Solar power enthusiasts have long dreamed of replacing fossil fuels with clean-burning hydrogen gas. Although solar cells can be harnessed to rip apart the hydrogen and oxygen in water molecules, the cells haven't been economical. Now researchers report in tomorrow's Science  that they've created a new solar water splitter that nearly doubles the energy output of previous devices, bringing the vision of clean energy closer than ever before.
Hydrogen is often viewed as the ultimate clean energy, because water is the only byproduct produced when it burns. The fuel is also easy to make: Just stick a pair of electrodes in water and apply a voltage between them. But this process, known as electrolysis, is expensive and typically uses fossil fuels to generate the electricity. So energy researchers have been working to cut the cord to the power grid by using solar energy to drive the electrolysis.
Solar water splitters work with the help of semiconductor-based solar cells that absorb photons, creating mobile electrical charges that get channeled to water-splitting electrodes. But splitters that can create electrical charges with just the right amount of energy to be taken up by hydrogen and oxygen are typically poor absorbers of sunlight.
To get around this problem, chemists John Turner and Oscar Khaselev of the National Renewable Energy Laboratory in Golden, Colorado, decided to combine two separate semiconductor layers. The first, made from gallium indium phosphide, absorbs ultraviolet and visible light and produces electrons with just the right amount of energy needed to produce hydrogen at one of the electrodes. The other, made from gallium arsenide, absorbs infrared light and creates mobile positive charges with the right amount of energy to produce oxygen at the other electrode. The result is a sunlight to hydrogen efficiency of nearly 12.5%.
This efficiency is "impressive," says Marye Anne Fox, a chemist at the University of Texas, Austin. But she, Turner, and others note that the semiconductors in the new devices are still too expensive to make solar hydrogen production cheap enough to compete with fossil fuels. So for now, Turner and Khaselev are working on finding cheaper semiconductors capable of accomplishing the same energetic feats.