Focus: Entangled Atomic Beams

Published November 3, 2000  |  Phys. Rev. Focus 6, 20 (2000)  |  DOI: 10.1103/PhysRevFocus.6.20

Squeezing and Entanglement of Atomic Beams

L.-M. Duan, A. Sørensen, J. I. Cirac, and P. Zoller

Published November 6, 2000
Figure 1
Phys. Rev. Lett. 85, 3991 (2000)

Entanglement mass-produced. Collisions between spin 0 atoms in a specially prepared Bose-Einstein condensate would produce oppositely directed beams of entangled atoms, which could be used for investigations of entanglement and quantum information theory. In the past only beams of entangled photons have been produced.

According to quantum mechanics, a pair of specially prepared particles can be far apart but retain a bizarre “entangled” relationship, where measurement of one has an immediate impact on the state of the other. Most experiments with entanglement have used isolated pairs of photons or two continuous light waves consisting of many entangled photon pairs. In the 6 November PRL, two teams independently propose a simple way of creating two entangled beams of atoms, using a Bose-Einstein condensate–a cloud of atoms cooled to its quantum mechanical ground state. These entangled beams would allow new tests of the weirdest ideas in quantum mechanics and new avenues of research into the fundamental concepts of quantum information, the basis for quantum computing and quantum cryptography. Another more distant application may be atomic clocks that beat conventional limits of precision.

Researchers can create a pair of photons in which one is horizontally polarized and the other vertically polarized, but the polarization of each one remains unknown–and according to quantum mechanics, undefined–until one is measured. If entangled photon pairs are produced at a high enough rate, the individual particles merge into continuous light beams which have been used to test the so-called Bell inequalities, the fundamental equations describing entanglement. Researchers have also used them to look at entanglement-related effects such as quantum state teleportation. Atoms are thought to be better suited to such applications than photons because almost every atom can be detected, and because they should be less prone to “disentangling” in response to outside disturbances. Although a few labs have generated pairs of entangled atoms–one spin up, the other spin down–none have produced entangled atomic beams, which could maximize the advantages of atoms over photons.

The two teams–one based at the University of Arizona in Tucson and one at the University of Innsbruck, Austria–propose essentially the same recipe. Imagine atoms that have three spin states: 0, +1, and -1. Start with a Bose-Einstein condensate of atoms entirely in the spin 0 state, and raise its energy (with microwaves) above that of the +1 and -1 states. The condensate will no longer be in the ground state, and atoms will tend to drop to the lowest energy states via spin-conserving collisions. Pairs of spin 0 atoms will collide and generate pairs of entangled spin +1 and -1 atoms, and these will increase their kinetic energy (speed) to compensate for their loss of “spin energy.” After each of these collisions, the two fast moving entangled atoms will then head straight out of the atom trap in opposite directions.

Han Pu of Arizona says that for a cigar shaped condensate, most of the entangled pairs will emerge as two beams along the cigar axis, and additional traps at each end could capture a large number of escaping atoms. These clouds could serve as plentiful sources for experiments. The Innsbruck team, led by Ignacio Cirac and Peter Zoller, calculates the unusual statistical properties of the beams: The total spin is exactly zero, and the fluctuations about this value are much smaller than would be expected for any classical set of atoms.

Mikhail Lukin of the Harvard-Smithsonian Center for Astrophysics says the results “may open a new chapter in the physics of non-classical states.” According to Lukin, the suppressed fluctuations may lead to atomic clocks with unparalleled precision, since today’s best clocks are limited by fluctuations. As for the experimental implementation of the proposals, Pu says at least one lab has already gone part-way, and experimentalists need only use current technology. “All the pieces are already there,” says Pu.