news and views Hamburg defining ten parts per million of Such fuels react with oxidizing species at problems, such as leakage through the poly- nitrogen oxides as the upper level of effluent lower temperatures without carbon forma- mer membrane or cracking of the ceramic from fossil-fuel burners. This is quite a diffi- tion and may be ideal power supplies for electrolyte, remain to be sorted out. It is cult limit to achieve because ordinary flame computers or mobile phones7. But such worth recalling that Nernst, the inventor of burners give off roughly ten times this intermediates, as with hydrogen or pure the zirconia electrolyte in the 1890s, praised amount of nitrogen oxides. methane, might not be readily available in the simultaneous invention of the telephone The logical answer is to imitate nature the quantities and economics we expect for because it enabled him to call home to switch and to oxidize the hydrocarbon fuel catalyti- regular fuels. the device on, the warm-up time being so cally, while extracting the electrons and use- Of course, the fuel reactions in the device long. Solutions to these problems are now ful energy directly through a membrane. are not the only areas of ignorance. Fuel- being found8. Only when this happens will But our knowledge and expertise of such cell systems are already semi-commercial the expected applications in domestic processes is shockingly weak, despite the but a number of issues remain to be resolved heaters, hybrid vehicles and distributed early discoveries of catalysis and fuel cells if the expected billion-dollar consumer power plants become reality. ■ by Davy4 and Grove5. A fuel cell works by market is to emerge in 2004 — the date set Kevin Kendall is in the School of Chemical 1 reducing oxygen to O2 at the cathode, while for car manufacturers in California to pro- Engineering, University of Birmingham, oxidizing fuel, ideally hydrogen, at the anode duce 5% zero-emission vehicles. One prob- Edgbaston, Birmingham B15 2TT, UK. & (Fig. 1). Ions created by this process (H lem is the cost of electrolyte membranes, e-mail: [email protected] 21 or O ) diffuse through the electrolyte for both solid polymer and ceramic fuel 1. Grove Symposium J. Power Sources March (2000). membrane to balance the reactions. At the cells, so the cost of materials will need to 2. Park, S., Vohs, J. M. & Gorte, R. J. Nature 404, 265–267 same time, electrons flow from the anode to fall by an order of magnitude. The perfor- (2000). 3. Houghton, J. Global Warming Ch. 3 (Lion, Oxford, 1994). the cathode through an external circuit. If mance of a fuel cell may also be limited by 4. Davy, H. Phil. Trans. R. Soc. Lond. 107, 45–85 (1817). the fuel is hydrogen, the only waste product peripheral equipment, which may represent 5. Grove, W. R. Phil. Mag. 14, 127–130 (1839). is water. Other fuels, such as methane, can be 75% of the total device. Existing systems, 6. Murray, E. P., Tsai, T. & Barnett, S. A. Nature 400, 649–651 used, but they have to be converted to hydro- such as the d.c./a.c. converter and heat (1999). 7. Büchi, F. N. (ed.) Portable Fuel Cells Conference (Bossel, gen before entering the cell. exchanger, may not be well matched to the Switzerland, 1999). Lately, evidence has emerged that new fuel-cell technology. And particular 8. Kendall, K. US Patent No. 5,827,620 (1998). methane can be converted directly by a solid- oxide fuel cell, without any flame, using cerium oxide as the electrode catalyst6. A typ- Evolution ical solid-oxide fuel cell uses a hard ceramic material such as zirconia (which conducts O21 ions) as the electrolyte, and has nickel- In search of the whales’ sisters based anodes. Park et al.2 now suggest that Zhexi Luo copper with cerium or samarium oxide may be a better anode catalyst for direct oxidation of more complex hydrocarbons that exist in hich mammals are the closest rela- molecular evidence1,2. They show that both realistic fuels such as kerosene or diesel. The tives to cetaceans (dolphins, por- issues hinge on the long-extinct fossil groups challenge is to find pathways through which Wpoises and whales)? That is, which that were split early from extant cetaceans. ethane and butane, or even aromatics such as mammalian group is their ‘sister taxon’? The topology, or branching sequence, of the toluene, can be reacted without fouling up There is a wide gulf between the morpholog- family tree of living artiodactyls and the process through damaging side reactions ical and molecular evolutionary studies on cetaceans depends on inclusion of these that tend to produce tar or carbon. the question, for they give conflicting divergent lineages of cetaceans and their This is the challenge that has been picked answers. In a paper in Systematic Biology, putative extinct relatives. up by Park and colleagues. They used a O’Leary and Geisler1 highlight the impor- Among those possible fossil relatives are copper catalyst intimately in contact with tance of the early divergent lineages, or the mesonychids that lived from 60 to 30 cerium oxide to produce a porous anode extinct fossil taxonomic groups, in resolving million years ago. Their feet were even-toed material. When this operated on pure this intriguing problem. and adapted for running, like those of butane for 48 hours there was no sign of car- The descent of whales from land-dwelling modern artiodactyls, and their fish-eating bon deposition, but good evidence of com- mammals is a compelling example of evolu- or carrion teeth closely resemble those of plete chemical conversion and electron tion. It is documented by a rich fossil record the earliest whales. Since the 1960s3, several transfer through a zirconia electrolyte of intermediate forms spanning the morphological studies1,2,4,5 (with one excep- membrane. An ordinary cell would die in land–water transition; and by morphological tion6) have considered that extinct mesony- minutes under the same operating condi- and molecular lines of evidence, both of chids and cetaceans are sister taxa (Fig. 1a). tions because of carbon fouling. When the which testify that cetaceans have a close affin- The combined mesonychid–cetacean group copper–ceria–samaria catalyst was used, ity to the artiodactyl mammals, or ungulates is, in turn, related to living artiodactyls. In even toluene showed reasonable reactivity with even-toed feet and double-pulley ankle, this scheme, living artiodactyls are recog- and electron conversion. This is very such as today’s cow, camel and hippopota- nized as a clade1,2,5,7, a genealogical group encouraging because aromatics are usually mus. But two central issues have yet to be set- that includes all descendants of a common prone to rapid graphite deposition, which tled. One is the identity of the cetacean sister ancestor. destroys the catalyst. taxon. The other is whether the cetaceans are In contrast, molecular studies8–12 have The hope is that even gasoline or diesel ‘nested’ within the group of artiodactyls as generally concluded that hippopotamuses fuel may soon be catalytically converted to a branch on the family tree of the living artio- and cetaceans are more closely related to energy and heat in a clean and powerful dactyls — in other words, whether cetaceans each other than either group is to other liv- manner by such devices. But it may be easier represent some ancient artiodactyls that ing artiodactyls (Fig. 1b). Consequently, the to use partly oxidized fuels such as methanol became adapted to aquatic life. artiodactyls would be a grade that has similar or formic acid because these will work in What O’Leary and Geisler have done is to evolutionary adaptations, but not a genea- aqueous solutions with catalytic electrodes. evaluate all the available morphological and logical clade. Moreover, some molecular NATURE | VOL 404 | 16 MARCH 2000 | www.nature.com © 2000 Macmillan Magazines Ltd 235 news and views studies12 suggest that mesonychids are not suggest), then cetacean adaptations such as closely related to cetaceans; this is because the underwater nursing of offspring and nearly mesonychid–cetacean relationship implies a hairless skin could have originated in the larger gap in the fossil record9, and is there- most recent common ancestor of both fore less plausible than the hippo–whale groups (Fig. 1b); this implies that certain connection12. So, for all their congruence on aquatic features had evolved before the ori- the broad picture of ungulate–cetacean gin of cetaceans9. The similar ear regions4,5 100 YEARS AGO evolution, there is a big disagreement and the fish-eating and carrion teeth must On reading the letter of Prof. Dexter on between morphological and molecular stud- have evolved independently in mesonychids “Assaults and Drunkenness” (p. 365), I ies over these two phylogenetic issues. and in the earliest cetaceans because these notice that there is one great fallacy in the What does that mean in terms of the ori- derived features are absent in living hippos argument. When a man is intoxicated and gin of cetacean adaptations to aquatic life? If and their artiodactyl allies. commits an assault, the result is entered in hippos and whales are sister taxa to the exclu- Alternatively, if mesonychids and ceta- police reports as “assault”, the more serious sion of mesonychids (as molecular studies ceans are sister taxa (as the best morphologi- offence overshadowing the less. So that, in all probability, many of the cases of assault referred to in the statement were also Other mammals drunkenness, but were not tabulated as such. The temperature is an important Horse element; for its variations are probably the cause of the change of character of the Rhino offences recorded.