Quantum nondemolition experiments.

Quantum nondemolition experiments

In the physics of the microcosm, a measurement disturbs what it measures. This is one of the consequences of the famous uncertainty principle. A second measurement of the same quality of the same system would not give the same result, because the back action of the previous measurement would have changed it. In the case of measurement of the amplitude of a beam of light, the back action amounts to total demolition: The measuring device destroys the photons or light quanta as they come in, converting their energy to something else.

In recent years scientists have been learning to get around this quantum demolition. Their results are having important repercussions in basic quantum physics and may have applications to communications technology. The first successful experiment was done by Marc D. Levenson, Bob Shelby and Steve Perlmutter of the IBM Research Division in San Jose, Calif., according to a review of the subject by Levenson.

The trick is to use two light beams in a single silica fiber of the sort used in optical communications. One beam measures the other through an interrelation mediated by the substance of the fiber. Passage of a light beam through the fiber changes the fiber's index of refraction slightly. The phase of a wave propagating through this kind of medium will change in relation to changes in the amplitude, and if two beams propagate through the medium simultaneously there is a reciprocal relationship by which changes in the amplitude of one produce changes in the phase of the other.

The experimenters combined light beams of two different colors with a prism, sent them through 100 meters of fiber and separated them at the end. They measured the phase change of one with an interferometer, and inferred a value for the amplitude change of the other from that. The amplitude of the other beam was measured directly in the usual kind of quantum demolition detector. The two results agreed.

A main purpose of such procedures is to measure noise -- that is, fluctuations in the amplitude of the light -- before the beam is sampled by some detector. One conceptually significant result that these experiments have already shown concerns "shot noise," the noise arising from quantum-mechanical fluctuations. One might think such noise arises from random fluctuations in the source of the light or from similar fluctuations in the detector. Levenson says the experiments prove both opinions wrong. The noise is "vacuum noise" inherent in the light itself, he says.