VoxSolaris: The Voice of the Sun
Fuel Cells

The principle

Fuel cells convert chemical energy directly to electricity by means of an anode, a cathode and an electrolyte, just like battery cells do. The reactants are not stored in-situ and are instead pumped into the cell in the same was as they are in a flow battery. But where a battery is a closed system with a fixed quantities of reactants whose reaction may or may not be reversible, the fuel cell is one-way and open. The reactants are pumped in from an external source and the products exhauted to an external sink. The fuel cell is further distinguished from the battery in the general case by the choice of reactants. Fuel cells with few exceptions if any, consume hydrogen or compounds containing hydrogen such as methanol or ammonia (i.e. combustable fuels). The hydrogen is reacted with oxygen to produce electricity in a process that is essentially the reverse of water electrolysis.

The fuel cell has gathered considerable mystique which is helpful to debunk. There is a notion that they are clean 'zero emission' devices but this is only true when the supplied fuel is devoid of carbon. A fuel cell is the functional equivelent of an IC engine combined with a DC generator which would also be zero emission if the supplied fuel is devoid of carbon. Both transfer the environmental burdens to the fuel source so an IC engine runing on bio-deisel is more carbon neutral than a fuel cell running on methane drilled from the ground.

The notion that fuel cells are more efficient holds more truth but not for all types. Interestingly, the fuel cells that are efficient are the ones running on fuels devoid of carbon. And here the efficiency is impressive. For fuel cells reacting pure hydrogen and oxygen, the process has a maximum theorectical chemical to electrical energy efficiency of about 83 percent with real fuel cells in the feild achieving of the order of 60 percent. This significantly exceeds the 60 percent theoretical and 30 to 40 percent real efficiencies provided by the typical IC engines. But it is worth noting that large diesels such as those powering ships are also more efficent than typical IC engines giving real efficiencies of better than 50 percent. It should also be noted that an IC engine optimized to run on pure hydrogen and oxygen would also be much more efficient.

The fuel cell in and of itself, is certainly much simpler than an IC engine and it has no moving parts. The fuel cell is not by any means a new idea though. Sir William Grove first demonstrated the conversion of hydrogen to electricity using an acid-electrolyte fuel cell in 1839. Unfortunately almost 180 years later, there are still significant challenges that lay between the fuel cell and commercial viability and hence widespread adoption.

Different types

Fuel cells are generally categorized by their electrolyte as it is the characteristics of the electrolyte that determine the optimal operating temperature of the fuel cell and the choices of fuel that can be used to generate electricity. Each type comes with its particular set of pros and cons. The lower temperature fuel cells are easier to construct but need expensive platignum catalysts. Higher temperature fuel cells are harder to construct but do not need platignum catalysts and offer the important advantage of 'cogeneration' in which waste heat can be used in a heat engine and thus contribute to higher overall electrical efficiency.

Fuel Cell Type Electrolyte Anode gas Cathode gas Temperature (C) Efficiency (%)
Proton Exchange Membrane (PEM) solid polymer membrane hydrogen air/oxygen ~75 35 - 60
Alkaline (AFC) potassium hydroxide hydrogen pure oxygen ~80 50 - 70
Direct Methanol (DMFC) solid polymer membrane aquous methanol solution air/oxygen ~75 35 - 40
Phosphoric Acid (PAFC) Phosphorous hydrogen air/oxygen 210 35 - 50
Molten Carbonate (MCFC) Alkali- Carbonates hydrogen, methane air/oxygen 650 40 - 55
Solid Oxide (SOFC) Ceramic Oxide hydrogen, methane air/oxygen 800 - 1000 45 - 60

Polymer Electolyte Membrane Fuel Cell (PEMFC)

PEMFCs or PEMs as they are often known have a very low operating temperature allowing rapid startup. They are also very responsive to varying loads making them potentially suitable for a very wide variety of applications, subject to the availability of hydrogen. PEM fuel cells use a solid polymer electrolyte, typically Nafion, through which protons (hydrogen ions) pass from anode to cathode. The electrolyte is often regered to as a 'proton exchange membrane'. The low operating temperature comes at a very high price though. As reactions at low temperatures are very slow (reaction speeds tend to rise exponentially with temperature), a platinum catalyst at the electrodes is required. Historically, the platinum has made these devices prohibitively expensive, but new application techniques have thinned the platinum layer, reducing the capital costs. In spite of these advances, we are still looking at somewhere in the region of $2,000 to 2,500 per kW. This is far to high for cars where a hydrogen supply is feasible although problematic. It is a lot more acceptable in a laptop but here the hydrogen supply is rather less feasible. Consequently the PEMFC has failed to enthrall the market although a number of car makers have built prototypes cars, presumably to boost their 'green' credentials. If PEMFCs are ever to be viable, a cheaper catalyst will need to be found or a massive platignum deposit will need to be discovered. The verdict is simple. PEMs are no match for lithium ion batteries so why bother?

Alkaline Fuel Cell (AFC)

The AFC was was developed by NASA to provide power on board spacecraft. The cells use graphite electrodes and the electrolyte is potassium hydroxide. AFCs are very efficient but can run only on very pure hydrogen and oxygen as the electrolyte is poisoned by traces of impurites such as carbon dioxide. Although the operating temperature is higher than for the PEMFC, it is still far too low to avoid the expensive platignum catalysts and techniques to reduce the quantities of platignum are less advanced for this type of cell. The cells are quite tempramental and are less suited to varying loads than PEMFCs. Taken together, these factors make the AFC considerably less attractive then the PEMFC and as a result, AFCs will find only niche markets.

Direct Methanol Fuel Cell (DMFC)

DMFC is similar to the PEMFC in that the electrolyte is a polymer and the charge carrier is the hydrogen ion (proton). However it is a much more convienent configuration likely to find use in laptops and mobile phones. Instead of using hydrogen gas as the fuel it uses methanol (CH3OH), a liquid at ordinary temperatures and pressures. The methanol is oxidized in the presence of water at the anode generating CO2, hydrogen ions and the electrons that travel through the external circuit. The hydrogen ions travel through the electrolyte and react with oxygen from the air and the electrons from the external circuit to form water at the anode completing the circuit.

Anode Reaction: CH3OH + H2O => CO2 + 6H+ + 6e-
Cathode Reaction: 3/2 O2 + 6 H+ + 6e- => 3 H2O
Overall Cell Reaction: CH3OH + 3/2 O2 => CO2 + 2 H2O

The DMFC is not particulaly efficient, largely because energy released by the oxidation of methanol at the anode is not put to use. Another drawback is that this low temperature oxidation requires a more active catalyst, which typically means a larger quantity of platinum than in PEMFCs. Methanol is also highly toxic and this has led some companies to develop a direct ethanol fuel cell (DEFC). This currently has even lower efficiency and power density than the DMFC although this is expected to improve. Our verdict would again be that this technology is inferior to lithium ion. Its market is small devices such as the mobile phone but the prevelence of such devices means there are chargers everywhere. And in situations where there are no chargers, such as on safari, rather than carry a bottle of methanol which will eventually run out, carry a solar mobile charger that won't.

Phosphoric Acid Fuel Cell (PAFC)

This type has been the most developed commercially. The cells use concentrated phosphoric acid as the electrolyte and platignum alloy electrodes 'printed' in a matrix form on a silicon carbide substrate. They are moderately efficient and are reliable with low maintenance and low suseptibility to poisoning. The hydrogen fuel can be less pure and the cell can run well on air rather than oxygen. The construction of the cell makes recycling of the platinum electrodes easier. Again a poor verdict. At the end of the day it is pretty simple. The dependence on platignum is a clear barrier to the mass market.

Molten Carbonate Fuel Cell (MCFC)

These have an electrolyte consisting of a mix of sodium and potasium carbonates held in a ceramic matrix of LiAlO2. The operating temperature is some 650 to 800°C in order to keep the electrolyte highly conductive. The carbonates are molten at this temperature. The anode and cathode are substrates of mickel and nickel oxide respectively. The high operating temperature places severe demands on cell components which are not helped by the molten electrolyte. The temperature rules out use in mobile applications and it also means the cell cannot be started quickly. However there are major advantages arising from the high temperature. The expensive platinum is not required as the reactions proceed rapidly at high temperature. The high grade waste heat facilitates cogeneration to boost overall efficiency and the cell can run on fuels such as methane without a separate reformer. The reactions in the cell are as follows:-

Anode reaction: H2 + CO3= = H2O + CO2 + 2e-
Cathode reaction: l/2O2 + CO2 + 2e- = CO3
Overall reaction: H2 + l/2O2 + CO2 (cathode) = H2O + CO2 (anode)

Note that CO is not directly used by the electrochemical oxidation, but produces additional hydrogen by the water gas shift reaction: CO + H2O = H2 + CO2. Typically the CO2 generated at the anode is recycled to the cathode where it is consumed. This requires additional equipment to either transfer the CO2 from the anode exit gas to the cathode inlet gas or produce CO2 by combustion of the anode exhaust gas and mixed with the cathode inlet gas. Fuel such as methane is reformed within the cell in the presence of a suitable catalyst to form H2 and CO by the reaction: CH4 + H2O = 3H2 + CO. Such fuels would need to be free of any sulphur though as this poisons the catalysts.

Solid Oxide Fuel Cell (SOFC)

These have a solid oxide electolyte (usually zirconium oxide ZrO2), a nickel zirconium oxide anode and a lanthium manganese oxide cathode. These oxides are not cheap but are much cheaper and adundant than platinum! The operating temperature is of the order of 800 to 1000°C. The electrolyte conducts oxygen ions rather than protons. The cell is more reliable than the MCFC as the electolyte remains solid at all times and it shares the advantages of the MCFC, namely cogeneration, no platinum and when run on hydrocarbons or ammonia, no separate reformer. The reactions in the cell are as follows:-

Anode reaction: 1/2 O2 + 2e- = O
Cathode reaction: H2 + 1/2O= = H2O + 2e-
Overall reaction: l/2O2 + H2 = H20

The cells can produce around 500 mW/cm2 and by use of using thin-film ceramics, high power densities can be obtained. These devices have advanced considerably in recent years and are now approaching efficiency and cost targets ($400/kW) for use in CHP systems. Use in cars would seem some way off though as this would need costs to come down to $100/kW. Both SOFC and MCFC suffer from thermal cycling problems but the SOFC are worse at this. For use in a vehicle, either this problem would have to be solved by finding more suitable materials (not particularly likely), better fabrication techniques (maybe). Our view on this is that it is better to deal with poor thermal cycling by simply keeping the cell at its operating temperature. Use of a multi layer vacuum vessel would strongly mitigate thermal loss.

Material Properties Limtitations

Many of the materials used in SOFCs and PEFCs today have changed little in decades and most of the fuel cell advances have come from fabrication techniques. The reason this is so is because the materials used have properties that are very unusual. Platinum is the most obvious example of these. There simply is no other material apart from the equally rare and expensive palladium, that posseses the quantum properties to make hydrogen and oxygen combine rapidly at temperatures well below their ignition temperature. The materials we do have are frought with problems. They have to be produced to high levels of purity or in the case of alloys, to very accurate proportions of the component metals. For the same reason, the materials are generally very sensitive to poisoning. Even when poisoning is eliminated by use of very pure hydrogen and oxygen (an expensive issue in its own right), the cell materials degrade.

Our conclusions

We are not convinced by low temperature fuel cells thusfar and we note that in all the potential applications we could think of, they are trumped by lithium ion. Some research into such fuel cells is ongoing but it is understandably harder to justify and fund than it was before lithium ion became established. It is difficult to see how low temperature fuel cells can break out of this vicious cycle although they are likely to benefit from research into better battery electrodes.

The outlook for high temperature fuel cells however is much more promising. Although the electrolytes are a weak point they are cheap and easily replaced. We forsee scenarios that would mitigate electrolyte weakness whilst benefitting from an operational efficiency that has the edge on that available from the IC engine the fuel cell would replace. The key to mitigating electrolyte weakness is for the cell to retain operating temperature. This limits the range of road transport applications as at present, most vehicles spend most of their time parked making thermal losses harder to justify. This is a problem for the Sodium Sulphur Battery with an operating temperature of 350°C and is a worse problem with fuel cells whose operating temperatures are closer to 800°C. There are some potential road transport scenarios where vehicle downtime is limited such as long distance frieght operators and the advent of autonomous vehicles is likely to significantly increase these scenarios. Another possible use is in aircraft.