American Institute of Physics
SEARCH AIP
home contact us sitemap
Physics News Update
Number 707, November 3, 2004 by Phil Schewe and Ben Stein

Accelerator for BECs

Two research groups have banged quantum gases together at record high velocities. Both groups begin by cooling clouds of rubidium atoms to ultralow temperatures. Next, through magnetic manipulation the clouds could be split into two separate clouds, each containing a native population with a characteristic spin value.

Physicists in the Netherlands (FOM Institute for Atomic and Molecular Physics and the University of Amsterdam) further cool the clouds to produce Bose-Einstein condensates (BEC) before using the same magnetic control over the atoms to urge the clouds back together again at an increasing speed. Earlier experiments had managed to "collide" separate BEC samples at slow speeds of mm/sec (slow in relation to the velocity of sound in the BEC---several mm/sec) in order to observe characteristic interference stripes, and affirm the intrinsic wavelike nature of BEC as a whole.

Now, the Dutch experiment is able to achieve speeds of 20 cm/sec; in effect their apparatus is a linear accelerator for BECs. The respective clouds are about 10 microns in size; the relative size of the clouds and their initial separation (up to record distances of 4 mm) is analogous to the separation of two tennis balls on opposite sides of a tennis court. When the two "tennis balls" collide, a spherical interference pattern shows up (see animation at staff.science.uva.nl/~walraven/walraven/Highlights.htm).

Why is the higher speed important? It’s because below sound speed, the superfluid BEC behaves like one giant matter wave, while above sound speed the BEC behaves like a collection of individual atoms. So in this experiment it is more accurate to think of 100,000 atoms (in the one cloud) scattering with 100,000 atoms (in the other cloud) rather then to think of two interacting clouds.

Furthermore, because the speeds are still slow, the atom-atom collision can still be thought of as being the collision of two waves (like separate ripples in a pond passing through each other). In other words, the experiment probes the interaction between atoms rather than between BECs.

In the BEC accelerator, matter waves of atom pairs are scattered out of the clouds at an energy of 10-7 eV. (Compare this to Fermilab’s 1012 eV energy scale.) These matter waves are a superposition of spherical-shaped "s" and dumbbell-shaped "d" waves and hence show quantum mechanical interference. This interference is being directly imaged (Buggle et al., Physical Review Letters, 22 October 2004; contact Jeremie Leonard, jleonard@science.uva.nl), and yields accurate measurement of the interaction properties between ultracold atoms.

Comparable observations are being reported by physicists from the University of Otago in New Zealand, although in this experiment the atoms were at 200-nanokelvin temperatures but did not constitute a BEC. (Thomas et al., Physical Review Letters, 22 October 2004; contact Niels Kjaergaard, nk@physics.otago.ac.nz; lab website at http://www.physics.otago.ac.nz/research/bec/Files/collisions/collisions.
html
).

Cooper Pairs Unpaired

In a low-temperature superconductor electrons don’t travel singly but in weakly tethered pairs, Cooper pairs. In a new experiment at the Forschungszentrum Karlsruhe in Germany, physicists have been able to send the two partners from Cooper pairs down separate wires spaced more closely than the effective size of the Cooper pairs themselves (see figure).

The Cooper pairs (which have the property that if one electron’s spin is up, then the spin of its partner must be down) start out in a piece of superconducting aluminum and proceed to a frontier where they can travel down either of two normally-conducting and magnetized iron wires. (In general, when Cooper pairs move from a superconducting into a normally-conducting material they can maintain their pair status for a bit into the new material---a distance referred to as the normal-metal coherence length---before breaking up.)

By magnetizing the wires so as to filter out pairings of any electrons that don’t have the characteristic Cooper opposite-spin-orientation, and by varying the distance between wires, and by measuring the resistance across the iron wires, the experimenters can learn specific things about the Cooper pairing mechanism (such as how large the pair is under various circumstances).

This work is part of the larger study of spintronics---the exploitation of electron spin for performing high-control electronics---and entangled states---the quantum behavior in which two spatially separated objects have a correlated behavior. (Beckmann et al., Physical Review Letters, 5 November 2004; contact Detlef Beckmann, detlef.beckmann@int.fzkde, 49-7247-82-6413).

Back to Physics News Update