## Physicists Observe Alteration in Fundamental Interaction of Light and Matter

Scientists refer to the interplay of light and matter as fundamental atomic radiative processes. These processes span the range from the linear absorption and scattering familiar from everyday optics to nonlinear interactions on which for example the laser is based. This recently seen alteration in one of these atomic radiative processes opens up a new field in optical physics and points the way to new phenomena that before now have been inaccessible.

"Optical processes can be qualitatively different with nonclassical light," said H. Jeff Kimble, professor of physics at Caltech. "You can see things that would never be possible with classical light. Now we've realized the first example of this by using squeezed light to alter the way light interacts with atoms." The results of the experiment appeared in the November 6 issue of Physical Review Letters.

Classical vs. Nonclassical Light Light waves, even from a laser, are not perfectly regular. There are slight fluctuations both in amplitude and in phase for any beam of light. Physicists call these fluctuations "noise" because of their random character, with the fluctuations existing even in the absence of light. This random noise in utter darkness is a property of the quantum vacuum state.

Nature's rule about these quantum fluctuations is simply that the product of the amplitude noise times the phase noise must have a minimum value set by Planck's constant. This is the Heisenberg uncertainty principle for light. Quantum mechanics demands these fluctuations; they are intrinsic and fundamentally unavoidable.

For simplicity, Kimble assigns the vacuum-state fluctuations for both dimensions—amplitude and phase—a size of "one" in some arbitrary units. By various techniques, Kimble and other specialists in quantum optics are able to reduce the uncertainty in one of the dimensions below "one," with the result that the uncertainty in the other dimension expands, in order to keep the same minimum product. These states of light with uncertainty lower than "one" are called nonclassical states. They were first created in 1977 by Kimble, Mario Dagenais, and Leonard Mandel. In recent years, such states have been exploited to enhance measurement sensitivity beyond the standard quantum limits.

The Experiment Kimble and his experimental colleagues, Nikos Ph. Georgiades, a Caltech graduate student; Eugene S. Polzik of Aarhus University in Aarhus, Denmark; and Keiichi Edamatsu of Tohoku University in Sendai, Japan, were studying how the rate of two-photon excitation varies with incident light intensity, using nonclassical light and a collection of ultracold cesium atoms held stationary in an optical trap. Two-photon excitation means that two photons of different energy are needed to excite the atom. One boosts the atom from a low state (level 1) to a higher state (level 2), with the second photon lifting it finally to level 3. The scientists recorded the number of cesium atoms that reached level 3 as a function of the illuminating light.

For classical light, such as a laser or a light bulb, there is a quadratic (power of 2) relationship between the light intensity and how many atoms become excited. For example if the light intensity is cut to one-tenth of some initial value, the number of excited atoms will fall to one-hundredth of the initial value (one-tenth, squared). This quadratic relation is a well-known property of two-photon excitation, and was heretofore thought to be immutable.

The Result But the Caltech scientists found that when a low-intensity beam of nonclassical light—photon twins, to be specific—is directed toward the cesium atoms, the number of excited atoms is larger than that expected for a comparable flux of classical light. In marked contrast to the quadratic relationship observed with classical light, the scientists found that for low-intensity photon twins, the number of excited atoms depends on the intensity raised to a power that approaches 1.3. For example if the light intensity is reduced to one-tenth of some initial value, the number of excited atoms would be near one-twentieth of the initial value (one-tenth to the power 1.3), a much higher rate than seen with classical light.

A. Scott Parkins, a theorist now at the University of Waikato, in Hamilton, New Zealand, has worked with Kimble and his experimental collaborators to provide a quantitative description of the result. In qualitative terms, this unusual result can be understood as follows. When classical light excites cesium atoms, two photons of different wavelengths must arrive within a brief instant of each other. One photon bumps the atom from level 1 to 2, the second takes it from 2 to 3. As the intensity of the light drops and the photons become rarer, the chances of two photons arriving at just the right instant decreases rapidly (in fact, quadratically).

By contrast, in the experiment described here, the scientists employed photon twins with a combined energy just right to bump the atom directly from level 1 to 3. Because photon twins are produced and travel in pairs, they always arrive together at almost the same instant, and thus can take the atom from level 1 to 3 in one fell swoop. Since the probability of pairs of photons in this case becomes equal to the probability of singles, the two-photon rate approaches a linear (power of 1), as opposed to a quadratic, dependence.

Beyond the current work, the Caltech group is pursuing another manifestation of nonclassical atom-field interactions, namely the modification of atomic radiative relaxation. More broadly, the research effort is directed toward the exploration of optical interactions with the "normal" vacuum state replaced by the "quantum quietness" of a nonclassical field.

Contact: Jay Aller (818) 395-3631 aller@caltech.edu