

The work of Segev's group might have remained theoretical, but by coincidence, a group led by John Dudley at the University of Franche-Comté in Besançon, France, has been performing its own experiments on self-bending light. People thought that there was no proper shape, that the solution would always fall apart-but we've shown that that is wrong." "If you want to go to large angles, must have the proper shape. "The Airy function is a solution for an approximate case," says Segev. After laborious mathematics and guesswork, the researchers found solutions to Maxwell's equations that precisely describe the initial phases required for truly self-bending light, as they report this week in Physical Review Letters. So his group turned to Maxwell's equations, the 150-year-old quartet of mathematical formulas that describe the propagation of electromagnetic waves such as light. The problem with the Airy function, says Segev, is that the shape of its oscillations specify the right phases only at small angles at angles much greater than 8°, the shape becomes a crude approximation. Now physicists Mordechai Segev and colleagues at Technion, Israel Institute of Technology, in Haifa say they have a recipe for making light self-bend through any angle, even through a complete circle. The Airy function, which contains rapid but diminishing oscillations, proved an easy way to define those initial phases-except that the resultant light would bend only up to about 8°. By carefully controlling the initial position of the wave peaks-the phase of the waves-at every step along the strip, it is possible to make the light traveling outward interfere constructively at only points on a curve and cancel out everywhere else. Now, imagine light emitted from a wide strip-perhaps a fluorescent tube or, better, a laser whose output has been expanded. For example, a peak passing a trough cancels each other out to create darkness a peak passing another peak "interferes constructively" to create a bright spot. How did this self-bending work? Light is a jumble of waves, and their peaks and troughs can interfere with one another. That work was largely ignored until 2007, when Demetri Christodoulides and other physicists at the University of Central Florida in Orlando generated optical versions of Airy waves by manipulating laser light, and found that the resultant beam curved slightly as it crossed a detector. In the late 1970s, physicists Michael Berry at the University of Bristol in the United Kingdom, and Nandor Balazs of the State University of New York, Stony Brook, discovered that a so-called Airy waveform, a wave describing how quantum particles move, can sometimes bend by a small amount. In each instance, light-bending has an external cause: For water, it is a change in an optical property called the refractive index, and for stars, it is the warping nature of gravity.įor light to bend by itself, however, is unheard of-almost.

Out in space, light rays passing near very massive objects such as stars are seen to travel in curves. When light rays pass from air into water, for instance, they take a sharp turn that's why a stick dipped in a pond appears to tilt toward the surface. The effect is actually an optical illusion, although the researchers say it could have practical uses such as moving objects with light from afar. But now researchers have shown that light can also travel in a curve, without any external influence.

Any physics student knows that light travels in a straight line.
