Device Turns Water, Sunlight to Fuel

Aaron J. Hand

With current technologies, the Earth's sun could supply all the electricity we need. About 10,000 square miles of photovoltaic panels could generate -- in about six hours of daylight -- a day's worth of electricity for the US, according to John A. Turner, senior scientist at the National Renewable Energy Laboratory. To make that work, however, you have to be able to store the energy generated for future use. This is where Turner comes in.
Turner has spent several years developing a device that could take sunlight and water -- two of our most abundant natural resources -- and generate hydrogen, a renewable energy source that does not pollute the atmosphere when used. With postdoctoral fellow Oscar Khaselev, he has developed a monolithic photovoltaic-photoelectrochemical device that splits water into its basic elements. The hydrogen that is created in this process is effectively stored sunlight.

Increased efficiency
Other systems use sunlight to split water into oxygen and hydrogen, but Turner and Khaselev's device does so more economically and with greater efficiency. These other systems use photovoltaic cells to generate electricity from the sun, linked with an electrolyzer that separates hydrogen from water. They typically convert 4 to 6 percent of the sunlight into hydrogen.
The energy laboratory's new device combines the two functions into one cell, reducing some of the high costs associated with previous systems and increasing efficiency to 12.4 percent. The device, based on a tandem cell developed at the laboratory, is immersed in an electrolyte. When sunlight shines on the device, the top layer, made of gallium indium phosphide, absorbs the visible light and produces hydrogen. The bottom layer, which is gallium arsenide, absorbs near-infrared light that passes through the top junction and the interconnect and produces oxygen.

Cost is still a problem
This device is more efficient and cost-effective than previous solutions, but it still is not an economical means of energy production. It still is three to four times more expensive than producing hydrogen through steam reforming of natural gas, Turner noted. The researchers plan to explore the use of other materials -- perhaps amorphous silicon or nitrides -- to bring costs down further. Turner foresees a potential efficiency of 16 or even 24 percent, but a device that is more efficient may not be more cost-effective, he said.
Cost, however, is unlikely to be the catalyst behind this technology, Turner said. Government mandates brought about by environmental concerns will probably play a bigger role, since what is good for the environment is rarely the cheapest solution. "If fossil fuels didn't pollute the atmosphere and were basically infinite in their size, we wouldn't have to worry about this," he said.