Capturing Solar Power With Antennae

necro81 writes "Researchers at the University of Missouri and the Idaho National Laboratory have demonstrated a new method of capturing solar power. Rather than using semiconductors to capture photons of sunlight, they fabricated small coiled antennae (several um square) that resonate with the wave nature of light. The antennae are tuned towards midrange infrared light (5-10 um), which is abundant on our cozy-warm Earth — even at night. They also demonstrated a way to imprint these coils on a substrate, like how CDs or vinyl records are produced, but could be scaled to roll-to-roll mass production. The usual caveat applies: it may be 5-10 years until this could hit the market."

Comments

Most important point not in summary

[GameboyRMH]

The summary fails to mention the most important advancement here: 90%+ efficiency. That's a game-changer for solar power.

Re:Most important point not in summary

[joe_frisch]

What are they using to rectify the signal to convert to DC? The antenna is neat - but not at all surprising, its size should just scale with wavelength. You could make a 125nm long antenna that would resonate with visible light (well withing the resolution of existing lithography). The problem is how to convert the 100THz signal you get to a DC signal. You need a fantastically fast diode.

If they have managed this, that would be an impressive achievement. The fastest diodes I am aware of are around 1THz, but its well outside my field and there might be something faster out there

BTW: the efficiency isn't all the impressive. For single frequency light, conventional solar cells can be quite efficient (~80%???), but they don't do will with broad thermal light (like sunlight). The photons that are less than a band-gap don't do anything, and the ones above a bandgap waste any excess energy.

Re:Most important point not in summary

[lupine]

From TFA:
The individual nantennas can absorb close to 90 percent of the available in-band energy.

So the total system efficiency depends on how wide that band is in relation to total solar energy available and whether nantennas can be stacked and designed to capture energy over a range of bands.

Thermodynamics is a bitch.

[Anonymous Coward]

This is, of course, utterly useless for harvesting power from ambient thermal radiation. Even if you can make a diode that's remotely capable of rectifying current at high enough frequencies, the diode has to be kept colder than the source of the radiation. It's the electrical analog of a Brownian ratchet.

Picking up a bit of the IR tail that conventional photovoltaics don't catch? Maybe, but there isn't very much power down there even if you got the efficiency usefully high. Turning ambient heat into usable energy? Sorry, no.

Re:Rectification is the hard part

[Anonymous Coward]

The simple answer is that electronics aren't instant. Every single wire and component in any device you can build actually acts like a resistor, capacitor, and inductor. The combination of these effects means that when you, say, apply a voltage to a wire, it takes some tiny amount of time to "charge up". Even with the gigahertz frequencies used in processors, things have to be specifically built (and be tiny) to work with these charge times. If you try and do anything with *terahertz* frequencies? Even a micrometer of wire won't be "charged' before the wave goes negative, at which point it discharges... and you wind up with an average of zero volts.

TL;DR: Until we have wire made out of superconductor, frequencies that high simply can't be transfered through circuits.