Accidental discovery leads to subwavelength transmission

17 Jun 2002

Thomas Ebbesen is a physical chemist for the NEC Research Institute and didn't intend to discover new light transmission properties. But as often happens in science, Ebbesen found what he wasn't looking for.

Ebbesen illuminated, in a silver film coating a quartz substrate, an array of subwavelength holes with 150 nm diameter 900 nm apart at wavelengths up to ten times the size of the hole. Theoretically transmission should have been reduced since the aperture was smaller than the wavelength, but more than twice as much light was transmitted as was impinged directly on the total area of the holes.

The discovery started nine years ago when Ebbesen was looking at making arrays of holes for quantum cavity effects. "While checking these arrays of little cavities, I discovered they had these transmission properties," he recounted.

"In the beginning a lot of people thought there was something wrong. How could they transmit wavelengths beyond their diameter?"

He worked on and off exploring the phenomenon. "The last two years we got all these things together."

Further experiments have increased the size of the circular apertures to up to 500 nm in diameter with similar results. Ebbesen and his colleagues believe that surface plasmons, that is electromagnetic fields associated with oscillations of charge on the surface of the metallic film, are the cause of this phenomena.

Periodicity, thickness, and type of metal were also used as variables in subsequent experiments, and the results were duplicated.

The process is begun when light incident on the film transfers its energy to surface plasmons on one side of the film. The energy flows to the opposite side of the film and is transferred to electromagnetic transmissions which propagate away from the film. The array of holes causes the surface plasmons to scatter off the wholes, shifting wave vectors. Some plasmons end up with a zero wave vector component along the surface, allowing the plasmons to couple with the light at normal incidence. After the coupling between surface plasmons, energy is coupled back to modes which propagate in free space.

"It took a long time to understand how this was possible," remarked Ebbesen. The research will also involve looking at applications; possibilities include optical switches, filters, near-field scanning optical microscopy, and photolithography. Even if the discovery is not commercialized, it has provided knowledge unknown a decade ago. "It's not obvious why you would set up an array of small holes to make good transmission, because that's not what you're taught," noted Ebbesen. "Maybe it was because I had no intent of measuring transmission.

"If it's obvious then it wouldn't be something new," he said. "Pure science, the amazing stuff always comes unexpectedly. Otherwise it wouldn't be so amazing."