Counting photons: squeezing a quantum limit.

Counting photons: Squeezing a quantum limit 

Even the best lasers aren't perfect enough for applications requiring extremely precise control of light. The problem lies in the quantum nature of light, which allows for a tiny amount of randomness that limits how much the noise, or fluctuations, in any signal can be lowered.

To overcome this problem, researchers have focused on reducing the uncertainty in one particular characteristic of a light wave at the expense of another, which would become more random. The light wave resulting from such a tradeoff is said to be in a "squeezed" state. By using the more predictable, or less noisy, component of this squeezed light, researchers can partly circumvent the system's inherent quantum fluctuations, permitting them to make more precise measurements.

One group has now devised a new way to squeeze light, using pulsed light from an ordinary laser to generate and then detect strings of photons to an unusually high precision. "These results takes us into a new regime of quantum optics," says Prem Kumar of Northwestern University in Evanston, Ill., who led the effort. The technique may eventually prove useful in optical communications systems by permitting light pulses to carry information more efficiently.

An ideal laser would produce light of a single wavelength in which every wave present is exactly in step with every other wave. Kumar and his colleagues, however, use a less coherent laser in which the waves are generally out of step. By shining pulses of green light from that laser into a crystal of potassium titanyl phosphate, they generate two separate beams of infrared light pulses.

In other words, the crystal in effect splits each green photon in the light pulses into a pair of infrared photons. "We know the two photons are generated simultaneously at the same location," Kumar says.

Measuring the number or location of photons in one beam necessarily modifies or destroys the photons in that beam but leaves photons in the other, duplicate beam unscathed. Because the beams are correlated, the measurement provides crucial information about the unaffected beam, which can then be used for some application.

"It gets around the problem of the random emission of photons in an ordinary laser where we don't know where the photonsare," Kumar says. "If we looked at either of the two beams, there would still be a lot of randomness in it. The point is that we have essentially duplicated the beam, so we can dispense with one to learn about the other."

The results show that a well-defined, coherent light source isn't essential for generating squeezed light. "It certainly broadens the range of optics where you might run into quantum light effects," says Richard E. Slusher of AT&T Bell Laboratories in Murray Hill, N.J.

Kumar and his colleagues describe their preliminary experiments in the Feb. 26 PHYSICAL REVIEW LETTERS and in a forthcoming issue of OPTIC LETTERS. They are now combining several of their ideas for producing and detecting pulses of squeezed light to achieve even better results.

"So far, they haven't generated much usable squeezing," sys Northwestern's Horace P. Yuen, who in 1976 first proposed the possibility of squeezing light. "But Kumar's group is now building a system tha has the potential of generating much larger squeezing."

"We're very excited about the next generation of experiments that we're doing," Kumar says. "We should get some new results within the next six months or so."

But the gap between laboratory results and practical applications, especially in optical communications, remains wide. "It takes time for people to invent all sorts of different ways to adapt the technique to a fiber-optic environment," Yuen says.

Some researchers are already trying to use techniques for squeezing light to make precision measurements. For example; Slusher is working with a group building a squeeze-light microscope that could detect the minuscule increase in thickness of a nerve membrane when a nerve is activated.