Ammonia-Based Fuel Cells

Back to Contents

    Given the decline in the availability of the fossil fuels that we use in our vehicles, even as world-wide demand increases, and given the problems associated with those fuels (such as flammability, toxicity, climate change, etc.), we should want something better. Many engineers have suggested the use of hydrogen as a fuel, but hydrogen has its own drawbacks, making it impractical as a vehicle fuel (with the obvious exception of rockets).

    The primary drawback comes from the need to contain a useful quantity of hydrogen in a small volume. Such containment requires either refrigerating the hydrogen until it achieves the liquid state (at -423 F, about 20 centigrade degrees above absolute zero) or compressing it to 5000 psi (340 atm.). Both processes involve considerable expense. Further, the small hydrogen molecules can leak through holes and cracks too small for other molecules and they can diffuse into the crystalline structure of metals and thereby embrittle them.

    No, for all of its advantages, hydrogen itself does not provide us with a useful fuel. However, we might consider using a chemical rich in hydrogen as a means of storing hydrogen for use as a fuel. In particular we might consider using ammonia, a chemical whose molecules each consist of one nitrogen atom and three hydrogen atoms. As a fuel ammonia has several advantages over hydrocarbon fuels.

    First, ammonia is such a common industrial chemical that producers use pipelines to distribute it to the large agricultural regions where it is used as the basis for fertilizer. Producers also transport it and contain it in tanks under modest pressure, in a manner similar to the containment and transport of propane. Thus we already have a mature technology in place for producing, transporting and storing ammonia.

    Second, ammonia does have some toxicity when inhaled; air containing a 1% concentration of ammonia inhaled for 1 hour has a 1% fatality risk. However, ammonia inhalation is easily avoided, because the gas has a readily detected odor and even a rather pungent feel when it enters the nose. Furthermore, ammonia vapor is lighter than air, so it will rapidly dissipate in a spill, thereby minimizing the risk of poisoning anyone.

    And third, it differs from gasoline or diesel fuel in that it does not readily catch fire in an accident: it has an ignition temperature of 650 Celsius. If no parts of an ammonia-based power system reach that temperature, then any ammonia spilled in an accident will simply dissipate.

    Now we must ask what kind of power system will make the best use of ammonia as fuel. Fuel cells provide a good alternative to an internal combustion engine. NASA has successfully used hydrogen-oxygen fuel cells to provide electricity for its spacecraft and some experimenters have used such fuel cells to power automobiles. The main difficulty in using such fuel cells lies in the requirement to store enough hydrogen to make the cell practical. To avoid that difficulty we need an ammonia-oxygen fuel cell, a fuel cell that exploits the relative ease of storing enough ammonia to make the cell practical for most applications. More specifically, we want a direct ammonia fuel cell, one that requires no preprocessing of the ammonia before it enters the cell.

    In a direct ammonia fuel cell ammonia passes directly into the cell without first being cracked into nitrogen and hydrogen; the cracking occurs in the cell itself. Solid oxide fuel cells provide a good example of what we want. Such a fuel cell consists of proton-conducting ceramic electrolytes or molten salt electrolytes sandwiched between suitable anodes and cathodes. The fuel cell operates at a high temperature that cracks the ammonia into nitrogen and hydrogen and then produces electricity with high efficiency. On the cathode side of the cell air enters and oxygen gets adsorbed onto the cathode material, gaining electrons as it does so. Ammonia enters on the anode side and gets cracked into nitrogen and hydrogen. The hydrogen gives up its electrons and the protons diffuse through the electrolyte and react with the oxygen to form water, the cell’s waste product. The excess ammonia and the water then exit the cell as exhaust while the unused gases (i.e. nitrogen) also pass out of the cell as exhaust. The electrons stripped from the hydrogen pass from the anode into an electric circuit where they do work before they go to the cathode to complete the chemical reaction that forms water.

    United States Patent 7,157,166 describes a fuel cell that directly utilizes ammonia as a fuel without prior treatment to decompose the ammonia and remove traces of undecomposed ammonia. This fuel cell produces electrical energy based on the chemical reaction sketched above. To understand the need for such a fuel cell, consider the following facts.

    A polymer exchange membrane hydrogen-oxygen fuel cell gives us an example of a fuel cell used in spacecraft, such as those of Project Apollo, and then developed further for widespread commercial application. However, such a fuel cell needs highly pure hydrogen to avoid the poisoning that would degrade the cell’s performance. Further the cell needs sources of hydrogen with high density, especially for mobile applications. Both compressing and liquefying hydrogen to provide reasonable densities offer substantial technological difficulties and use up a sizeable fraction (up to about 30%) of the stored hydrogen's energy. Chemical storage of hydrogen in metal hydrides can achieve storage of only a few percent of hydrogen by weight, another fairly inefficient method of fuel storage. But ammonia contains approximately 17 percent hydrogen by weight and it liquifies relatively easily or dissolves readily in cold water. Ammonia therefore provides an attractive source of hydrogen for polymer exchange membrane hydrogen-oxygen fuel cells.

    The hydrogen comes from the ammonia in an endothermic reaction carried out in a device separate from the fuel cell. Ammonia decomposition reactors catalytically decompose ammonia into N2+3H2. In theory the reaction consumes approximately 13% of the energy in the ammonia. However, this reaction requires high temperatures of 400 - 1000° Celsius. Polymer exchange membrane hydrogen-oxygen fuel cells typically run at 80° Celsius. The necessity to cool the products of the ammonia decomposition leads to inefficiencies, which, in practice, can use up to 40% of the energy in the ammonia, a loss significantly higher than the 13% theoretical loss noted above. Further, ammonia decomposition absorbs heat, so energy must continually flow into the decomposition reactor to keep it at the required temperature. That energy flow plus the need to cool the gases coming out of the decomposition reactor results in a loss in efficiency for the fuel processing/fuel cell system as a whole.

    Moreover, a polymer exchange membrane fuel cell cannot use ammonia directly because the ammonia would not decompose at the temperatures at which such fuel cells operate and the undecomposed ammonia would poison the fuel cell’s catalyst. Even the N2+3H2 mixture coming from a decomposition reactor cannot be fed into the anode of a hydrogen-oxygen fuel cell, because any residual ammonia will poison the anode of a polymer exchange membrane fuel cell. Note that residual ammonia results from an incomplete decomposition reaction or, depending on the decomposition temperature, from the decomposition equilibrium, in which the reactor recreates ammonia at the same rate at which it decomposes it.

    United States Patents 5,055,282 and 5,976,723 disclose a method for cracking ammonia into hydrogen and nitrogen in a decomposition reactor.

    The method consists of exposing ammonia to a suitable cracking catalyst under conditions effective to produce nitrogen and hydrogen. In this case the cracking catalyst consists of an alloy of zirconium, titanium, and aluminum doped with two elements from the group consisting of chromium, manganese, iron, cobalt, and nickel.

    United States Patent Application Number 2002/0021995 discloses an alternative apparatus and method for decomposing ammonia. In this apparatus a fluid containing ammonia passes in contact with a membrane that consists of a homogeneous mixture of a ceramic and a first metal. The desired ceramic consists of cerates, zirconates, or stannates of beryllium, magnesium, calcium, strontium, barium, or radium or some suitable mixture thereof doped with one or more of calcium, yttrium, ytterbium, indium, neodymium, or gadolinium. The first metal that we mix with the ceramic consists of platinum, silver, palladium, iron, cobalt, chromium, manganese, vanadium, nickel, gold, copper, rhenium, ruthenium, osmium, iridium, or suitable mixtures thereof. The membrane thus made has a catalytic metal on the side in contact with the fluid containing ammonia, which catalyst drives the decomposition of ammonia into nitrogen and hydrogen. When the hydrogen then contacts the membrane it breaks up into ions (protons), which pass through the membrane, thereby driving the decomposition of the ammonia toward completion. We now need to incorporate that device into a fuel cell for generating electrical energy from ammonia and air.

    The invention described in United States Patent 7,157,166 satisfies the needs laid out above, describing a novel ammonia fuel cell for directly generating electrical energy from a simple chemical reaction. A preferred version of the fuel cell itself consists of a three-component plate. On the ammonia side we have a decomposition catalyst (which causes NH3 to decompose to N2+3H2) in contact with a high temperature proton conducting membrane (through which the protons diffuse to a catalytic cathode where they combine with oxygen to form water). In contact with that decomposition catalyst we have a catalytic anode, which dissociates and ionizes H2 into protons and electrons. Finally, the fuel cell includes a catalytic cathode (on the oxidizer side) on which protons, electrons and oxygen come together to form water.

    The complete fuel cell also includes an external circuit extending from the catalytic anode to the catalytic cathode, the circuit consisting of a wire through which the electrons flow through an external load. It also needs an ammonia source for introducing ammonia into the fuel cell, a gas exit for the nitrogen left over from the decomposition of the ammonia, an oxygen source, and a means of discarding the oxygen-depleted air and the water produced at the cathode.

    Decomposition of ammonia (on a proton exchange membrane that operates at temperatures above 400° C.) enables direct use of ammonia in a fuel cell. The process yields nitrogen and hydrogen atoms adsorbed onto the membrane. Pairs of adsorbed nitrogen atoms combine to form nitrogen molecules, which then desorb to form nitrogen gas that exits the fuel cell. Meanwhile the adsorbed hydrogen atoms dissociate into electrons and protons. The electrons flow to an external circuit, where they do work on an external load, and the protons diffuse through the ceramic membrane to a catalytic cathode. At the cathode oxygen, either from air or an oxygen source, reacts with the diffusing protons and the electrons from the external circuit to form water.

    Since we can obtain oxygen from air, we need only worry about storing the ammonia for use in our fuel cell. We have available to us several methods of storing the ammonia prior to its use in the fuel cell. Ammonia exists as a liquid at room temperature at a pressure of less than10 bar (106 Pascals or nearly 10 atmospheres), so we only need a fuel tank that can contain a pressure of about 150 pounds per square inch. Or, because ammonia liquifies at standard atmospheric pressure at a temperature of 33.35 Celsius degrees below zero (about 28 Fahrenheit degrees below zero) liquid ammonia can be stored in a moderately insulated fuel tank. At that temperature ammonia has a density of 681.9 kilograms/cubic meter, just about the same density as gasoline. Yet another method of storing ammonia involves using absorbents and chemical compounds containing coordinated ammonia. Possible absorbents include activated carbon, silica gel, and zeolites, or, most simply, cold water. Cold water (20 degrees Celsius, 78 degrees Fahrenheit) dissolves 899 kilograms/cubic meter while hot water (near the boiling point at 100 degrees Celsius) dissolves 74 kilograms/cubic meter. Because heating the water drives off the ammonia for use in the fuel cell, we need only arrange for a thin stream of water to run over a hot plate, then pass through a convector to cool off before going back into the fuel tank.

More Details

    In further detail, we add to our description of the fuel cell the following:

    Fuel cells provide electrical energy by electrochemically oxidizing fuels without combustion. Electrochemical oxidation avoids the inefficiencies associated with heat engines and their approximations to the Carnot cycle. We thus have a direct ammonia fuel cell comprising a catalytic anode, a proton conducting ceramic membrane, and a catalytic oxygen cathode. The overall reaction consists of the oxidation of ammonia’s hydrogen component. The components of the fuel cell consist of the following possibilities.


    We can select the catalytic materials for the anode from any of the formulations designed for ammonia decomposition, such as those described in United States Patents 5,055,282 and 5,976,723. Typically, these catalysts comprise mixtures of early transition metals together with various group VIIIb elements. Reasonable reaction rates have been achieved with those materials in the range between 400° Celsius and 1000° Celsius. The anode must also function as a current collector in order to supply electrons to an external circuit. This requirement can be met by mixing the catalytic material with a conducting metal that is also usually a group VIIIb element. The production of such structures employs techniques analogous to those used in making solid oxide fuel cells.

    Some examples of possible anode materials include Raney nickel, Nickel-nickel oxide composite, Platinum paste, porous Platinum and mixtures thereof.

    To facilitate transfer of protons from the catalytic anode to the high temperature proton conducting membrane, the membrane material, such as described below, may be incorporated into the anode as a mixture.

Proton-Conducting Membrane:

    In the fuel cell we have the anode supported on a proton conducting ceramic membrane. This membrane can be selected from any of a group of ceramics developed for high temperature proton conduction. Several recent examples include those developed for high temperature hydrogen-oxygen fuel cells. Examples consist of cerates, zirconates, and stannates of group IIA metals and doped perovskites.

    As the name implies, proton conducting membranes conduct protons in much the same way in which copper conducts electrons. We have available to us a variety of protonic conductors, which differ in conductivity (ideally high); differ in the variation in conductivity with changes in temperature; differ in their electron and oxide ion conductivities (ideally low); and differ in their chemical stability (ideally high). The desired operating temperature and the current/voltage characteristics of the system dictate the choice of a specific proton-conducting membrane. The membrane can be produced by standard techniques e.g. starting from oxides, carbonates, or hydroxides with milling, sintering, and pressing. Solgel techniques provide an alternative. The structure can be a homogeneous monolithic plate, disk or tube. It can also be a composite with a electrochemically inert ceramic (alumina, zirconia) powder or a porous foam used to improve strength. Another possibility is a layered structure composed of two or more different electrolyte materials. In this case thin layers of more chemically inert material can be used on the outside of a less stable material. The thickness depends on conductivity and desired operating current voltage characteristics. For good performance the thickness (T in cm) and the conductivity (sigma in Ohm/cm) are related as T/sigma=0.15 Ohm/cm2.

    A few examples from many possibilities for electrolytes include:

1. Rubidium nitrate/aluminum oxide composite,

2. Lithium oxide/aluminum oxide composite,

3. Doped strontium-cerium-scandium oxide, or

4. Doped calcium-zirconium-indium oxide.

    Generally, the suitability of the proton conductivity ceramic component candidates will depend on the stability requirements of the fabricated homogeneous composite.


    We have a cathode analogous to those used in solid oxide fuel cells or high temperature hydrogen-oxygen fuel cells. This provides the oxidizer for the chemical reaction that drives the fuel cell. Oxygen can come to the cathode from air or from a pressurized oxygen source. Some examples of cathodes comprise:

1. Nickel oxide/Silver oxide mixture (NiO--AgO),

2. Carbon paste,

3. Silver paste,

4. Platinum paste, and

5. porous Platinum and mixtures thereof.

    To facilitate transfer of protons from the proton conducting membrane to the cathode, the proton conducting ceramic comprising the membrane may be a component of the cathode as a mixture.

    The development of high temperature proton conducting membranes has made use of ammonia directly as a fuel feasible. Using that technology, we can make the physical configuration of the anode/membrane/cathode structure either tubular or planar. For convenience the planar form gives us the better option because we can easily stack the plates, as we do in lead-acid batteries, to increase the output voltage and/or current, with the ammonia and air inlets alternating.

    The fuel cell has an entrance for ammonia coming from the fuel tank. The ammonia flows directly to a combined decomposition catalyst and catalytic anode, where the ammonia decomposes into nitrogen, which leaves the fuel cell through the exit, and hydrogen. The hydrogen leaves the decomposition catalyst and interacts with the catalytic anode, where the hydrogen molecules dissociate into hydrogen atoms that then get ionized. The combined catalyst and anode ionizes the adsorbed hydrogen atoms into hydrogen ions (i.e. protons) and electrons.

    The catalytic anode is supported by a high temperature proton conducting ceramic membrane. The protons diffuse through the membrane toward a catalytic cathode. The electrons flow to an external circuit. The external circuit is connected between the catalytic anode and the catalytic cathode. The diffused protons from the catalytic anode, electrons from the external circuit and oxygen, from air as the oxygen source, react at the catalytic cathode to form water.

    The external circuit includes a load to which the fuel cell applies power. In the fuel cell the potential is higher at the catalytic anode than it is at the catalytic cathode. The difference in the potential is equivalent to the potential of the load. At 500° C., the enthalpy change for the partial oxidation reaction is -316.0 kiloJoules, the entropy change is 34.74 Joules/Kelvin, and the free energy change is -342.9 kiloJoules. For this reaction 3 electrons flow through the external circuit for each ammonia molecule decomposed at the anode and so the open circuit voltage is 1.18 Volts. The fuel cell produces nitrogen and water as ecologically benign waste. We can count this as a further advantage of the ammonia fuel cell.

Producing Ammonia

    Of course, to make the fuel cell feasible we need a cheap, efficient means to produce large quantities of ammonia. In one method the reverse of the process used in the fuel cell can manufacture ammonia from streams of nitrogen separated from air and hydrogen created by dissociation powered by high-temperature process heat and electric power from LFTR (liquid fluoride thorium reactor) electric power generators.

    Molten salts sandwiched between suitable anodes and cathodes at 673 Kelvin can produce ammonia from hydrogen diffusing through the anode at one atmosphere and nitrogen diffusing through the cathode at one atmosphere. The efficiency of the operation, with electrons driven from anode to cathode by the power supply, can reach 72%. At the anode we have

3H2+2N3- = 2NH3+6e-

At the cathode we have

N2+6e- = 2N3-

So in total we have

3H2+N2 = 2NH3.

    In another method we can use Solid State Ammonia Synthesis. In this method the ammonia synthesis unit consists of two main parts, the air separation unit (ASU) and the solid state ammonia synthesis unit (SSAS). The entire process can be run by electricity from any source (e.g. such as thorium-based nuclear power). Air goes into the inlet of the ASU and pure nitrogen comes out of one outlet and oxygen and other gases come out the other outlet. Water (for its hydrogen) and nitrogen go into the SSAS and ammonia comes out one outlet and oxygen comes out the other. The chemical synthesis goes as

6H2O + 2N2 = 3O2 + 4NH3.

    The hydrogen electrolysis or thermal dissociation step can be eliminated via solid-state ammonia synthesis, operating like a solid-oxide fuel cell, but in reverse. It similarly has a ceramic proton conducting membrane. It has the advantage that there is never any separated explosive hydrogen gas and it operates at low pressure. Nitrogen is obtained from the ASU (air separation unit). Water supplies the hydrogen. The ceramic membranes are tubes and the SSAS (solid state ammonia synthesis) can be scaled up by using more tubes. The SSAS process is safer and cheaper than the standard Haber-Bosch process. The key cost advantage is that SSAS is projected to make ammonia at 6800 kWh per ton. With factory reactor production, LFTR electric power is projected to cost $0.03/kWh, leading to ammonia costs of about $200 per ton. This is half the cost of ammonia produced today from natural gas, and it avoids the release of carbon dioxide in that widespread manufacturing process.

Ammonia fuel cost relative to gasoline

    We calculate the cost to provide fuel as follows:

Gasoline; $2.70 for crude + $0.90 for taxes et al. + $0.40 refining = $4.00/gal. Gasoline density = 719.7 kg/cubic meter or 6.073 lb/gal.

Ammonia; $0.90 for ammonia + $0.90 for taxes et al. + $0.40 refining =$2.20/gal. Liquid Ammonia density (-33.35 Celsius; -28 Fahrenheit) = 681.9 kg/cubic meter.

We use those costs to calculate the energy costs for California in 2011 as:

Fuel         Heat of Combustion         Price         Energy Cost

Ammonia     22 MegaJoule/kg         $0.20/kg       $0.01/Joule

Gasoline     132 MegaJoule/kg         $4.00/gal      $0.03/Joule.

That looks like gasoline has the advantage of energy density: it takes fewer gallons of gasoline to produce a given amount of energy. But if we take into account the inefficiency of the gasoline engine, ammonia’s disadvantage diminishes. In a typical automobile only 20% of the energy in the gasoline gets turned into kinetic energy of the car while the rest gets exhausted as waste heat. In a fuel cell driven car 75% of the energy in the fuel gets turned into kinetic energy. So effectively we get 26.4 Megajoules per kilogram from gasoline and 16.5 Megajoules per kilogram from ammonia.

    We thus need 1.69 times as many gallons of liquid ammonia to get the same benefit of a given quantity of gasoline. A car with a 20-gallon gas tank would thus need a 34-gallon ammonia tank.

    How might this lower energy cost translate into vehicle fuel costs? Most of the cost is for the crude petroleum that provides the energy content of the gasoline. The refining costs are only about 10%, even though refineries are complex, expensive investments. We don’t really know the cost of SSAS chemical plants, but simply assume that the talented chemical engineers who built petroleum refineries can build similarly large ammonia production plants at about the same cost.

    In summary, ammonia liquid fuel can replace petroleum liquid fuels at less cost while also eliminating CO2 emissions. With such a system in place we can begin to replace fossil-fuel based power systems with ammonia-based fuel cells. Trains offer a good first start for such an application, getting the standard diesel-electric locomotives replaced with ammonia-electric locomotives. Cars and trucks, as well as the carriers of autorail systems (see "California AutoRail on this website) provide another opportunity to replace hydrocarbon-burning engines. Boats and ships offer yet more opportunities. Aircraft seem unlikely to use ammonia-based power: even though NASA used liquid ammonia as fuel in its X-15 rocketplane, ammonia doesn’t have the energy density needed by modern long-range aircraft.


Back to Contents