Harry Potter's cloak? No. But it's still invisible

Call it the case of the vanishing bump.

For the first time, two teams have shown that, in principle, they can render a object invisible to light.

It's a small feat -- literally. The size of the experiments' pieces are measured in billionths and millionths of a meter. And it's not exactly visible light, but close, in the near-infrared. But the work has interesting implications -- ones that have nothing to do with Harry Potter-like cloaks or Star Trek's Romulan "wessels," at least for now.

For one thing, the approach could make it easier to design computer chips that use light, rather than electrons to do their work. For another, reversing the approach -- concentrating light rather than bending it in unusual ways -- could lead to more efficient solar cells.

What did the teams do? They took advantage of light's tendency to bend when it moves through mediums of different density -- like air and water. This is why a straw looks like it bends at the point where it enters water in a drinking glass.

Instead of air to water, however, the teams from Cornell and the University of California at Berkeley fooled light into behaving as though it was traveling trough different materials by introducing tiny structures into the silicon material they used.

It's part of a field known as metamaterials, where scientists fabricate existing materials in ways that give the materials properties they don't naturally have.

And for cloaks, this is taking place essentially in two dimensions, so think a very Flat Stanley here.

How to make a cloak

The Cornell team built its "cloak" from a tiny triangular sliver of silicon a couple of hundred nanometers thick. They dotted the cloak with tiny silicon pillars, arranged in a predetermined pattern. These pillars, in effect, were designed to fool light into acting as though it was moving through different materials with different densities, even though it was actually the same material.

They fabricated another triangle with pillars that were uniformly distributed. This triangle was designed to behave as material with uniform density.

Then, along the triangles' bases, they fabricated Lilliputian mirrors, one with a bump in it.

The test? If the pillars in the "forest" are uniformly distributed, a tightly formed beam of light should enter the forest on one side of the triangle. It should get spread out as it encounters the forest. When the light reflects off a bumpless mirror along the base, it should exit through the third side of the triangle as far wider beam, but with a fairly constant brightness along its width.

If the forest is uniform and the mirror has a slight bump, the light will leave the forest with a very dim spot along its width - in effect, the bump's shadow. But with the cloak and its specially arranged pillars, light should leave the forest with no shadow to reveal the bump's presence.

And that's exactly what happened.

It worked

Lucas Gabrielli, who is the lead author on the preprint describing these results, describes the effect in terms of a fun-house mirror. The mirror is warped to give the tall-skinny, short-fat effect in its reflection. Design a cloak, and even the warped mirror "will reflect light from its bumps as if it was a plain mirror," he says.

The Berkeley team used a similar experimental approach. But instead of fabricating pillars atop a silicon triangle, they in effect drilled holes in it. Then they shined their beam into one edge of the triangle. The holes were uniformly distributed throughout the triangle, except for a region near the bump in the mirror. There the team formed a rectangle -- a bit like the penalty box in front of a soccer goal -- filled with holes in a pattern designed to cloak the bump behind it.

As with the Cornell results, the Berkeley team's results represent a proof of concept for cloaking at visible wavelengths, notes Jason Valentine, a graduate student at UC Berkeley and one of three lead authors reporting the results in a recent issue of Nature Materials.

He explains that scaling the work up to a human scale would be extremely difficult given the materials involved. And both groups note that the next challenge will be trying to work with three-dimensional objects, not just two dimensional objects.

So, Harry's cloak or the Romulan warbird's vanishing act? "Those will stay in the science fiction section of the library" for quite some time, Mr. Valentine says.

But optical computing? Such approaches could help improve chip designs by placing groups of components, say transistors, on a chip where it would be best to place them, then route light around them by cloaking the transistor array.