Is night falling on classic solar panels?

http://www.newscientist.com/article/mg20827915.000-is-night-falling-on-classi...



Is night falling on classic solar panels?

* 20 December 2010 by Duncan Graham-Rowe


A new breed of electronic solar cells that harvests power from heat could double the output of conventional panels

SOLAR cells that work at night. It sounds like an oxymoron, but a new breed of nanoscale light-sensitive antennas could soon make this possible, heralding a novel form of renewable energy that avoids many of the problems that beset solar cells.

The key to these new devices is their ability to harvest infrared (IR) radiation, says Steven Novack, one of the pioneers of the technology at the US Department of Energy's Idaho National Laboratory in Idaho Falls. Nearly half of the available energy in the solar spectrum resides in the infrared band, and IR is re-emitted by the Earth's surface after the sun has gone down, meaning that the antennas can even capture some energy during the night.

Lab tests have already shown that, under ideal conditions, the antennas can collect 84 per cent of incoming photons. Novack's team calculates that a complete system would have an overall efficiency of 46 per cent; the most efficient silicon solar cells are stalled at about 25 per cent. What's more, while those ideal conditions are relatively narrowly constrained for silicon solar cells - if the sun is in the wrong position, light reflects off a silicon solar cell instead of being absorbed - the antennas absorb radiation at a variety of angles. If the antennas can be produced cheaply, the technology could prove to be truly disruptive, says Novack.
Solar arrays of billions of the tiny antennas have an efficiency as high as 84 per cent

Unlike photovoltaic cells, which use photons to liberate electrons, the new antennas resonate when hit by light waves, and that generates an alternating current that can be harnessed.

To build an array that could capture both visible and infrared radiation, researchers envision multiple layers of antennas, with each layer tuned to a different optical frequency.

So far, two main challenges have stood in the way of fomenting a revolution in solar power. First, the length of the antennas must be close to the size of the wavelength being captured, which in the case of the solar spectrum can be very small - from millimetres down to a few hundred nanometres.

Second, the currents produced will be alternating at frequencies too high to be of use unless they are first converted into a steady direct current. The problem here is that silicon diodes, which are crucial to the conversion, typically don't operate at the high frequencies required, says Aimin Song, a nanoelectronic engineer at the University of Manchester, UK.

Both of these barriers are now being broken down. Earlier this year, Novack and colleagues perfected a technique for creating arrays of billions of antennas. Although these antennas were only just small enough to harvest energy at the far end of the infrared spectrum, Novack says it should be possible to modify the process and build smaller antennas to work with mid and near-infrared.

Meanwhile Song, and Garret Moddel's team at the University of Colorado in Boulder, have independently taken a significant step in tackling the current-conversion problem by creating novel diodes capable of handling high optical frequencies (see "The devil's in the diodes"). Both groups expect to combine the diodes and antennas into working prototypes within months. "There's a potential for this to be a real game-changer," says Moddel.
The devil's in the diodes

Semiconductor diodes act like valves, converting alternating current into direct current. To work with the novel antennas, they have to operate at the AC frequencies being received and match the conductive properties of the antenna.

Semiconductors are ill-suited for this, as they tend to become less conductive when shrunk to the size of the antennas. Several groups have tackled this problem, creating diodes based on different concepts. One is that at tiny scales, the physical geometry of the device influences current flow: by creating asymmetry in the geometry, electrons can be funnelled to flow one way only.
Tweet