for pure hydrogen has slowed down work on AFCs in recent years.

In most terrestrial applications, the KOH electrolyte is circulated through the stack, an option which has some advantages over the alternative immobilized systems chosen for space applications such as Apollo and the Space shuttle. The use of a circulating electrolyte allows thermal and water management to be easily controlled. It provides a very simple and effective way of cooling the stack via a heat exchanger. Moreover, impurities (e.g. carbon from electrodes or carbonates) can be easily removed, making the circulating electrolyte systems less sensitive to CO2 poisoning than the immobilized electrolyte systems.

The electrolyte circulation loop consists of a KOH tank, a pump and a heat exchanger. During start-up the KOH is heated to the desired operating temperature, typically 70°C. An air blower forces air into a CO2 scrubber (usually containing soda lime), from where the air is directed to the air inlet. The outlet air is directly exhausted to the atmosphere whereas the hydrogen is recirculated or dead-ended.

Historically, four different types of AFC cells have been developed:

Cells with free liquid electrolyte;

Cells with liquid electrolyte in the pore-system;

Matrix cells where the electrolyte is fixed in the electrode matrix;

Falling film cells.

All these cells rely on porous electrode architectures like those used in metal-air batteries. More recently, the alkaline anion-exchange membrane (AAEM) has attracted increasing attention. However, while the AAEM fuel cell holds great promise, developments still need to be made to achieve suitably conducting and stable membranes.

The anode electrode for AFCs has been less studied than the cathode, where catalyst containing platinum-group metals such as Pt/Pd has shown good performance and stability. Nickel, and in particular high surface-area Raney nickel, has been shown to be one of the most active catalysts for the hydrogen oxidation reaction (HOR) in alkaline media, despite reported deactivation effects.

b. A short history of AFC

Alkaline fuel cells (AFCs) were the first practical modern fuel cell following the pioneering work of Francis Thomas Bacon in the 1930s to 1950s. After his invention of the double-layer porous gas diffusion electrode made from activated nickel powder and his choice of an alkaline electrolyte (potassium or lithium hydroxide) instead of the sulfuric acid electrolyte employed by Grove, the english scientist developed and presented a series of alkaline fuel cell stacks running at 200°C under pressurized conditions (50 bar!) with increasing output powers up to 6 kW. The use of porous electrodes allowed increasing the surface area in which the reaction between the electrode, the electrolyte and the fuel occurs. Pressure was applied to hinder flooding phenomena in the tiny electrode pores. Electrodes were developed with increasingly stable interfaces and less corrosion problems. In 1961 Bacon founded his own company, Energy Conversion Ltd., and started commercialize AFC products. Meanwhile, the patents for the fuel cell were licensed by Pratt & Whitney, the aircarft engine manufacturer and industrial gas turbine division of United Technologies Corporation (UTC), which then successfully developed an AFC power plant for the NASA manned Apollo space program in the 1960s.

Due to their high CO2 sensitivity, AFCs were demonstrated in special applications where pure hydrogen and pure oxygen are easily supplied, such as space. They were originally used to provide electric power and drinking water to the astronauts in the Apollo and Skylab spacecrafts, and also by the Russian space program. NASA developed an asbestos-based porous matrix in which the KOH electrolyte was immobilized. Later in the 1970s, the 12 kW Orbiter fuel cell system was supplied by UTC for the Space Shuttle, giving impressive performance at 4 bar and a lower operating temperature of 90°C, but at the expense of high noble metal loading of 10 mg cm-2 at the anode (80% Pt–20% Pd), and 20 mg cm-2 at the cathode (90% Au-10% Pt)!

Alkaline fuel cells were also the first fuel cell technology to be put into mobile applications: demonstration of the first fuel cell powered vehicle was made with a farm truck in 1959 by Allis-Chalmers Manufacturing Company. The tractor with power of 5 kW was followed by an AFC-powered golf cart in 1962 where the AFC unit was fuelled by hydrazine and provided 4 kW of continuous power and 10 kW of peak power. Allis Chalmers Manufacturing Company also produced the world’s first fuel cell powered submersible. In the 1960s Dr Karl Kordesch from the Union Carbide Corporation developed the first circulating electrolyte systems leading to the manufacturing of an AFC motorbike running on hydrazine and of an Austin A40-based fuel cell car. Both vehicles were driven for several years on U.S. public roads in the early 1970s. The fuel cell car has been used by Dr Kordesch for his own personal transportation for three years; it had a driving range of 180 miles (300 km).

The AFC was developed and studied extensively throughout the 1960s until the 1980s, prior to the emergence of the proton exchange membrane fuel cell (PEMFC), which has subsequently attracted most of the attention from developers. However, primarily because the forecast cost reduction in PEMFCs has not become a reality yet, a renewed interest in AFCs has occurred during the last decade or so. The trend among academic research groups and several fuel cell companies is to check up the existing AFC technology and find potential ways to reduce costs while further improving its versatility.

For example, fuel cell companies such as UK-based Eneco (formerly Zetek Power) have recently focused on the design of circulating electrolyte low temperature unpressurised systems for backup power, stationary and mobile applications. The aim was to achieve a low cost mass production fuel cell. Injection-molded plastic frames were used to house the electrodes and build the stack. Low cost electrode production was ensured by the use of standard industrial processes such as rolling (calendaring) and pressing. A process was developed to eliminate CO2 from the fuel by selective absorption onto a carbon-based composite fiber. Eneco has applied this design to the production of a light AFC-powered bus and a hybrid 70 kW battery-AFC taxi.

c. AFC applications and perspectives

Today, due to the cell’s sensitivity to CO2 and the need to purify the hydrogen fuel, the AFC has only conquered predominantly niche transportation markets, powering forklift trucks, boats and submarines. It is still used in space applications and other controlled aerospace and underwater applications, when price is not an issue and high electrical efficiencies are requested. NASA continues to operate several 12 kW units in the Space Shuttle fleet. AFC systems now need to meet the challenging requirement of low cost, high performance and durability to become competitive in mainstream terrestrial applications.

The AFC technology suffers from the predominance of PEMFC, DMFC and SOFC technologies, which have left only small market opportunities due to their generally better performances. Yet, AFCs still have the potential of major improvements with modest investment: the utilization of non-noble metal catalysts and liquid electrolyte makes the AFC a potentially low-cost technology compared to PEMFCs, which employ platinum catalysts and specifically engineered membrane electrolytes. AFCs can produce up to 20 kW of electrical power and new designs have been reported to operate at temperatures close to ambient 23-70°C. Cost analyses have shown that ambient-air AFC systems for mobile and low power applications are less expensive than their PEMFC equivalents. Moreover, AFC fabrication methods (rolling, pressing, spraying, and screen-printing) among which rolling is the most commonly applied, are easily scaleable for mass production.

The future of the AFC technology will highly depend on the improvement of electrodes, especially the cathode, which causes the most part of cell losses. The development of new catalyst systems is more likely in alkaline media because of the wide range of options for the materials support and catalyst, as compared to acidic media which offer limited materials choice. For example, researchers in Wuhan University (China) have recently developed a cheap AFC prototype that uses a new membrane material, a silver cathode and an anode coated with nickel nanoparticles decorated with chromium ions that is more tolerant to corrosion than previous nickel cathodes. The power output is relatively low (0.05 W/cm²), but it provides a first proof of principle of a potentially much less expensive fuel cell.

The development of circulating electrolyte systems has shown to have advantages compared to the immobilized electrolyte systems for terrestrial applications that could create further commercial applications. Attempts are being made to push the already well developed AFC technology forward on this route. But electrolyte leakage and parasitic power losses are still challenges with the circulating electrolyte system and require the development of improved stack designs.

A 20,000 h lifetime has been achieved by Siemens and 15,000 h by UTC Fuel Cells with AFC stacks running on pure hydrogen and oxygen; but when air is supplied to the system, lifetime is significantly less (8,000 h) due to carbon dioxide poisoning. Durability is therefore a main issue for AFCs, especially when using non-noble metal catalysts and air. The 40,000 h target for stationary applications has not been reached yet, due to material issues.

In summary, the major issue for in the AFC technology today is a lack of R&D: alternative catalysts have been identified to replace platinum, but substantial efforts are now needed to meet the durability targets required for commercial applications.

4. The Phosphoric Acid Fuel Cell

a. How does a PAFC work?

The Phosphoric Acid Fuel Cell (PAFC) uses concentrated liquid phosphoric acid (H3PO4) as electrolyte and a graphite-supported platinum catalyst to speed up both electrode reactions like in PEMFCs and DMFCs. Anode and cathode reactions are also similar to PEMFCs but operating temperatures are slightly higher (150-220°C) making them more tolerant to reforming impurities like CO: a PAFC can withstand about 0.5-1.5% CO in the fuel without detrimental effects on efficiency. Optimal performance occurs at 180-210°C because above this temperature, H3PO4 undergoes a phase transition making it unsuitable as an electrolyte.

Protons formed at the anode by the oxydation of the fuel (1) migrate through the electrolyte to the cathode in the form of hydronium ions H3O+. Electrons generated at the anode travel through an external circuit, providing electric power along the way, and go to the cathode. There the electrons, hydrogen ions and oxygen (from oxygen gas or air) form water (2), which is expelled from the cell. The overall electrochemical reaction is the combination of hydrogen and oxygen to form water and heat (3). The thermodynamic open circuit potential between the oxygen reduction reaction and the hydrogen oxidation reaction is 1.23 V.

Anode 2H2 ? 4H+ + 4e (1)

Cathode O2 + 4H+ + 4e ? 2H2O (2)

Overall O2 + 2H2 ? 2H2O (3)

In a PAFC both electrodes are made of platinum nanoparticles coated on porous graphite. The binder is polytetrafluoroethylene (PTFE). Pt loadings are typically in the 0.1-0.5 mg/cm² range: 0.1 mg/cm² at the anode and 0.5 mg/cm² at the cathode. For more information about the electrochemical reactions you may go to the How works a PEMFC? section.

Phosphoric acid has a low tension of vapor at high temperature but solidifies at +42°C, which requires maintaining the fuel cell above this temperature at shutdown. An advantage is that concentrated phosphoric acid electrolyte can operate above the boiling point of water, unlike other acid electrolytes that require water for conductivity. It requires, however, that other components in the cell resist corrosion. Unlike AFC systems, the electrolyte is not circulating but immobilized in porous silicium carbide (SiC) matrix. The SiC matrix provides mechanical strength to the electrolyte, keeps the two electrodes separated, and minimizes reactant crossover. Operation temperature is controlled by a cooling fluid (air, water, or oil) flowing inside plates inserted between cells to hinder electrolyte evaporation. Nevertheless, acid losses are often observed upon long term operation, and H3PO4 must be replenished on a regular basis.

In the standard bipolar stack configuration, electrical interconnection and gas supply are achieved by carbon plates grooved with 1.5-2 mm deep channels. Bipolar plates for PAFCs can be of two types: they are either of dense nature and grooved on their both faces with perpendicular orientation of the channels, or they are porous and grooved on one (gas) side only. Bipolar plate materials are commonly carbon/phenolic resin composites.

Another advantage of phosphoric acid is that it does not react with CO2, which allows use of reformed hydrogen. PAFCs use hydrocarbon sources such as natural gas, propane or waste methane. Hydrogen is extracted by an external reformer. Due to the use of platinum catalyst they are susceptible to carbon monoxide and sulfur poisoning at the anode. This is not an issue when running on reformed or impure gases. Susceptibility depends on temperature: since the PAFC operates at higher temperatures than the PEMFC, it exhibits a greater tolerance. Carbon monoxide tolerance to CO can be as high as 1.5% depending on the exact conditions, but sulfurous acid tolerance is not more than 50 parts per million (0.005%). If the hydrocarbon fuel is gasoline for example, sulfur must be removed before entering the fuel cell to avoid poisoning the electrode catalyst.

The net electrical efficiency is in the range 36-45% and an overall 80-85% total efficiency is achieved with cogeneration of electricity and heat (T » 85°C). These values are maintained in a range between 50 and 100% of the nominal power. Power density usually lies between 0.2 and 0.3 W/cm². However, efficiency has a tendency to decay with time due to stack ageing, i.e., mainly electrolyte evaporation and electrode corrosion. PAFC modules of 200-250 kW are in commercial operation, and units of 11 MW have been successfully tested in Japan during the 1990s.

b. A short history of PAFC

The PAFC is the technology with the greatest experience in consumer applications. More than 200 stationary PAFCs have been installed worldwide, providing power and heat to hospitals, hotels, office buildings schools, utility and waste water treatment power plants, landfills and even an airport terminal. UTC Fuel Cells paved the way for the technology, selling robust PAFC systems since the early 1990s. The company recently reached the milestone of one billion kW/h of energy with its latest model of power plant.

However, the use of phosphoric acid as electrolyte was not reported in the literature before the 1960s due to its poor electrical conductivity as compared to sulfuric acid. In 1961, promising results were obtained by G.V. Elmore and H.A. Tanner with an intermediate fuel cell based on a 35:65 H3PO4/SiO2 powder electrolyte pasted into a PTFE gasket. This fuel cell run an air instead of oxygen and was particularly stable. In the late 1960s and 1970s, major advances in electrode materials and lingering problems with other types of fuel cells spurred new interest in PAFCs. The U.S. Army explored the potential for PAFCs to operate on “logistic” fuels, i.e., hydrocarbon fuels commonly available in the field. A cell was produced by Allis-Chalmers Manufacturing Company that integrated a plastic-bonded electrolyte and a steam reformer by Engelhard Industries.

Carbon paper substrates and PTFE-bonded carbon layer as catalyst support were first developed by Karl Kordesh and R.F. Scarr at the Union Carbide Company. A partnership known as TARGET, for Team to Advance Research for Gas Energy Transformation also made significant contributions to research on PAFCs. Funded primarily by Pratt & Whitney and the American Gas Association, TARGET research led to fuel cell power plants from about 15 kW in 1969 to nearly 5 MW in 1983. The military also installed a number of stationary PAFC units from International Fuel Cells Corporation in army buildings where air quality is critical between 1993 and 1997. The systems have been operating either in connection with the utility grid or off-grid.

During the 1970s to 1990s, research on PAFC systems was further focused on transport applications, both in Japan and in the U.S.A. With an aim at developing electric vehicles, researchers at Los Alamos National Laboratory designed a golf cart powered by a phosphoric acid fuel cell. In1994 a 50 kW PAFC from Fuji Electric was integrated in a transit bus and run in Georgetown. In 1998 the U.S. Department of Transportation began new tests in a bus fleet with a 100 kW module from International Fuel Cells Corporation, a joint venture between Toshiba and U.T.C. Fuel Cells. However, due to the long warm-up period of PAFCs, their applicability in private vehicles remains limited.

c. PAFC applications and perspectives

PAFCs are typically used for medium to large-scale stationary power generation. The power output of existing modules is 25-250 kW and units can attain the MW range. They are essentially commercialised by the company O.N.S.I., a subsidiary of Toshiba and U.T.C. Fuel Cells and more than 200 units are mainly installed in the U.S.A. and in Japan. A few hundred more units between 50 and 200 kW have been developed by other Japanese manufacturers: Mitsubishi Electric, Fuji Electric. The two biggest PAFCs are also located in Japan, i.e., the 5 MW power unit from Fuji and the 1 MW from Toshiba. By comparison, few PAFCs are installed in Europe. Most of them are distributed power units from O.N.S.I. or Fuji Electric. One 200 kW co-generation system running on natural gas supplied by GDF has been developed by EDF/GDF for demonstration purposes.

Today PAFC is a mature technology having probably reached its full potential of development. This type of fuel cell exhibits good reliability: 40000 h durability (and over) has already been demonstrated. But further significant improvements are not expected. Relative weaknesses of the technology include expensive platinum catalyst, sensitivity to CO and S poisoning, and corrosion by the concentrated phosphoric acid electrolyte. With the exception of the bus program of U.T.C. Fuel Cells with the University of Georgetown (cf. A short history of PAFC section), all applications have been stationary. Through demonstration programmes experience has been gained about the dynamic response and the maintenance of a fuel cell under various load conditions, the electric and thermal efficiency, availability, servicing costs, emission levels of NOx, CO2, etc. This has contributed to make the technology more popular.

5. The Molten Carbonate Fuel Cell

a. How does an MCFC work?

The Molten Carbonate Fuel Cell (MCFC) is a high temperature fuel cell that uses an immobilised liquid molten carbonate salt as the electrolyte, which is retained in a ceramic matrix made of lithium aluminate LiAlO2. In the MCFC the carbonate anion CO32- serves as ionic conductor under the operating temperatures of 620-660°C. Electrical efficiency is 45-60% and up to 85% if the waste heat is used for cogeneration.

Anode 2H2 + 2CO32- ? 2H2O + 2CO2 + 4e (1)

Cathode O2 + 2CO2 + 4e ? 2CO32- (2)

Overall 2H2 + O2 ? 2H2O (3)

Upon heating, the ionic salt melts and frees carbonate ions. These anions flow away from the cathode towards the anode where they combine with hydrogen producing water, carbon dioxide and electrons (1). The electrons are directed through an external circuit to the cathode and generate electricity. At the cathode oxygen and carbon dioxide recycled from the anode side react together to give back carbonate ions (2). The overall reaction is the same as for the other types of fuel cells, i.e. the electrochemical combination of hydrogen and oxygen to form water (and heat) (3).

In this particular case however, carbon dioxide is involved both at the anode (produced) and at the cathode (consumed). CO2 formed at the anode must therefore be recycled prior its utilization at the cathode. In practice, anode waste stream is burned; the resulting mixture of steam and CO2 is mixed with fresh air and then recirculated to the cathode. The heat released at the burner is used to pre-heat the reactant air, thus improving the efficiency and maintaining the operating temperature of the system. In MCFCs the open-circuit potential of reaction (3) not only depends on hydrogen, oxygen and water concentrations but also on the partial pressures of carbon dioxide, which are generally different in each single cell compartment.

At the cathode, oxygen gas first dissolves into the molten alkali carbonate. Two possible electrochemical processes then take place giving the reaction (2):

Peroxide path: O2 + 2CO32- = 2O22- + 2CO2 (4)

O22- + 2e ? 2O2- (5)

2O2- + 2CO2 = 2CO32- (6)

Superoxide path 3O2 + 2CO32- = 4O22- + 2CO2 (7)

O2 + 3e ? 2O2- (8)

2O2- + 2CO2 = 2CO32- (6)

Reaction (6) is the rate-determining step in both cases. In MCFCs the oxygen reduction reaction (ORR) does not depend on the electrode material but on its porous structure.

At the anode, like in other types of fuel cells the hydrogen oxidation reaction (HOR) proceeds much faster than the ORR. It is also not influenced by the electrode material. Exchange current densities of HOR are about one order of magnitude higher than the values of ORR (» 100 mA/cm² vs » 10 mA/cm²). HOR is decomposed on three successive steps:

H2 absorption H2 + 2M = 2M-H (9)

2M-H + 2CO32- = 2OH + 2CO2 + 2M + 2e (10)

2OH + CO2 = H2O + 2CO32- (11)

Although there are many kinds of carbonates available the alkali salts are the most stable at these temperatures, i.e., lithium, potassium and sodium carbonates. No pure carbonate melts in the 550-700°C range but a few eutectic mixtures do. One of the eutectics is a 62% mol lithium-38% mol potassium carbonate Li2CO3:K2CO3 melting at 488°C. This mixture contains a high proportion of light lithium element, which is a good ionic conductor. Moreover, it is nontoxic. Given its low melting point and high conductivity in the 600-700°C range, this eutectic mixture has been preferred in the MCFC systems since the 1970s.

The electrolyte is immobilized in a ceramic matrix (thickness » 40 µm). In order to avoid electrode flooding and achieve a stable electrode/electrolyte interface, the pore size distribution of the electrolyte must be smaller than that of the electrodes. Since the supporting matrix impregnates the molten carbonate salt, it is a soft structure that serves as gas sealing under compression. Nevertheless, problems associated with this type of electrolyte arise from creepage, corrosion and evaporation phenomena. Gas crossover may also occur through the matrix leading to overheating and performance losses.

Like in other types of fuel cells, a porous electrode is used to effectively enhance gas diffusion and increase electrochemical reaction rates. Platinum and other precious metals are not necessary and MCFC electrodes are made of porous nickel, which is stable in the carbonate medium up to a maximum temperature of 600-650°C. It is this electrode stability that limits the operating range of MCFCs. At both electrodes, the nickel provides catalytic activity and conductivity. At the anode, nickel is generally in reduced form and alloyed with 2-10% of chromium (or aluminium). Cr addition enhances porosity and surface area of the anode structure. At the cathode, it is oxidized in NiO and lithiated with 2-3 mol% of Li+ ions in order to minimize nickel dissolution, which could otherwise adversely affect fuel cell performance.

In the common stack configuration, ten to several hundreds of cells are connected in series by metallic separator plates. Their role is to divide the active cell components, make the electrical connection between adjacent cells and distribute fuel and oxidant gas to each cell (feed and exhaust). A coolant system is also integrated to control the stack temperature. Channels are grooved on both side of each bipolar plate with a cross flow pattern. As the atmosphere is reducing in the anode compartment and oxidizing in the cathode compartment, the stainless steel plates must be coated by a nickel alloy on the anode side. Soft gas sealings are attached between bipolar plates. Gas sealing design is a crucial aspect of the technology for high temperature fuel cells: an important issue with MCFCs comes from corrosion of the plates and/or sealer materials and progressive loss of gas tightness. Because the separator plate corrodes at high temperatures, the sealing surface is typically processed with an aluminized treatment.

The advantages associated with the MCFC system are mostly the result of high operating temperature. This feature allows achievement of high electric efficiencies of 50% and up to 60% thanks to the dramatic increase of the reaction kinetics. It also removes the need for noble metal electrodes, which at the same time limits carbon dioxide poisoning faced by low temperature fuel cells. The high operating temperature of the MCFC (650°C) provides fuel flexibility: the fuel (coil, oil or natural gas) is possibly converted to a hydrogen-rich fuel gas by external or internal steam reforming (cf. Stack and systems section). Cogeneration and coupling with a gas turbine are possible. Since the MCFC can accept carbon monoxide as fuel, an external device to remove CO is not required.

Technological difficulties with MCFCs are related to the use of a corrosive liquid electrolyte rather than a solid and the requirement to inject carbon dioxide at the cathode since carbonate ions are consumed at the anode during HOR (1). There have also been some concerns with high-temperature corrosion of the nickel oxide bipolar plates by the electrolyte (see above), but the issue has been addressed in more recent systems to achieve sufficient lifetime. Due to stresses created by the freeze-thaw cycle of the electrolyte during start-up/shut-down cycles, MCFCs are considered best suited for providing stationary, continuous power, and cogeneration of heat and electricity for industrial applications. They are generally operated at high voltage and low current, e.g. 0.8-0.85 V single cell voltage and 0.15 A/cm² giving power output densities of ca. 0.12 W/cm² or 1.2 kW per m² of active area.

b. A short history of MCFC

The history of high temperature fuel cells, i.e., Molten Carbonate Fuel Cells (MCFCs) and Solid Oxide Fuel Cells (SOFCs) has common roots back in the 1930s: Swiss scientist Emile Baur was the first one who tested different ceramic and alkaline silver electrolytes for use in fuel cells in 1937. Tests with alkaline materials were unsuccessful due to technical limitations such as parasitic reactions at the electrodes with atmospheric gases including CO and insufficient conductivity for gas processing. In 1946, based on the previous results of the Swiss team the Russian Oganes Davtyan developed a fuel cell with an ionic solid conductor as electrolyte operating at 700°C with aim at increasing the mechanical strength by addition of monazite sand (a phosphate mineral), sodium carbonate, tungsten trioxide and soda glass. But his system was not very operative.

In the next decade, the German researchers Broers and Ketelaar decided to go further with this electrolyte mixture considering that it was more accurately described as a molten liquid immobilised in a solid matrix rather than an all-solid electrolyte. As a result, they began to focus on electrolytes made of carbonate salts. By 1960 they had developed a cell where the electrolyte mixture (lithium, sodium and/or potassium carbonates) impregnated a cylindrical porous matrix of sintered magnesium oxide. The cathode was silver and the anode was platinum/nickel. This first MCFC was tested for six months between 550 and 700°C on a variety of fuels. A progressive loss of electrolyte was observed yet. Also in 1960 the industrial association Gas Technology Institute began to study MCFCs and demonstrate its potential for high power generators. An electrolyte retained between a silver cathode and a nickel anode was used in the MCFC system. In 1963 internal steam reforming of methanol at the anode surface was assessed, and in 1965 further improvements of the interface between the electrolyte and the two electrodes were achieved. Work at General Electric involved finding a new method for holding the molten carbonate electrolyte within a porous “diffusion” electrode rather than a solid one, unlike Francis T. Bacon who was using a “free molten” electrolyte between two solid electrodes. Other groups were also developing porous electrodes and semisolid or “paste” electrolytes.

In the mid-1960s, demonstration tests started on a wider scale as several MCFCs ranging from 100 to 1000 W output power manufactured by Texas Instruments were evaluated at the U.S. Army’s mobility Equipment R&D Center (MERDC). These specially designed units were fed by externally reformed “combat gasoline” to demonstrate the possibility to use fuels directly available on the field.

Materials and design improvements have continued since then: in the early 1990s, Ishikawajima Heavy Industries (I.H.I.) in Japan successfully operated a 1 KW MCFC power unit for 10,000 h continuously. M-C Power Corporation of Illinois installed a 250 kW cogeneration unit at the Miramar Marine Corps Air Station in San Diego in 1997, and two years later installed a 75 kW stack then gradually upscaled to 300 kW to demonstrate the feasibility of a MCFC commercial plant. At the same time, the U.S. Company Fuel Cell Energy operated a 2 MW MCFC demonstration plant during 3,000 h in Santa Clara, California, with the ultimate goal to scale up their units to 3-6 MW and even 40 MW.

c. MCFC applications and perspectives

Due to their high operating temperature (» 650°C) molten carbonate fuel cells are mostly employed in large stationary applications and looking forward, it is unlikely that they will find other end-use applications. Although the market for MCFC appears somewhat limited, its advantages are numerous. The high temperature opens the opportunity of using waste heat to generate steam for space heating, industrial processes, or in a steam turbine to generate more electricity as is already the case in modern gas combustion power plants. The current MCFC technology can therefore provide in-situ power generation and/or a 75-400 kW distributed power supply system, but when using combined heat and power (CHP) generation it can produce several MW of power. In addition, the use of fuel cells to produce power makes the plants quieter and cleaner thus allowing the units to be placed closer to consumers. It also improves reliability and efficiency and would possibly remove the strain from existing power grids (cf. Benefits of fuel cells section).

To date, MCFCs have operated on hydrogen, carbon monoxide, natural gas, propane, landfill gas, marine diesel, and synthesis coal gasification products. Since the current manufacturing technology for the electrodes and the electrolyte-supporting plate can uniformly fabricate components up to 1 m², this unit surface area is commonly used. For identical stack surfaces, the power output can be varied by more than 30% by increasing the operation pressure. MCFC stacks typically generate 250 to 350 kW depending on the pressure. Demonstration units have produced up to 5 MW, but designs exist for units of 50-100 MW capacity.

The MCFC systems must meet a durability target of 10% power degradation after 40,000 h operation to be of practical use. The main technological hurdle in reaching this durability standard is nickel shorting with concomitant electrolyte loss. During long term operation (> 10000 h), nickel shorting in the electrolyte is caused by the reduction of nickel ions dissolved from the bipolar plates into nickel metal nanoparticles that cluster progressively and form electrical paths between the two electrodes, resulting in performance loss. The supporting matrix for the electrolyte made of LiAlO2 is a stable ceramic under the MCFC conditions, but its actual behavior in molten carbonates under prolonged periods is not well understood. Although the amount of evaporated carbonate is considered negligible for cell performance, the alkali oxides evaporated may influence peripheral devices such as gas turbines, etc.

Cooperative projects have been and are being consistently established between electric utilities and commercial companies in order to achieve commercialization of MCFC plants. In the current MCFC market there are a number of academic institutions (» 38% of the total) and a majority of commercial companies (» 51%) both addressing issues for mass production such as performance, durability and cost. Government agencies account for less than 10% of the sector. The general trend is a focus on customer-driven commercial prospects. Nevertheless, commercialization is not imminent: the highest level of activity in the MCFC sector is in the research and development field.

Today, over ten Japanese companies are developing MCFCs. The main developers of the technology are the German MTU CFC solutions and the American Fuel Cell Energy, its partner that manufactures the stacks. MTU is now selling CHP modules installed in various locations, primarily in Germany. To demonstrate the feasibility of the MCFC technology for distributed power generation in electrical utilities, Fuel Cell Energy in the U.S.A. and I.H.I. in Japan have been developing and manufacturing stacks operating at ambient pressure or pressurized conditions for twenty years. Systems including internal reforming can be directly fed by methane, natural gas or biogas. Fuel Cell Energy has already demonstrated their 250 kW power unit at a number of sites in the United States and in Japan. It has been awarded several grants for R&D projects and made a number of technical advancements in recent years. I.H.I. has also been granted subsidies from the Japanese government for developing their MCFC module and is currently working under the New Sunshine Project about the development of a 300 kW pressurized system.

Several government funded organisations are fostering the development of the MCFC technology worldwide: for example, the California Energy Commission through the Energy Technology Programme in the U.S.A.; the C.I.E.M.A.T through R&D on new cathode materials and other components in Spain; the Central Electrochemical Research Institute (C.E.C.R.I) through development of MCFC prototypes in India; and the MCFC Research Association in Japan with a main interest in power generation systems for small (10 kW) and large stationary applications.

The installation of MCFC power generation into plants has begun in 2003. Large increase in the number of installed units has been observed since then (> 200), thanks to lower costs and local government incentives, for example the Project 100 from the Connecticut Department of Public Utility Control under which 6.6 MW of MCFC units from Fuel Cell Energy have been approved for installation in 2009. Nevertheless, efforts of commercial companies to fit the technology to customer’s needs are of upmost importance. The industry players are very focused on trading their products on a mass commercial scale. It is vital to adopt this approach if full commercial potential is to be exploited in the near to mid-term future.

Attempts to reduce costs are focused on the engineering and processing of the plant. The market for remote power generation, in the 100-500 kW range, is potentially large when the efficiency of the given power generator is more than 45% higher than that of a grid power plant. In the future, increases in durability and performance should enable the MCFC to be utilized in many small and large stationary applications. It may be feared, however, that due to the concurrent SOFC technology, when most problems with SOFCs are solved, work on the MCFC might be stopped.

6. The Solid Oxide Fuel Cell

a. How does a SOFC work?

A solid oxide fuel cell (SOFC) uses a solid ceramic as electrolyte. An important feature of SOFCs comes from the fact that ionic transport through the electrolyte is achieved by oxide ions O2-, i.e., oxidant-related species, and not protons as in PEMFC for example. Therefore, the electrolyte must be a good oxygen-ion conductor: O2- ions are transported through oxygen vacancies in the crystalline structure of the solid oxide, which is a high-temperature process in order to achieve practical conductivity values. The most popular electrolyte used in SOFCs is yttria (º yttrium oxide Y2O3)-stabilized zirconia (º zirconium oxide ZrO2) (formula (ZrO2)1-x(Y2O3)x or YSZ). The intrinsic ionic conduction of zirconia has been improved by doping it with yttrium ions Y3+ that introduce electronic vacancies in the Zr4+-based zirconia structure. As a consequence, the temperature range 800-1000°C of SOFCs has been imposed by the choice of this material. According to the same principle, doping of zirconia with Ca2+ is also possible and the electrolyte’s formula becomes (ZrO2)1-x(CaO)x.

Electrical efficiencies vary from 45 to 55% depending on the operating conditions, and the total efficiency may attain 65-70% in cogeneration mode. Conversion efficiency of SOFCs is expected to be very high, although the thermodynamic value is lowered with increasing temperature. This is mainly because of the good kinetics of the cathode and anode reactions and also because of the possible utilization of waste heat in the reforming process.

Anode 2H2 + 2O2- ? 2H2O + 4e (1)

Cathode O2 + 4e ? 2O2- (2)

Overall 2H2 (+ CO) + O2 ? 2H2O (+ CO2) (3)

During operation of the cell, oxygen (in the form of air) is supplied at the cathode where it is reduced into two oxygen ions per molecule (2). The ceramic electrolyte conducts the oxygen ions from the cathode to the anode through the crystal lattice whilst electrons are pushed round an external circuit to produce electricity. At the anode, on consideration that the fuel is not always pure hydrogen but often a reformate fuel composed of H2 and CO; the anodic reaction may alternatively be a mixture of (1) and (4):

Anode CO + O2- ? CO2 + 2e (4)

The oxygen ions combine with hydrogen to produce water (1) and with carbon monoxide to produce carbon dioxide (4). Unlike in most fuel cells, water is produced at the anode and not at the cathode. The overall process (3) generates heat as well. High operation temperature eliminates the need for precious metal catalysts replaced by nickel-based compounds, and can reduce cost by recycling the output heat for internal steam reforming of hydrocarbon fuels to hydrogen and carbon monoxide. As sketched in the anode reaction sequence above, the tolerance of SOFCs to CO poisoning allows CO and H2 derived from coal gas for example, to be employed as a direct fuel source without further purification as in low-temperature fuel cells. The remaining fuels can be burned and used in the combined gas turbines.

In a SOFC, the ceramic electrolyte is coated with two specific anode and cathode porous materials. On the anode side, the fuel electrode must be resistant to the highly reducing and high-temperature conditions, while on the cathode side the air electrode must be resistant to the highly oxidizing and high-temperature conditions.

The most common anode electrodes are nickel-YSZ cermets, i.e., metal-ceramic mixtures where the nickel particles dispersed at the surface of the ceramic oxide provide electronic conductivity and catalytic activity while the YSZ provides mechanical stability combined with adequate high-porosity and high-surface area structure, plus ionic conductivity and good thermal expansion compatibility with the electrolyte.

The cathode electrode is usually a mixed ionic and electronic conducting ceramic compound. Typical materials are perovskite oxides (structure type ABO3) and include strontium-doped lanthanum manganite ((La,Sr)MnO3), lanthanum-strontium ferrite ((La,Sr)FeO3), lanthanum-strontium cobaltite ((La,Sr)CoO3), and lanthanum-strontium cobaltite-ferrite ((La,Sr)(Fe,Co)O3). These materials exhibit good oxidation resistance and high catalytic activity in the cathode environment. Calcium may be used for a better match between thermal coefficients and for addressing problems of formation of a passivation layer between the electrode and the electrolyte.

Due to the high temperature of SOFCs, interconnect materials are mostly ceramic oxides that must be good electronic conductors, gas tight and exhibit high resistance in both reducing and oxidizing environments. Preferred materials are lanthanum chromites doped with strontium and calcium, and occasionally include nickel and/or cobalt. For planar designs, metallic materials based on chromium with Y, La, Ce, Zr are possible.

Fuel flexibility is one of the characteristic features of SOFCs when compared with other fuel cells. In contrast with the PEFC, which is closely linked to the hydrogen technology, the utilization of hydrocarbon fuels is an important point: an anode SOFC can be fueled by either pure hydrogen or a mixture of hydrogen and carbon monoxide gas produced by reforming a number of available fuels like natural gas, propane, liquid alcohols, coal gas, biogas, biomass, ammonia, etc. Internal reforming is possible for this type of fuel cells, and endothermic steam reforming is the process generally employed at such high temperature levels (700-800°C) (cf. Stack and systems section). Internal reforming occurs directly on Ni catalyst particles distributed onto the anode surface, and preferentially under high partial water pressure to avoid side reactions forming carbon. As compared to external reforming, it offers the huge advantage of compactness and simplicity.

Among the factors influencing the operation of SOFC, pressurization of the system is beneficial to electrical performances, whereas any temperature decrease will be detrimental due to the negative effect on electrolyte conductivity. In general, SOFC materials must be chemically stable under both reducing and oxidizing atmospheres, and against each other. At such high temperature level, it is highly desirable that the different components have identical thermal dilatation coefficients in order to minimize thermal stresses in the fuel cell.

b. A Short history of SOFC

In the SOFC technology, development of materials, processing techniques and stack design has been concomitant since the beginning of R&D due to very close links between the main technological issues.

Seminal work on SOFCs is common to MCFCs (cf. A short history of MCFCs) and was done in the 1930s by Swiss scientists Baur and Preis based on a solid compound that had been discovered thirty years before by Walther Nernst: this mixed oxide composed of 85% zirconia (ZrO2) and 15% yttria (Y2O3) exhibited good oxygen ion conductivity at high temperatures. It had been used in the “Nernst lamp” patented by Nernst in 1897. Baur and his colleague synthesized and measured the ionic conductivity of different ceramic materials using zirconium, yttrium, cerium, lanthanum, and tungsten derived from the “Nernst-Mass” material. For a few compounds, they obtained values compatible with practical use as fuel cell electrolyte, but as their price was quite prohibitive, they concluded that it was necessary to find out less expensive and more performant materials for really commercial applications. They also observed unwanted chemical reactions with fuel cell gases including carbon monoxide and for this reason discontinued efforts.

Many technical problems were persistent during the two following decades despite a noticeable acceleration of research into high-temperature fuel cells in the 1950s. General feeling among researchers was that MCFCs showed better short-term promise than SOFCs. But the hope for finding a stable solid electrolyte and building a cell tolerant to carbon monoxide was still there.

A big technological breakthough was obtained in 1962 by the U.S. Company Westinghouse, which invented the tubular design in which the different active layers are concentric around a central cylindrical support. Experimentations were performed at temperatures between 800 and 1100°C with a cell using again a variation on the Nernst-Mass material as electrolyte. In the tubular design, the cathode (at the center) is fabricated by extrusion, and then the electrolyte, the anode and the interconnect materials are deposited by electrochemical vapor deposition. No seal is added. An array of fuel cell unit tubes is connected in parallel by nickel plates. Air flows inside the porous central zirconia support, while the fuel is fed externally where it is partly oxidized; the rest is recycled with outflowing air for inlet gas preheating. High conversion efficiency and long lifetime (over 70000 h) were achieved with this stack, confirming that degradation in SOFCs could be quite low with a proper choice of materials. But this success was somehow counterbalanced by an intrinsically low power density of the tubular geometry and a very high fabrication cost due both to the vapor deposition method and the expensive ceramic materials used.

In the following period, several groups aimed to overcome these issues in three ways:

  1. Reduce the amount of expensive elements in fuel cell components such as yttrium, scandium, gallium, lanthanum, etc.
  2. Adopt a cheap processing technique for the ceramic oxides, i.e., a “wet” method (for example rolling, screen-printing, sol-gel) with a smaller number of sintering steps.
  3. Increase the power density par area or volume.

Firstly, a simpler planar configuration was considered with an electrolyte support working at 1000°C; then with an anode support that helped reduce electrolyte thickness while allowing good performance at lower temperature (700-800°C). The decrease in working temperature enables the use of metallic interconnects, which are much less expensive than ceramic interconnects used in first-generation SOFCs. The planar design also gives higher power densities but needs efficient gas sealing at the edges and good contact across the interfaces. In particular, the “monolithic” technology proposed by Argonne National Laboratory in the 1980s, in which the anode and cathode are folded, and gas flows are parallel or crossed, is interesting in the sense that it is based on a cheaper “wet” fabrication process and allows achieving high power densities.

In parallel, Westinghouse continued to develop the tubular SOFC concept and in 1998 joined with Siemens to found the Siemens-Westinghouse Power Corporation. More recently, climbing energy prices and advances in materials technology have revived work on SOFCs. Nevertheless, the SOFC remains a sophisticated technology. The strong financial support of the military sector has been a key driver for the renewed interest that is observed, especially for portable applications. In addition, several car manufacturers, e.g., B.M.W., are working on APUs for onboard comfort equipment in luxury-class vehicles (air conditioning, music and video player powering, etc.). Finally, government funded programmes favor a continuing commitment of industry and institutions to achieve new development milestones (cf. SOFC applications and perspectives).

c. SOFC applications and perspectives

The applications of SOFC systems may be classified into three main groups: stationary (large stationary power > 10 kW and small stationary power < 10 kW), niche transportation and portable. Due to their attractive efficiency and fuel flexibility advantages associated to an all-solid technology, the primary application of SOFCs has been large stationary power for the electric utility and distributed generation. Demonstration power plants have capacities up to 100-300 kW.

Today, the majority of systems installed are being used for small stationary applications including domestic power-and-heat supply (cogeneration mode), industrial applications requiring uninterruptible power supply (UPS) and military applications. Due to their low noise and low emissions, SOFCs are also well suited for medium-to-large scale, on-site power generation or combined heat and power (CHP) units for hospitals, schools, office buildings, hotels, etc.

In recent years, there has been a high level of diversification in the SOFC sector. A strong interest has grown in developing smaller modules in the range of several kW to several tens kW for stationary and automotive use. The compact size and cleanliness of SOFCs indeed make them especially attractive for urban locations. SOFCs are for example evaluated for use in telecom backup power. In the niche transport sector, the possibility of using SOFC as auxiliary power unit (APU) has given rise to more general concern on how to use the SOFC technology in transportation fields. As a result, small SOFC units in various sizes ranging from several hundreds W to several hundreds kW are now developed as stand-alone systems to supply auxiliary power to selected vehicles: cars, trucks, ships, trains and military vehicles.

In the large stationary sector, efforts have been driven towards increasing the conversion efficiency up to 85-90% by combining SOFC units with gas turbines (GT) in larger MW-size systems in order to recover some of the high-quality waste heat as additional power. The first demonstration was made in 2000 by Siemens Power Generation at the University of California with a 220 kW pressurized SOFC-microturbine cogeneration unit. Other developers like Mitsubishi Heavy Industries are currently working on similar SOFC-GT hybrid solutions. If successful, this hybrid option could open up new markets that were previously closed when considering SOFC technology as a stand-alone power solution.

In comparison to other fuel cell technologies, SOFC appears to be more focused on R&D than commercialization because there is still much developmental work to do before getting to the different possible markets. A few examples of main research topics are presented:

1. In recent years, it was established that only SOFCs based on metallic supports onto which are deposited the active ceramic layers could possibly reach acceptable costs. In this new generation design, the quantities of expensive ceramic materials are minimized. In addition, a metallic support has better mechanical, electrical and thermal properties than a ceramic oxide, which allows increasing robustness and durability of the stack.

2. There has also been recent interest in SOFC systems that can be processed in a simpler manner (at the expense of lower conversion efficiency). A typical example is the microtubular design, in which the interconnection is made at low temperature outside the stack without using interconnect materials. Fabrication is easier but electrical losses are higher.

3. Micro SOFCs based on silicon-chip technology are developed for portable electronic applications by companies like Lilliputian Systems, which is a spin-out from the Massachusetts Institute of Technology. The little power generator is fueled by recyclable fuel cartridges and provides a factor ten improvement in volumetric energy density when compared to conventional lithium-ion batteries.

4. In combination with internal reforming, compact “direct-hydrocarbon” or “one-chamber” SOFC units have certainly a great potential for commercialization. Small-scale SOFCs less than 1 kW are being developed for mobile applications using this concept. In turn, this implies that the interconnect technology is crucial to determine whether the manufacturing of SOFC stacks may be successful or not.

Interestingly, the operation temperature of SOFC stacks mostly depends on the interconnect materials. When oxide interconnect is adopted, the operational temperature is set around 1000°C, whereas a lower temperature is mandatory when metal interconnect is used in order to avoid irreversible degradation.

An intermediate-temperature (400-700°C) SOFC design could remove most of the disadvantages associated with high-temperature operation while maintaining the main benefits of the technology. Such SOFCs could employ much cheaper sealing technologies and robust, inexpensive metals rather than ceramic stack components. At the same time, this new design could still provide reasonably high efficiency and fuel flexibility. But to operate SOFCs at lower temperature, further development is required to improve the oxide ion conductivity of the electrolyte and the reaction rates at the electrodes.

As a result, extensive studies on new solid oxide electrolytes and electrode materials are being pursued with aim at filling the so-called “conductivity gap” between low-temperature proton conductive materials and high-temperature ion-oxide conductive materials: typical new materials are scandia-stabilized zirconia, lanthanum-strontium-gallium-magnesium oxides as electrolyte, lanthanum-strontium ferrite-cobaltite as cathode. Oxide anodes without metals less sensitive to carbon deposition and sulfur impurities than current nickel anodes are also investigated extensively.

In summary, for the stack developments, the following points are important:

For the high-temperature stacks with oxide interconnects, once an inexpensive fabrication process is adopted, the increase in power density will be the most important issue to reduce costs and lead to commercialization.

For the intermediate-temperature stacks with metal interconnects, an inexpensive coating process of ceramic active materials onto metal supports is still needed. In addition, the use of costly elements such as Sc or Ga should be limited.

Although the reliability of tubular SOFCs has been demonstrated, it is important to confirm whether such long stability can also be achieved when metallic interconnect or sealing materials are used since diffusion phenomena leading to degradation are different. Moreover, problems may be exacerbated at lower temperatures.

The majority of SOFC research, development and commercialization are currently done in North America, Europe and Japan.

In the U.S.A., investigations have been well advanced mostly on the basis of tubular stacks.

In Europe, on the other hand, planar SOFCs with metal interconnects have been built to increase the power density compared to the tubular cells and to reduce the materials cost by using Cr-based or ferritic alloys instead of expensive (LaCrO3-based) oxide interconnects.

In Japan, the major achievement has been made in adopting the low cost fabrication methods based on ceramic technology for both tubular and planar stacks. Particularly, the technology associated with oxide interconnect has been well developed.

In terms of government investments, most regions have been supported. In the U.S.A, the Solid State Energy Conversion Alliance (S.E.C.A.) has brought together US government, industry and the scientific community since 1999 to promote the SOFC technology in the 3-10 kW range. The target cost is $400/kW and a minimum of 50000 units produced by the end of 2010. Different companies are involved in the S.E.C.A. programme, e.g., Acumentrics for UPS and CHP systems, Delphi Automotive Systems for automotive and truck APUs, Fuel Cell Energy for small stationary and portable military applications. In Japan, there is also an ambitious residential fuel cell programme including SOFCs subsidized at 50% by the government (cf. Stationary applications section).

Finally, strong support from the U.S. Army drives SOFC research and provides funding for many projects in the portable and off-grid sector. Companies like Protonex Technology Corporation and Nanodynamics have been repeatedly awarded funds in recent years for developing their range of products specially designed for use by soldiers during field operation, for unmanned aerial vehicles (UAVs), or for adapting the products to alternative bio-derived fuels, etc.

In the future, target markets for end-use of SOFC technology remain likely to be stationary power generation, niche transportation and portable applications, with the small stationary sector continuing to be the largest one. There are numerous companies working towards commercialization and a significant number of research programmes aimed at further improving the different SOFC designs. Governments’ funding level is high and interest from military sector will probably continue to drive demand. The main R&D goals include lowering operating temperatures by use of metal interconnects and intermediate temperature systems thanks to new active ceramic materials. Size reduction of the fuel cell unit is also an issue to be addressed especially for portable applications. Most of all, reduction of costs (components, fuel cell stack manufacturing) is crucial to achieve full commercialization.

  1. Glossary

Anode: Electrode through which electric charge flows into a polarized electrical device. The flow of electrons is always from anode to cathode outside of the cell or device, regardless of the cell or device type and operating mode. In a power-consuming device, the anode is positive (+), and in a power-releasing device the anode is negative (-). In electrochemistry, the anode is the electrode where oxidation occurs. In fuel cells, the anode is the (-) terminal.

Auxiliary Power Unit (APU): power-generation system on a vehicle, whose purpose is to deliver electrical energy independently from the main engine and for functions other than propulsion. Different types of APU are found on aircraft, marine vessels as well as on some large ground vehicles. Where the elimination of exhaust emissions or noise is particularly important (such as yachts, camper vans), fuel cells or photovoltaic devices are used as APUs for electricity generation.

Cathode: Electrode through which electric charge flows out of a polarized electrical device. From an electrochemical point of view, positively charged cations move toward the cathode and/or negatively charged anions move away from it to balance the electrons arriving from external circuitry. The cathode is the electrode where reduction occurs. Cathode polarity depends on the device type, and can even vary according to the operating mode. In fuel cells, the anode is the (+) terminal.

Combined Heat and Power (CHP): Family of energy conversion processes involving the simultaneous generation of usable heat and power (usually electricity) in a single process. CHP is a highly efficient way to use both fossil and renewable fuels. In its simplest form, it employs a gas turbine, or a steam engine to drive an alternator, and the resulting electricity can be used either wholly or partially on-site. In a sustainable version of the CHP the engine is replaced by a fuel cell stack. The heat produced during power generation is recovered in a boiler and can be used to raise steam for a number of industrial processes, to provide hot water for space heating, or even cooling. Because CHP systems make extensive use of the heat produced during the electricity generation process, they can achieve overall efficiencies in excess of 70% at the point of use. Unlike conventional power plants, CHP units are sited close to where their energy output is to be used. At home, a micro CHP unit resembling a gas-fired boiler will provide both heat for space and water heating, as a boiler does, but electricity as well to power domestic lights and appliances.

Complementary Metal Oxide Semiconductor (CMOS): Major class of integrated circuits fabricated today in the microelectronics industry. The term CMOS refers both to the type of digital circuitry design, and to the family of processes used to implement it on integrated circuits. CMOS technology uses a combination of p- and n-type Metal Oxide-Semiconductor Field-Effect Transistors (MOSFETs) to implement logic gates and other digital circuits found in computers, telecommunications and signal processing equipment. CMOS circuitry dissipates less power when static, and is denser than other implementations. For these reasons, the vast majority of modern integrated circuit manufacturing is based on CMOS processes. Typical commercial CMOS products are integrated circuits composed of millions (or hundreds of millions) of transistors on a rectangular piece of silicon between 0.1 and 4 cm² area called chip.

Electrode: Electrical conductor used to make contact with a nonmetallic part of a circuit (e.g. semiconductor, electrolyte or vacuum). An electrode in an electrochemical cell is referred to as an anode or a cathode. The anode is now defined as the electrode at which electrons leave the cell and oxidation occurs, and the cathode as the electrode at which electrons enter the cell and reduction occurs. Each electrode may become either the anode or the cathode depending on the voltage applied to the cell.

Electrolyte: Any material containing free ions that behaves as an electrically conductive medium. Because they generally consist of ions in solution, electrolytes are also known as ionic solutions, but molten electrolytes and solid electrolytes are also possible. Proton conductors are a special class of solid electrolytes, where hydrogen ions act as charge carriers. Finally, an ionomer (e.g. Nafion®) is a polymer that comprises repeat units of both electrically neutral repeating units and a fraction of ionized units (usually no more than 15 percent). Ionomers have unique physical properties including electrical conductivity.

Eutectic (mixture): Mixture of two solids at such proportions that their melting point has a minimum and that the constituents crystallize simultaneously at this temperature from molten liquid solution.

Internal Combustion Engine (ICE): Engine in which the combustion of the fuel occurs in a chamber placed inside and integral to the engine. It is the volume expansion of the high-temperature and pressurized gases produced by the combustion process that creates the mechanical force necessary to drive the movable component of the engine (piston, turbine blade…).

Light duty vehicle: In the U.S. legislation, the light duty vehicle (LDV) category includes all vehicles of less then 8,500 lbs (3,859 kg), and is further divided into passenger cars and light-duty trucks. In Europe, vehicles of less than 3,500 kg belong to the light sector and vehicles with more than 3,500 kg are referred to as heavy duty vehicles (HDV). The light duty vehicle technology is derived from passenger car developments, though the higher vehicle weight requires more engine power.

(Hydrogen) reformer: Chemical reactor that extracts hydrogen from other fuels, typically methanol, natural gas or gasoline. The most common method for producing hydrogen for different industrial and commercial applications is steam reforming. A high temperatures (700-1100°C) and in the presence of a metal-based catalyst such as nickel, steam reacts reversibly with methane to yield hydrogen and carbon monoxide. Steam reforming of liquid hydrocarbons has been regarded as an interesting way to provide fuel for fuel cells in the near to medium term and mitigate the current distribution problems with fuel cell vehicles (FCVs) due to the lack of a real hydrogen infrastructure. Basically, a fuel tank and an external reformer unit would replace the pressurized hydrogen tank. In an improved design, steam reforming of the fuel takes place directly on the anode surface producing the hydrogen molecules prior to their use in the fuel cell reaction. This solution, still under development, is called internal reforming and in that case, the reformer is the anode.

Uninterruptible Power Supply (UPS): Backup system that provides emergency power from a separate source when utility power is not available. Unlike an auxiliary power unit, it provides instant protection from a momentary power interruption. A UPS can be used to provide uninterrupted power to equipment, typically for 5-15 minutes until an auxiliary power unit (APU) can be turned on or utility power from the electric grid is restored. While not limited to safeguarding any particular type of equipment, a UPS is typically used to protect electrical equipment, telecommunication and data centers, hospitals, etc. where a power outage could cause injuries, fatalities, serious business disruption or data loss. UPS units come in sizes ranging from units which will backup a single computer to units which will power data centers or buildings (several megawatts).