Laser Cooling and Trapping Group

Metastable Xenon Project

Collisions in Optical Lattices

Optical lattices are periodic potential structures created by the interference of laser beams with the appropriate choice of polarizations. The light shift (AC Stark shift) potential felt by the atoms in the region of the interference varies periodically on a scale determined by the laser wavelength, creating potential wells in which atoms may become trapped. Three-dimensional lattices, formed from the interference of four linearly polarized beams, have wells separated by roughly one-half an optical wavelength, and atoms in those wells may be localized to significantly less than that. This can be expected to have a significant effect on the collision rate in the lattice, as atoms bound in discrete sites will be prevented from colliding with one another. Using ion detection to monitor the collision rate, we can estimate the rate at which atoms move from one site to another, and study the dynamics of atoms in lattices.

The lattice collision experiment consists of loading an ultra-cold sample of xenon atoms from our MOT into an optical lattice formed using ~1 W of laser light tuned several GHz to the red of the laser cooling transition. We hold the atoms in the lattice for some time, then turn off the lattice, and compare the rate of ion production in the lattice to the rate immediately after the lattice beams are shut off. As there is not time for the density of the sample to change significantly, this comparison provides a measurement of the effect of the optical lattice on the collision rate.

Figure Penning ionization counts measured during and after the lattice phase. The lattice is turned on at t = 0, maintained for 100 ms, then turned off abruptly. The sharp increase in the count rate after the lattice is turned off shows that the collision rate in the lattice is a factor of two lower than in the disordered sample. Inset: A close-up view of the "kink" in the collision signal. The 8 µs delay before observation of an effect reflects the time of flight of ions to our detector; the decrease and subsequent increase in the rate are explained by a simple theoretical model.

In these experiments, we see suppression of the collision rate by as much as a factor of two when the atoms are allowed sufficient time to reach thermal equilibrium and localize into the lattice potential wells. Before thermalization is complete, we see a significant increase in the rate, as atoms with energies too high to be bound, or atoms which "hop" from well to well, are "guided" by the potential toward bound atoms. From our measurements and a simple model, we extract an estimate of the rate at which atoms trapped in the lattice "hop" between wells.

Publications:

• J. Lawall, C. Orzel, and S.L. Rolston, Phys. Rev. Lett. 80, 480 (1998) [PDF Pre-print ( 358 kB)].

For more information, contact:

Steven Rolston
National Institute of Standards and Technology
PHYS A168
Gaithersburg, MD 20899
(301) 975-6581
e-mail: steven.rolston@nist.gov

Inquiries or comments: steven.rolston@nist.gov
Online: May 1998   -   Last update: June 2003 (format only)