A ruthenium catalyst has been developed that, at a few parts per million, releases hydrogen gas from methanol, a simple bulk chemical. The advance might allow methanol to be used as a hydrogen source for fuelling vehicles. See Letter p.85
Fossil fuels are a worry. For example, the phenomenon of global warming is linked to the presence of atmospheric greenhouse gases arising from their use1. Combustion of these fuels has provided convenient energy for centuries, and so there are also questions about the extent of remaining, accessible oil supplies2. Combine these issues with environmental concerns, and with humans' seemingly insatiable thirst for energy, and the result is a strong motivation to seek alternative energy sources. In this issue (page 85), Nielsen et al.3 report a dramatic advance that might pave the way to the practical use of one such alternative: hydrogen gas derived from the solvent methanol*.
The use of 'traditional' alternatives to fossil fuels, such as hydroelectricity and nuclear power, has increased, and technologies that exploit nature's wind, solar and tidal power are also being developed to supplement the energy needs of stationary users. For transportation applications, the advent of hybrid technologies — such as cars that use electric motors alongside petroleum-driven combustion engines — may reduce fossil-fuel consumption. However, a longer-term alternative to fossil fuels has been the subject of speculation by scientists, business leaders and futurist visionaries alike4.
A leading candidate for a 'clean' energy source is hydrogen, because the only product of its combustion is water5. This would seem to offer a solution to energy, pollution and greenhouse-gas problems in one fell swoop. But there are downsides. Public acceptance of hydrogen-based technologies is undermined by the gas's flammable nature — a fact indelibly linked to the explosion of the Hindenburg passenger airship in 1937. Finding a way to store large quantities of hydrogen in a portable volume is also a challenge. And even if these issues were overcome, distributing hydrogen to the millions of potential users would require massive new infrastructure.
Nobel laureate George Olah and his colleagues have advocated methanol (CH3OH) as an alternative fuel, and have proposed the development of a “methanol economy”6. Perhaps the most compelling aspect of their vision is the fact that methanol is hydrogen-rich (12.6% by weight) and, because it is a liquid at ambient temperatures, could be easily distributed using the existing infrastructure for petroleum. The development of chemical methods for extracting hydrogen from methanol is crucial to the success of this strategy. Currently, the best systems use platinum-based catalysts operating at temperatures above 200 °C, which limits the efficiency of methanol-based fuel cells to about 40%.
Nielsen et al. now report ruthenium-based molecules that catalyse the highly efficient liberation of hydrogen from methanol in water at less than 100 °C. The ruthenium catalysts ensure that, for each molecule of methanol and water consumed, three molecules of hydrogen are generated, along with one molecule of carbon dioxide (Fig. 1a). The authors observed that the reaction is facilitated in strongly basic solution, which also sequestrates the CO2 almost quantitatively as carbonate or formate salts. Using such strongly basic conditions, the researchers found that the ratio of hydrogen to CO2 in the evolved gas is consistently greater than 500:1.
To optimize the reaction conditions, Nielsen and colleagues studied the effects of catalyst concentration, the ratio of methanol to water, the basicity of the solution, and temperature. In the best case, the authors found that as little as 1.8 parts per million (p.p.m.) of the catalyst produced hydrogen from a 9:1 mixture of methanol to water, using a high concentration (8.0 moles per litre) of the base potassium hydroxide at 91 °C. Under these conditions, approximately 2,700 equivalents of hydrogen per equivalent of the catalyst were liberated every hour. This 'turnover frequency' rose to 4,700 equivalents per hour if pure methanol was used as the solvent, with just 1.6 p.p.m. of catalyst. However, for real-world applications in fuel cells, methanol–water mixtures and lower concentrations of base will be required.
To test their catalysts in real-world conditions, Nielsen et al. decreased the ratio of methanol to water to 4:1, used a much lower concentration of a different base (0.1 moles per litre of sodium hydroxide) and increased the catalyst concentration (to 21 p.p.m.). Under these conditions, the initial catalyst-turnover frequency decreased to about 800 equivalents per hour. The researchers observed that the pH of the solution fell from 13 to 10 during the first 4 hours of reaction and that the ratio of hydrogen to CO2 in the produced gas changed from 9:1 to 3:1 over the same period. However, the composition of the gas mixture then remained stable for up to a further 3 weeks of operation, equivalent to a remarkable 350,000 turnovers of hydrogen from the catalyst.
The authors did not test their catalyst in methanol fuel cells, but their findings suggest that the catalysts will improve the energy capacity of such cells. In the short term, this augurs well for applications such as portable electronic devices. For transportation applications, however, the catalyst-turnover frequencies will have to rise substantially.
The active form of the catalyst is generated in situ from a precursor complex in which ruthenium is bound by the nitrogen and phosphorus atoms of a 'tridentate' ligand molecule, HN(CH2CH2PR2)2 (where R is either a phenyl group, C6H5, or an isopropyl group, CH(CH3)2). The ruthenium is also bound by a hydrogen atom, a chloride ion and a molecule of carbon monoxide. Nielsen and co-workers probed the mechanism of action of their catalyst, and found that the dissolved base removes a proton (H+) from the nitrogen of the tridentate ligand to generate the active catalyst (Fig. 1b). This nitrogen and the ruthenium atom then interact with methanol to liberate a hydrogen molecule (H2), transiently generating formaldehyde (O=CH2). In the presence of water, the formaldehyde forms a gem-diol(ate) species (CH2(OH)O−), and then becomes a formate intermediate (HCO2−) by losing another hydrogen molecule. Loss of CO2 and of a third hydrogen molecule from the formate and the nitrogen of the ligand on ruthenium regenerates the active form of the catalyst.
Nielsen and colleagues' work is a seminal finding, because it demonstrates the viability of a soluble catalyst for the efficient and long-lived generation of hydrogen from methanol. Chemists know a great deal about the design and optimization of molecular catalysts, and so they will undoubtedly embark on further studies aimed at improving the catalytic activity of ruthenium complexes. They could thereby discover innovations that will bring us even closer to the methanol economy.
*This article and the paper under discussion3 were published online on 27 February 2013.