At the height of the cold war, a Russian named Victor Veselago took a close look at something called the refractive index. At its most simple, the refractive index is the reason why the speed of light is what it is (even vacuum has a refractive index). What Veselago saw was that a positive refractive index is the product of the details of a material's response to the electric and magnetic components of the light field. But if one could play with material structures on the scale of the wavelength of light, it might be possible to create materials that had a negative refractive indexâ€”or no refractive index at all.
Just to drive home how hard it is to think about these meta-materials, let's go through a couple of examples. Imagine I have two blocks of material: a normal one with a refractive index of one, and another with a refractive index of -1. I shine light on both of them and observe the consequences. One thing that I would notice is that, although the speed of light was exactly the same in both materials, it seems to be in opposite directions. That is, the light in the negative index meta-material appears to be going backwards. Yet the direction of energy flow remains unchangedâ€”both materials transmit the energy from the light field in the same directionâ€”it's just that one makes the light moonwalk.
Well, that was a negative index meta-material, but what happens when the refractive index is zero? In this case, you might think that the speed of light would drop to zero, but it doesn't (that actually happens with a very large refractive index). No, in this case, the speed of light seems to be infinite. Funnily enough, though, this looks eerily like the light is standing still.
That's because, if I were to examine how the fields were distributed in space, I would find that they never change. But, yet again, the energy flowâ€”if I were to momentarily increase the light intensity and measure the time for that pulse to appear at the other end of the meta-materialâ€”is no faster than the speed of light in a vacuum.
To analyze the behavior, they basically performed the experiments described above: split light from a laser in two, send part through the zero index material and part through a normal material, then recombine the two laser beams at the end. The output depends on the effective distance the light has travelled through each sample. If both distances are the same, then they will add up in phase and give a bright light beam. On the other hand, if they are a half wavelength different, then the two light fields add up out of phase to give no light at all.
So, in one arm we have a normal silicon waveguide, and in the other we have layers of silicon interleaved with layers of silicon with holes drilled in it (the negative index material)â€”a zero index meta-material. If this is truly a zero index material, then, as far as the light is concerned, it isn't there at all. In other words, it should have a zero length, at least for the appropriate wavelengths of light.