A new class of candidate materials for hydrogen storage — urgently needed to carry this fuel safely in a 'hydrogen economy' — has been made-to-measure by scientists in the USA. They have built highly porous materials from metal and organic building blocks, in which the pores are of nanoscale dimensions.

One member of this class, designated MOF-5, is capable of containing up to 4.5 percent by weight of molecular hydrogen at low temperature (78 K). Although this figure falls to 1.0 weight percent at room temperature and pressures considered safe for hydrogen refuelling (20 bar), that is still a very respectable storage capacity compared with some of the other materials that have been explored previously, such as carbon-based adsorbents.

But it falls considerably short of the figure of 6.5 weight percent hydrogen set by the US Department of Energy (DoE) as the target for a viable hydrogen-storage material for applications such as hydrogen-powered (fuel-cell) vehicles. Nevertheless, the researchers expect to improve their results substantially, and indeed they have already quadrupled the hydrogen uptake in preliminary work on other members of this family of materials.

A hydrogen economy is one of the few objectives shared by both environmentalist groups and the US government. By burning hydrogen in fuel cells, rather than fossil fuels in conventional internal combustion engines and power-generation plants, we would avoid the burden of greenhouse gases that seem now to be warming the planet. Hydrogen fuel cells would also eliminate the toxic substances, such as carbon monoxide, nitrogen oxides, and fine carbon particulates, currently emitted in vehicle exhaust fumes; the only by-product of hydrogen fuel cells is water. This would be environmentally beneficial, but its attraction to the US administration of George W. Bush is that it would also alleviate American dependence on foreign oil.

The biggest hurdle is finding a 'clean' source of hydrogen. Currently it is produced mainly by the reforming of natural gas by catalytic steam at high temperatures, or by electrolysis of water. Both processes consume a lot of energy from conventional sources. The ideal would be to make hydrogen by solar-powered electrolysis or photocatalytic water-splitting. The former is used to some extent at prototype hydrogen-refuelling stations in California, but neither process is yet commercially viable on a large scale.

But fuel-cell vehicles running on hydrogen will also have to store this explosive substance in a safe form. At present, such vehicles tend to use tanks of compressed or liquefied hydrogen. A better option is to absorb large quantities of the gas in some solid-state material. Various candidates have been explored, but none has so far come up to scratch.

Metal hydrides can contain a high weight fraction of hydrogen, but they tend to be heavy. DaimlerChrysler is investigating the use of sodium borohydride for fuel-cell vehicles. There was excitement in the mid-1990s when it was claimed that carbon nanotubes could act as miniature storage cylinders with a capacity exceeding the DoE goal. But these claims have not been replicated. Much of the work is now focused on other forms of carbon, such as porous 'activated' carbon and intercalated graphite. Typically these achieve storage densities of only a few tenths of a percent by weight.

The materials now reported by Omar Yaghi of the University of Michigan and co-workers¹ are metal-organic frameworks, in which linear organic groups are joined into a three-dimensional open framework by coordination to metal-containing clusters. Yaghi and colleagues have considerable past experience in assembling these structures from a variety of components to make materials laced with nanoscale pores, whose precise dimensions are set by the size of the organic molecules that serve as the 'struts' of the framework.

Their hydrogen-storage materials consist of oxygen-centred tetrahedral zinc clusters (OZn₄) at the vertices of a cubic framework, each coordinated to six carboxylate groups in the organic linkers. These linkers are linear dicarboxylates with an aromatic centre: they are in effect struts with two sticky ends.

In MOF-5 the linker is 1,4-benzenedicarboxylate. Having established that this microporous material adsorbs substantial amounts of hydrogen even at room temperature and moderate pressure, Yaghi and colleagues used inelastic neutron scattering spectroscopy to probe the adsorbed molecules and identify the hydrogen binding sites. They believe that there are two types of sites: one associated with the zinc atoms, and the other with the organic linkers.

They figured that a higher hydrogen loading might be achieved by making the linkers larger, providing more room for the hydrogen molecules to bind. And indeed, linkers in which the benzene groups are replaced by naphthalene produce a material with approximately four times the hydrogen uptake of MOF-5 at room temperature and 10 bar pressure. This is still not enough to meet the DoE target, but it is getting close. And most importantly, it shows the scope for improving the properties of these materials by tinkering with the nanoscale design