Light Exceeds Its Own Limit
Light Exceeds Its Own Speed Limit,
or Does It?

By JAMES GLANZ, New York Times

The speed at which light travels through a vacuum, about 186,000 miles per second, is enshrined in physics lore as a universal speed limit. Nothing can travel faster than that speed, according freshman textbooks and conversation at sophisticated wine bars; Einstein's theory of relativity would crumble, theoretical physics would fall into disarray, if anything could.

Two new experiments have demonstrated how wrong that comfortable wisdom is. Einstein's theory survives, physicists say, but the results of the experiments are so mind-bending and weird that the easily unnerved are advised--in all seriousness--not to read beyond this point.

In the most striking of the new experiments a pulse of light that enters a transparent chamber filled with specially prepared cesium gas is pushed to speeds of 300 times the normal speed of light. That is so fast that, under these peculiar circumstances, the main part of the pulse exits the far side of the chamber even before it enters at the near side.

It is as if someone looking through a window from home were to see a man slip and fall on a patch of ice while crossing the street well before witnesses on the sidewalk saw the mishap occur--a preview of the future. But Einstein's theory, and at least a shred of common sense, seem to survive because the effect could never be used to signal back in time to change the past--avert the accident, in the example.

A paper on the experiment, by Lijun Wang of the NEC Research Institute in Princeton, N.J., has been submitted to Nature and is currently undergoing peer review. It is only the most spectacular example of work by a wide range of researchers recently who have produced superluminal speeds of propagation in various materials, in hopes of finding a chink in Einstein's armor and using the effect in practical applications like speeding up electrical circuits.

"It looks like a beautiful experiment," said Raymond Chiao, a professor of physics at the University of California in Berkeley, who, like a number of physicists in the close-knit community of optics research, is knowledgeable about Dr. Wang's work.

Dr. Chiao, whose own research laid some of the groundwork for the experiment, added that "there's been a lot of controversy" over whether the finding means that actual information--like the news of an impending accident--could be sent faster than c, the velocity of light. But he said that he and most other physicists agreed that it could not.

Though declining to provide details of his paper because it is under review, Dr. Wang said: "Our light pulses can indeed be made to travel faster than c. This is a special property of light itself, which is different from a familiar object like a brick," since light is a wave with no mass. A brick could not travel so fast without creating truly big problems for physics, not to mention humanity as a whole.

A paper on the second new experiment, by Daniela Mugnai, Anedio Ranfagni and Rocco Ruggeri of the Italian National Research Council, described what appeared to be slightly faster-than-c propagation of microwaves through ordinary air, and was published in the May 22 issue of Physical Review Letters.

The kind of chamber in Dr. Wang's experiment is normally used to amplify waves of laser light, not speed them up, said Aephraim M. Steinberg, a physicist at the University of Toronto. In the usual arrangement, one beam of light is shone on the chamber, exciting the cesium atoms, and then a second beam passing thorugh the chamber soaks up some of that energy and gets amplified when it passes through them.

But the amplification occurs only if the second beam is tuned to a certain precise wavelength, Dr. Steinberg said. By cleverly choosing a slightly different wavelength, Dr. Wang induced the cesium to speed up a light pulse without distorting it in any way. "If you look at the total pulse that comes out, it doesn't actually get amplified," Dr. Steinberg said.

There is a further twist in the experiment, since only a particularly strange type of wave can propagate through the cesium. Waves Light signals, consisting of packets of waves, actually have two important speeds: the speed of the individual peaks and troughs of the light waves themselves, and the speed of the pulse or packet into which they are bunched. A pulse may contain billions or trillions of tiny peaks and troughs. In air the two speeds are the same, but in the excited cesium they are not only different, but the pulses and the waves of which they are composed can travel in opposite directions, like a pocket of congestion on a highway, which can propagate back from a toll booth as rush hour begins, even as all the cars are still moving forward.

These so-called backward modes are not new in themselves, having been routinely measured in other media like plasmas, or ionized gases. But in the cesium experiment, the outcome is particularly strange because backward light waves can, in effect, borrow energy from the excited cesium atoms before giving it back a short time later. The overall result is an outgoing wave exactly the same in shape and intensity as the incoming wave; the outgoing wave just leaves early, before the peak of the incoming wave even arrives.

As most physicists interpret the experiment, it is a low-intensity precursor (sometimes called a tail, even when it comes first) of the incoming wave that clues the cesium chamber to the imminent arrival of a pulse. In a process whose details are poorly understood, but whose effect in Dr. Wang's experiment is striking, the cesium chamber reconstructs the entire pulse solely from information contained in the shape and size of the tail, and spits the pulse out early.

If the side of the chamber facing the incoming wave is called the near side, and the other the far side, the sequence of events is something like the following. The incoming wave, its tail extending ahead of it, approaches the chamber. Before the incoming wave's peak gets to the near side of the chamber, a complete pulse is emitted from the far side, along with a backward wave inside the chamber that moves from the far to the near side.

The backward wave, traveling at 300 times c, arrives at the near side of the chamber just in time to meet the incoming wave. The peaks of one wave overlap the troughs of the other, so they cancel each other out and nothing remains. What has really happened is that the incoming wave has "paid back" the cesium atoms that lent energy on the other side of the chamber.

Someone who looked only at the beginning and end of the experiment would see only a pulse of light that somehow jumped forward in time by moving faster than c.

"The effect is really quite dramatic," Dr. Steinberg said. "For a first demonstration, I think this is beautiful."

In Dr. Wang's experiment, the outgoing pulse had already traveled about 60 feet from the chamber before the incoming pulse had reached the chamber's near side. That distance corresponds to 60 billionths of a second of light travel time. But it really wouldn't allow anyone to send information faster than c, said Peter W. Milonni, a physicist at Los Alamos National Laboratory. While the peak of the pulse does get pushed forward by that amount, an early "nose" or faint precursor of the pulse has probably given a hint to the cesium of the pulse to come.

"The information is already there in the leading edge of the pulse," Dr. Milonni said. "You can get the impression of sending information superluminally even though you're not sending information."

The cesium chamber has reconstructed the entire pulse shape, using only the shape of the precursor. So for most physicists, no fundamental principles have been smashed in the new work.

Not all physicists agree that the question has been settled, though. "This problem is still open," said Dr. Ranfagni of the Italian group, which used an ingenious set of reflecting optics to create microwave pulses that seemed to travel as much as 25% faster than c over short distances.

At least one physicist, Dr. Guenter Nimtz [[umlaut over u]] of the University of Cologne, holds the opinion that a number of experiments, including those of the Italian group, have in fact sent information superluminally. But not even Dr. Nimtz believes that this trick would allow one to reach back in time. He says, in essence, that the time it takes to read any incoming information would fritter away any temporal advantage, making it impossible to signal back and change events in the past.

However those debates end, however, Dr. Steinberg said that techniques closely related to Dr. Wang's might someday be used to speed up signals that normally get slowed down by passing through all sorts of ordinary materials in circuits. A miniaturized version of Dr. Wang's setup "is exactly the kind of system you'd want for that application, Dr. Steinberg said.

Sadly for those who would like to see a computer chip without a speed limit, the trick would help the signals travel closer to the speed of light, but not beyond it, he said.