Terahertz radiation fits into the electromagnetic spectrum between infrared waves and microwaves and can be used for all manner of extraordinary things. One application is medical imaging--terahertz waves can penetrate a short distance beneath the skin and so spot tumors. Another use is surveillance--terahertz waves pass through most types of clothing, allowing "intimate" body searches at a distance. Terahertz radiation also shows promise for real-time spectroscopy--molecules can be identified by the way they reflect short pulses of terahertz waves.

So how best to create and detect terahertz waves? We'll leave the creation aside and talk in this post about detecting terahertz waves for real-time video images.

That's been a tough task until recently. The current state of the art is to use an array of microbolometer pixels, consisting of tiny slabs of vanadium oxide that change their resistance when hit by infrared photons and heat up. Although designed for infrared up to wavelengths of 14 micrometers, these devices have a residual sensitivity at terahertz frequencies, and in the last couple of years, they've been used to make terahertz videos.

But a new kid on the block is set to take the microbolometer's crown. In the early 1990s, Michel Dyakonov at the University of Montpellier 2 in France and a pal calculated that a sufficiently small field-effect transistor ought to be able to produce and detect terahertz radiation.

The idea is that if a transistor is small enough, the electrons in the channel between its source and drain can be thought of as a 2-D plasma. And if the properties of this channel are tuned just right, the plasma can resonate when hit by terahertz waves. This, in turn, influences the current passing through the channel and allows the terahertz waves to be detected.

Of course, the size of these transistors is measured in nanometers, which is why it's taken so long to actually build them.

However, Dyakonov is now part of a group that has not only built transistors that detect terahertz waves, but also created the first images with them (see above).

That's impressive for several reasons. First, these transistors detect terahertz waves at room temperature (although nobody is quite sure how--something that will have to be worked out before scientists will want to use them). Second, the read-out rate is good enough for video. And finally, the transistors can be built into arrays using standard silicon CMOS (complementary metal-oxide-semiconductor) technology, which means they have the potential to be as cheap as conventional video cameras.

Dyakonov and his colleagues say "the experimental and theoretical results clearly indicate that nanometer transistors are promising candidates for a new class of efficient terahertz detector."

Can't argue with that. However, there might be more interest when the videos of at-a-distance "body searches" start appearing on YouTube.

Ref: arxiv.org/abs/0907.2523: Field Effect Transistors for Terahertz Detection: Physics and First Imaging Applications