Free-electron lasers are the must-have gadgets for all self-respecting modern laboratories. They work by sending a beam of electrons into an undulating magnetic field, called a wiggler. This changes the trajectory of the electrons, forcing them to emit coherent photons. That's cool, but their real flexibility comes from their tunability. Change the energy of the incident electrons or fiddle with the wiggler, and you can change the wavelength of the laser light they produce.

There's one problem though. Wigglers are big, complex devices that are expensive to build and difficult to operate and maintain. And that means that only the biggest and best-funded labs can afford to run them.

Now Nikolay Zheludev at the University of Southampton in the U.K. and a few mates are doing for the free-electron laser what Shockley, Bardeen, and Brattain did for the vacuum tube: they're turning it into a solid-state device that can be built into a single chip.

Here's how. They first create a slab of material consisting of alternating layers of gold (a metal) and silicon dioxide (a dielectric). They then drill a tiny hole--just 700 nanometers across--through the slab. Finally, they fire a beam of electrons through the hole. Instead of experiencing an undulating magnetic field, the electrons experience an alternating dielectric environment that has the same effect--it forces them to emit photons. And crucially, the team can tune the wavelength of the photons by changing the energy of the incident electrons.

Zheludev and his pals have already built a prototype that produces light in the range of visible to infrared with an emission intensity equivalent to 200 watts per square centimeter.That's an impressive device that could make a big impact in nanophotonic devices, in optical memory, and in next-generation displays.

But there's work to do yet. The light emitted by this device, which Zheludev calls a light well, is incoherent because the photon conversion process is relatively inefficient (the number of photons produced per incident electron is just 10^-5).So Zheludev's next job will be to improve this efficiency by reducing the energy lost to surface plasmons and the light lost inside the material slab. If he can improve the efficiency enough, lasing may become possible.

When that happens, Zheludev and his colleagues will have a device on their hands that could revolutionize lasing. In the meantime, they'll just have to settle for a revolution in nanophotonics.

Ref: arxiv.org/abs/0907.2143: The Light-Well: A Tuneable Free-Electron Light Source on a Chip