First Experimental Demonstration of a Trapped Rainbow Using Silicon

Engineers want to use trapped rainbows to make everything from optical buffers to optical signal processors. But the trapping techniques have never been practical…until now

When a beam of light reflects off a surface, the different parts of the beam can interfere in way that shifts the point of reflection. This process is known as the Goos-Hӓnchen effect after the physicists who discovered it in the 1940s.

Usually, this shift is in the same direction as the propagating light. But in recent years, physicists have realised that there are certain classes of materials — or more precisely combinations of materials — in which the Goos-Hӓnchen effect is negative. In other words, the shift in the point of reflection would be against the direction of propagation.

That gave a group of physicists a curious idea. Back in 2007, they suggested that if the negative Goos-Hӓnchen effect could be made big enough, it ought to bring a ray of light to a standstill. In other words, the negative shift in the point of reflection in one direction would be big enough to cancel out the movement of light in the other. So the beam itself would remain stationary.

The team went on to describe how this effect could be used in a waveguide to trap a rainbow. Their idea was to taper the waveguide to trap ever smaller wavelengths of light.

Since then various groups have attempted to create metamaterials with the required properties to trap rainbows. Indeed, several groups have claimed successes of various kinds, although none of them ended up using the Goos-Hӓnchen effect.

Today, Rui Yang and pals at the University of Illinois at Chicago change all that. These guys have finally trapped a rainbow in this way for the first time. And they say that unlike earlier experiments, their method is efficient enough to be used in future computing components such as optical memories.

The key to the new work is the discovery of an entirely new way to produce the Goos-Hӓnchen effect. Instead of using exotic metamaterials, which are hard to manufacture and hugely inefficient at optical wavelengths, Yang and co say it is possible to produce the Goos-Hӓnchen effect by placing an interference grating made of silicon on the surface of an ordinary dielectric such as a silica. And by changing the spacings in the grating, it is possible to trap light of different wavelengths.

These guys first demonstrate numerically that they can trap lots of different wavelengths in this way and then go on to demonstrate the effect using a flat slab of silica with silicon gratings in contact with the slab at different positions along its length. “The result clearly gives a “rainbow” trapped along the device,” they say.

Removing the silicon grating from the silica waveguide releases the light again.

That’s an interesting result that has the potential to be extremely useful. The ability to trap light is widely considered an important enabling technology for purely optical computers. It allows, for example, optical buffers, optical signal processing and so on.

Until now, however, it has only been possible to trap light efficiently by passing it into a Bose-Einstein condensate at ultra-low temperatures. That’s not a method that lends itself to small scale optical processing-on-a-chip. And although metamaterials can trap rainbows, they also absorb most of the light that enters them, making the process highly inefficient.

The new technique suggests a way of trapping rainbows in devices based on silicon chip technology. Handy!

Ref: arxiv.org/abs/1410.8196: Realization of “Trapped Rainbow” in 1D slab waveguide with Surface Dispersion Engineering