Take sunlight, add water, and there you have it: free energy. Plants have been doing this for quite some time, splitting water’s hydrogen apart from its oxygen, but our efforts to turn water into a source of free hydrogen fuel by mimicking them have borne no fruit. The problem is that splitting water takes more energy than conventional solar-cell technology can realistically deliver. But now we may be tantalisingly close to having economically viable sun-powered water splitters, and with it all the clean-burning fuel we want.
In 2008, Daniel Nocera at the Massachusetts Institute of Technology and his team unveiled a revolutionary approach to splitting water. They used a cheap cobalt-phosphate catalyst and titanium oxide electrodes that need far less electricity than conventional electrolysis to split water.
That raised the possibility of stealing plants’ trick and using sunlight to power the reaction. However, the number of photovoltaic cells needed for such devices mean it cannot compete on price with fossil fuels, says Daniel Gamelin, a chemist at the University of Washington in Seattle.
But Gamelin and his team thought they could bring down the costs by incorporating some of that photovoltaic technology in Nocera’s water-splitting device, creating a so-called photoelectrochemical (PEC) water splitter.
Nocera’s electrode was an indium-tin-oxide strip coated with cobalt and phosphate. Gamelin’s team also used cobalt and phosphate, but they started with hot glass, onto which they sprayed an iron solution. The iron oxidises in the air, forming a crystalline rust. The rust crystals give the electrode a large surface area, and it also happens to have photovoltaic properties.
The team then immersed their rust electrode in a solution containing cobalt and phosphate, and applied a current to electrochemically deposit the compounds on the surface. This created a PEC electrode that can at once generate current and catalyse the water-splitting process.
So far, the electrode cannot generate enough power to do this on its own, but even so it could reduce the amount of solar cells needed, making the process far cheaper, says Gamelin (Energy and Environmental Science, DOI: 10.1039/c0ee00030b).
Gamelin is also investigating the possibility of a so-called “tandem” device, which can generate enough energy from sunlight to power the water-splitting process on its own.
This device would have two cells housing electrodes, one on top of the other, with a rust electrode coated in cobalt and phosphate on top. Sunlight would strike the top electrode, which would absorb photons and catalyse the water-oxidation process. But not all the sunlight would be absorbed by this electrode: light with a wavelength longer than 600 nanometres isn’t absorbed by the rust-coloured water in the top cell so would pass through to strike the lower electrode, powering the production of hydrogen.
Like Nocera’s original device, Gamelin’s technology is also only able to produce oxygen gas and hydrogen ions. Teams around the world are searching for suitable cathode materials which can efficiently turn those ions into hydrogen gas, says Gamelin.
Meanwhile, other teams are working on alternative water-splitting devices. For example, a team led by Licheng Sun at KTH Royal Institute of Technology in Stockholm, Sweden, is working on a system that uses a photosensitised anode similar to those used in dye-sensitised solar cells (Chemical Communications, DOI: 10.1039/c0cc01828g).
Unlike Gamelin’s system, Sun’s device is already producing both oxygen and hydrogen gas. However, the current version uses an expensive, and externally powered, platinum cathode. To produce a commercially viable device, Sun is exploring the use of carbon and cobalt-based cathodes, with which he hopes to ultimately reach a solar-to-hydrogen efficiency of around 10Â per cent.
But almost all of the electrode materials studied to date are impractical, says John Turner at the National Renewable Energy Laboratory in Golden, Colorado. “The efficiency is abysmal,” he says.
Instead of using titanium oxide or iron oxide, researchers need to explore advances in photovoltaic devices, where the best today achieve an efficiency of 27Â per cent, he says.