"PV Antennas," Input/Output, ME Departments, March 2008

input/output

PV Antennas

by Jeffrey Winters, Associate Editor

A square mile of Nevada desert receives 5 trillion watt-hours of radiation over an average 12 months, the equivalent of a continuous 570 MW. No one has yet turned solar's great potential into a competitive power source. But researchers at the Idaho National Laboratory may have found a way to make solar cells that would be cheap and ubiquitous. Unlike today's photovoltaic cells, made from pure crystals that must be grown in contamination-free environments, these new solar cells could be stamped out using common industrial techniques.

What's more, the new solar cells could be as much as 80 percent efficient in converting sunlight to electricity.

Photovoltaic cells are so common that it's easy to forget they rely on some advanced solid state physics. When light hits a semiconducting material, such as silicon, it can dislodge an electron; streams of electrons flowing through the material create a current that can be tapped to make electricity. But the best-performing materials convert only around a third of the incoming energy, and commercially available PV cells are less than 20 percent efficient. Also, PV cells rely on rare material such as indium and gallium, or expensive-to-manufacture silicon crystals.

There are, however, other ways to harvest energy from sunshine. Sunlight can be concentrated onto a small area using mirrors, for example. The concentrated energy will heat up a fluid that can be passed through a turbine to generate electricity. Since modern turbines have efficiencies in the 30 to 40 percent range, this is an effective way to make electricity from solar energy, especially in large installations.

But what promises to be the most efficient means of harnessing solar power is not through heat or photovoltaic effects. Instead, Steven Novack and his colleague, Dale Kotter, tap solar energy much the way a radio antenna taps into an FM signal: The electric field created by traveling light waves makes electrons oscillate in a nanoscale antenna.

To catch a wave: Researchers in Idaho have developed minuscule antennas that can absorb energy from infrared radiation.

Generally, the wavelike nature of electromagnetic radiation is useful only at lower frequencies, such as radio and microwaves. The shorter the wavelength, the smaller the antenna. Because the wavelength of light is around one-millionth the length of a typical VHF signal, a "light" antenna would have to be measured in micrometers.

That much has been known for decades, and techniques developed to make integrated circuits can easily produce the right size structures. But making enough nano-antennas to cover a postage stamp, let alone an acre, was impractical.

Undeterred, Kotter stuck with the idea, testing different materials and antenna configurations until he found one that could resonate when struck by light with a wavelength of about 10 micrometers. That corresponds to infrared, which is plentiful in sunlight, though not as powerful. A square meter exposed to sunlight receives about 600 watts of infrared light from the sun.

The antennas—gold coils made of wires just 50 nanometers thick—can lie on many kinds of non-conducting materials. On a visit to a Kodak plant in Connecticut, Novack and Kotter saw machinery that produced continuous rolls of material and realized that a similar "roll-to-roll" process could manufacture their solar sheets.

Automated production could bring down the cost of solar power below that of all other sources.

The Idaho team hasn't gotten to that point, yet. But in December, they displayed a plastic sheet made by a stamp-and-repeat process that contains some 260 million embedded antennas. Novack said the process that created that sheet could be easily scaled up to make energy-absorbing material like so much wallpaper. And the antennas need not be made of precious material, Kotter said, although the gold in the prototype is so thin that it requires much less than a gram per square meter.

"Unlike photovoltaics, the size of the antennas determines the wavelengths where they resonate," Novack said. "Getting to visible wavelengths just requires more accuracy and expertise in nanotechnology."

"It's not physics," Kotter added. "It's just a manufacturing problem."

Even so, it may be some time before nanoscale antennas power the electrical grid. For one thing, the team still must find a cheap and efficient way of converting the current induced in the antennas—which buzzes at the frequency of light, more than a trillion cycles per second—into conventional 60 Hz electricity. "The conversion from these high frequencies is going to be a challenge,"

Novack said. The solution may come from work on optical computing, which similarly handles high-frequency signals. Novack said a first application for the material would avoid such issues. The nano-antennas could be used to manage heat, drawing in infrared radiation at one wavelength and radiating it away at another, less troublesome wavelength.

"The first thing that comes to mind is the large racks of computers," Novack said. "They need to be constantly cooled, which is a big power drain. These antennas could more efficiently take the heat out of that environment."