In today's nano-age, lasers are just so micro. To generate their trademark focused beam, lasers must contain a cavity in which to resonate and stimulate atoms. But even with an efficient resonance-promoting medium, light cannot resonate in cavities smaller than half its wavelength. This means that even the smallest optical laser would occupy a chunky 100–200 nanometres, a problem for engineers who need resonating cavities to be even smaller than this to design light-activated medical therapeutics and to build speedier nanometre-scale electronics.

Now, Mikhail Noginov of Norfolk State University in Virginia and his colleagues have followed up on hints that nanoparticles, which measure 2–100 nanometres in diameter, might serve as nanometre-scale resonators working at visible wavelengths. The researchers used visible light to excite free electrons in nanoparticles. The vibrating electrons form oscillations called surface plasmons, which, like a traditional laser cavity, can support stimulated emission, but do so using much less space (see page 1110).

The idea of surface-plasmon-based lasers, or 'spasers', was first proposed in Physical Review Letters by another group in 2003. Several years later, Noginov and his colleagues began brainstorming about the best way to build the right resonating cavity. That “very important part of the work”, Noginov says, helped them to target certain metal nanoparticles for their planned spaser. Pure metals are prone to losing some of the incoming energy, preventing the stimulated emission that is needed to produce a spaser. To get around this, co-author Ulrich Wiesner and his team at Cornell University in Ithaca, New York, fabricated hybrid nanoparticles with a 14-nanometre gold core surrounded by a 15-nanometre-thick silica shell embedded with dye molecules, making a 44-nanometre-diameter sphere. By combining gold with the dye, the team took advantage of the two materials' electrical properties: the gold provided the free electrons needed to create a surface plasmon; and the dye molecules transferred incoming energy to the surface plasmons.

Next, the researchers tested their creation using standard optical equipment. They sent light into water containing the nanoparticles, and detected the amplified light that emerged when the spaser began working. Noginov anticipated plenty of technical issues. “In real experiments, potential glitches are everywhere, like bugs in the swamp,” he says. At first, his team was seeing “nothing”, but eventually a sharp line appeared at 531 nanometres — the wavelength of the light emitted by the surface plasmons, and the first evidence of an optical spaser.

Noginov also drew on the expertise of co-authors Vladimir Shalaev and Evgenii Narimanov at Purdue University in West Lafayette, Indiana, who work with metamaterials — man-made materials with optical properties that do not occur in nature. These metamaterials demonstrate, “absolutely crazy phenomena” says Noginov, including optical invisibility cloaking like something straight out of Harry Potter. Shalaev's team measured the emission kinetics of the spaser using a sophisticated microscope.

The next step will be to shrink the resonator even further. How low will they go? “Theoretically,” says Noginov, “you can go as low as you want” until the metal atoms stop behaving collectively like a metal, which researchers predict occurs at diameters of about 1 or 2 nanometres. In future nanocircuits, such tiny spasers could fulfil the gatekeeping role that transistors held in the microelectronics era.