Nano lasers offer hope for scalable optical processing

The pretenders to the throne—optical devices—continue to sit off to one sidewhile their supporters develop cunning plans for a takeover. All of those plansfocus on making optical circuit elements smaller.

One of the main barriers to reducing the size of opticalcomponents is the wavelength of light. Visible light has a wavelength of around500nm, so devices that manipulate light, like lenses and waveguides, must havecomparable sizes. At least up until now—a long-awaited development has nowprovided a proof of principle, demonstrating that lasers can be made as smallas 50nm, sizes that are comparable to current electronic features.

Normally, the absolute smallest length that laser hardwarecan have is a half wavelength, so the blue laser diode in your PS3 could be asshort as 200nm (it's not though; it's considerably longer). Even worse, whilethe length may now be 200nm, the width and height have to be much bigger sothat the end mirrors provide a good reflecting surface. These size issues havekept on-chip optics at something close to a stand-still for well over a decade.

In the last few years, scientists have begun to take freshinterest in the optical properties of metals. The way that light can cause theelectrons in a metal oscillate coherently has led to new ideas for scaling downoptical elements. These electron oscillations, called surface plasmonpolaritons, have been shown to travel down wires just a few nanometers indiameter, providing the opportunity to scale things down.

These surface plasmon polaritons don't last very longthough. The electrons in the metal quickly dissipate their energy by bouncingoff the atoms in the wire, heating them up, while the remaining energy isre-radiated as light. To overcome this, you need an amplifier, and, as aninitial source of light, a laser. Such a laser, called a spacer (surfaceplasmon laser), has been proposed on several occasions, but until now, nobodyhad actually produced one.

To create the spaser, researchers took gold nanospheres andcoated them with a sodium glass. On top of that, they placed an outer shell ofdye-impregnated glass. The inner gold sphere acts as the resonator for thelight. The light remains on the outside of the gold sphere as the electronsinside slosh back and forth—the light field extends far enough that it passesthrough the dye-impregnated outer shell, which provides gain.

The middle layer of sodium glass acts as a separator,preventing the dye molecules from interacting directly with the metal.Otherwise, the interaction would act to broaden the range of colors the dyemolecule could emit and shorten the amount of time it spends in the excitedstate, where it can emit light.

The basic idea is that blue-green light is shone on theparticles and absorbed by the dye. The excited dye molecules then begin tospontaneously emit green photons, some of which hit the chewy gold center,exciting a surface plasmon. The surface plasmon starts sloshing back and forth,and its field sweeps through the excited dye molecules, stimulating them toemit, adding to the surface plasmon. In the meantime, part of the energy storedin the surface plasmon radiates as coherent light.

The researchers observed that their spaser had all thecharacteristics expected of a laser: a threshold, narrow emission line, andrelaxation oscillations. A threshold means that it requires a certain amount ofinput energy before a sufficient number of dye molecules contribute light toovercome the losses that occur as the electrons slosh about. Below this energy,a weak broad range of colors are emitted, while above, only a single color isemitted.

Finally, after the energy is dumped into the dye, it takessome time for the spaser to start going but, once going, it quickly producesshort, intense pulses of light, called relaxation oscillations—typically, weprefer to operate lasers so that only a single pulse is emitted on theseoccasions, but we can't always be choosy.

So, we now have 50nm laser hardware, which could conceivablybe combined with nanowires to start developing optical circuits that really dolook like electronic circuits (e.g., small and cheap). This laser was poweredby a very powerful pump laser, meaning that it's only small if you ignore theenormous power supply. But that was a side effect of how the experiment was puttogether. A single spaser used about 20 microwatts of power, so much smallerpump sources are feasible. If they can achieve continuous wave operation, theresearchers are on to a winner.