Deregulation of the electric power industry and more stringent emission controls are stimulating investment into fuel-cell systems, which are virtually free from noxious emissions of nitrogen oxides normally associated with burning fossil fuels. They can convert chemical energy directly into electricity with greater efficiency than most other devices, which both conserves fuel resources and reduces CO₂emissions. Although hydrogen is considered the ideal fuel for many energy-conversion systems, its widespread use is dependent on technological breakthroughs in its cost and storage. So it is best to assume that in the immediate future, except for a few niche markets, fuel cells will have to use hydrocarbons (such as methane) or alcohols (such as methanol) as fuel. In this context, it is timely to examine the merits of direct electrochemical oxidation of methane in a fuel cell reported by Murray et al.¹ on page 649of this issue.

Using methane as the primary fuel favours devices that operate at high temperatures², such as molten carbonate and solid-oxide fuel cells (SOFC). These are able to electrochemically oxidize a mixture of hydrogen and carbon monoxide, which is generated internally from the natural gas (methane) within the fuel stack. The development of SOFC technology has been principally directed towards large-scale power generation, and early commercial devices operating on natural gas at around 900 °C are expected to produce 1-3 megawatts (MW) of power³. Such fuel cells use anodes that contain nickel to catalyse the conversion of methane to H₂and CO, but large amounts of steam have to be added to prevent carbon deposition on the nickel-ceramic anodes, which severely degrades their performance. The steam-conversion reaction is endothermic and has to be sustained by significant thermal energy, which is available in the high-temperature SOFCs. In contrast, low-temperature fuel cells need an external processor to produce H₂and remove CO, which not only reduces the efficiency of the system but also increases its complexity and cost (Fig. 1).

Figure 1: Fuel processing reactions (reforming of methane to hydrogen fuel) and their influence on the complexity and efficiency of different types of fuel cell.

The intermediate-temperature solid-oxide fuel cells (SOFCs) developed by Murray et al.¹ bypass the need for methane steam reforming by directly oxidizing methane to CO₂and H₂O. Devices that use direct methane oxidation should find new applications in areas where small-scale units are appropriate. MCFC, molten-carbonate fuel cell; PAFC, phosphoric-acid fuel cell; PEMFC, proton-exchange-membrane fuel cell.

High resolution image and legend (40K)

Over the past decade there has been a growing realization that there could be an important market for small (1-10 kW) thick-film SOFC stacks operating at intermediate temperatures (500-700 °C). These systems could provide heating and electricity for buildings⁴ (domestic and small commercial units), or provide electrical power for auxiliary functions (such as electric windows or air conditioning) in vehicles⁵. Although the provision of excess steam is appropriate for relatively large (> 1 MW) SOFC systems, it introduces cost and efficiency penalties as the system size is reduced. Direct electrochemical oxidation of dry methane in intermediate-temperature SOFCs, as developed by Murray et al.¹, would have many advantages in small-scale applications.

Methane oxidation and activation has been investigated with a wide range of oxide catalysts⁶,⁷, but at present no consensus exists about the mechanisms involved. It has been suggested⁸ that oxygen surface species (O¹, O²¹, O₂²¹ and so on) and lattice oxygen in the electrodes are responsible for different types of behaviour, and that high catalytic activity is often associated with mixed electronic and oxygen-ion conductivity.

In the SOFC research community, doped La_0.8Ca_0.2CrO₃ (LCC) and doped Ce_0.9Gd_0.1O_1.95 (CGO) have received the most attention. LCC is a promising anode material⁷,⁹ because of its wide thermodynamic stability range, and existing use as an electrical connector between single fuel cells. Catalysts that contain ceria (such as CGO) have attracted interest¹⁰ because it has been known for decades¹¹ that the use of doped ceria can improve the performance of SOFC anodes. Both LCC and CGO have good catalytic activity for methane oxidation. Steady-state measurements⁷ with a limited supply of oxygen gas indicate that LCC oxidizes methane to CO₂and H₂O at around 400 °C (Fig. 2). The reaction with CGO occurs at lower temperatures, around 300 °C. Exposure to residual excess methane confirms that only small amounts of carbon are deposited on both oxides at high temperatures (800-900 °C). Moreover, if carbon is intentionally deposited on the oxides it can be removed¹² at relatively low temperatures (300-400 °C) by the introduction of oxygen. The excellent catalytic behaviour of the ceria-based materials is not totally unexpected because CeO₂-ZrO₂ (CZO) catalysts are used in catalytic converters fitted to car-exhaust systems. It appears that CGO and CZO promote the reduction of Ce⁴⁺ to Ce³⁺, which enhances oxygen-exchange processes¹³ and associated catalytic reactions.

Figure 2: Conversion of methane by an oxide catalyst⁷.

The percentage conversion of CH₄by La_0.8Ca_0.2CrO₃powder is shown as a function of temperature when fed with a mixture of methane and oxygen (diluted in helium) in the ratio 5:1 (by volume). Methane is directly oxidized to CO₂and H₂O at temperatures around 400 °C. At higher temperatures CH₄is reformed to H₂and CO fuel with minor carbon deposition on the oxide catalyst.

High resolution image and legend (33K)

These catalytic studies suggest that the use of oxide electrodes instead of the standard composite nickel-ceramic anodes could have advantages in producing direct electrochemical oxidation of dry methane. But so far it has been impossible to synthesize a single-phase oxide material that satisfies all the criteria specified¹⁴ for use at intermediate temperatures.

The experiments of Murray et al.¹ use electrodes consisting of a thin layer (0.5 m) of yttria-doped ceria (YDC) together with a thicker layer (2 m) of a nickel-ceramic composite as the current collector, demonstrating one way to overcome the relatively poor electronic conductivity of YDC. Yet although it works in the laboratory, this solution is unlikely to be adopted by SOFC developers for two reasons. First, natural gas usually contains higher alkanes (C₂H₆, C₃H₈and so on) that are more easily dissociated than methane, resulting in carbon deposition on nickel at lower temperatures. Second, a number of investigations with intermediate-temperature SOFC stacks has revealed that when nearly all the fuel is converted to steam and CO₂the nickel can be oxidized to NiO with a concomitant degradation in the anode performance.

A more promising strategy is to fabricate a composite anode in which doped CeO₂is mixed with another oxide exhibiting good electronic conductivity, so avoiding the use of nickel altogether. This approach is being examined by a number of laboratories around the world and, if successful, is likely to speed up the commercialization of intermediate-temperature SOFC for small-scale applications.

References

1. Murray, E. P. , Tsai, T. & Barnett, S. A. Nature 400, 649–651 (1999). | Article | ISI | ChemPort |
2. Steele, B. C. H. CR Acad. Sci. 1, Serie IIc, 533-543 (1998).
3. Casanova, A. J. Power Sources 71, 65–70 (1998).
4. Diethelm, R. et al. in Proc. 3rd European SOFC Forum (ed. Stevens, P.) 87-93 (European Fuel Cell Forum, Oberrohrdorf, Switzerland, 1998).
5. Morton, I. in Automotive Engineer June (1999).
6. Doshi, R. , Alcock, C. B. , Gunasekaran, N. & Carberry, J. J. J. Catalysis 140, 557–563 (1993).
7. Middleton, P. H. et al. in Solid Oxide Fuel Cells III Vols 93-94 (eds Singhal, S. & Iwahara, H.) 542-551 (Electrochem. Soc., Pennington, New Jersey, 1993).
8. Gellings, P. J. & Bouwmeester, H. J. M. Catalysis Today 12, 1–105 (1992).
9. Steir, J., van Herle, J. & McEvoy, A. J. in Proc. 3rd European SOFC Forum (ed. Stevens, P.) 267-276 (European Fuel Cell Forum, Oberrohrdorf, Switzerland, 1998).
10. Marina, O. A. , Bagger, C. , Primdahl, S. & Mogensen, M. Solid State Ionics 123, 199–208 (1999).
11. Takahashi, T. in Physics of Electrolytes Vol. 2 (ed. Hladik, J.) 989-1049 (Academic, London, 1972).
12. Ramirez-Cabrera, E., Atkinson, A. & Chadwick, D. Solid State Ionics Abstr. I-10 (in the press).
13. Naito, H., Sakai, N., Otake, T., Yugami, H. & Yokawa, H. Solid State Ionics Abstr. C-07 (in the press).
14. Steele, B. C. H. , Middleton, P. H. & Rudkin, R. A. Solid State Ionics 40/41, 388–393 (1990). | Article |

1. Brian C. H. Steele is in the Department of Materials, Prince Consort Road, Imperial College, London SW7 2AZ, UK.
   e-mail: Email: b.steele@ic.ac.uk