Yb Trapping Overview

Yb Project Overview

Atomic Yb is a rich system in which to explore specific atomic properties and more fundamental physical questions. Like the alkaline earth elements, Yb has a J = 0 ground state and many stable isotopes with comparable relative abundance. The lack of magnetic substructure greatly simplifies both the interpretation of experimental results and theoretical calculations. The ability to perform experiments with different isotopes opens many experimental possibilities, including the study of BEC mixtures and sympathetic cooling of fermions, as well as provides theorists with more data to test their predictions. In particular, Yb is an excellent candidate for further investigation of Doppler and sub-Doppler cooling mechanisms, cold and ultra-cold collisions, photoassociation, Bose-Einstein condensation and degenerate Fermi systems, symmetry violation, and high precision frequency standards.

In our experiment, we use the allowed electric dipole transition from the ground state to the ¹P₁ for our Zeeman slower and blue magneto-optical trap (MOT). The blue MOT lifetime is limited to less than a second by branching to the ³D states, so we transfer the atoms to a second MOT which uses the green intercombination line between the ground state and ³P₁. This transition, only allowed because the spin-orbit interaction mixes in a bit of the ¹P₁ state, is much weaker, leading to much lower temperatures than in the blue MOT. The lifetime of our green MOT is limited by collisions with background gas.

Blue MOT

I_sat = 66 mW/cm²
T_D = 690 mK
lifetime ~ 800 ms

Green MOT

I_sat = 0.14 mW/cm²
T_D = 4.4 mK
lifetime ~ 60 s

Summary of Cooling and Trapping Apparatus:

Blue MOT:

Ar⁺ laser pumps Ti:Saph laser
for 798 nm light
798 nm light is frequency-doubled to 399 nm by LBO crystal in ring cavity

Green MOT:

Ar⁺ laser pumps ring dye laser to produce 556 nm light
Pound-Drever-Hall lock to Fabry-Perot cavity narrows laser linewidth to 100 kHz
Atomic beam lock prevents
slow frequency drift

Imaging System:

Time of flight images captured
by CCD camera

Laser system detail

Our group has recently completed a study of cooling mechanisms in the green MOT. In addition to studying pure Doppler cooling in the even isotopes, we made one of the first observations of sub-Doppler cooling mechanisms using only nuclear spin in the odd isotopes. The top curve on the right shows temperature as a function of laser intensity for ¹⁷⁴Yb. Because the even isotopes of Yb have zero ground state spin, we expect to see pure Doppler cooling. The temperature depends linearly on intensity, as expected for Doppler cooling; however, the slope of the line is steeper than theory predicts. These results are consistent with observations in Sr and alkali atoms and are still not well understood. The deviation from linearity at the lowest temperatures is due to the influence of gravity.

The lower graph on the right shows temperature as a function of intensity for the odd isotopes. In these isotopes, we see polarization gradient cooling due to interaction of polarized trap laser light with hyperfine sublevels in the atoms. One might naively expect to see pure corkscrew cooling in the MOT because it consists of beams of counterpropagating, oppositely directed circularly polarized light. In fact, the three-dimensional MOT has a comparable amount of Sisyphus and corkscrew gradients, and Sisyphus cooling dominates. Our results are consistent with Sisyphus cooling theory: temperature depends linearly on intensity and higher nuclear spin results in a lower slope. More detail and a comparison to theory can be found in PRA 68 011403(R) (2003).

All temperatures were calculated using a time of flight method which compares absorption images of the atoms as they expand when the trap is turned off.

We are currently exploring the possibility of creating an extremely precise atomic frequency standard using the¹S₀ - ³P₀ transition. This transition is forbidden in the even isotopes, but in the odd isotopes the hyperfine interaction allows a weak electric dipole transition to proceed. The resulting narrow line is an attractive candidate for an optical clock. In general, narrow optical transitions are of interest as clocks because their line Q is high compared to microwave transitions. Ensembles of cold neutral atoms have a high signal to noise ratio compared to single ions, but have thus far been limited to an interrogation time of a few milliseconds. The model we're pursuing, first presented by Katori at the Sixth Symposium on Frequency Standards and Metrology in 2001, is to hold the atoms in an optical lattice operating at a "magic wavelength" at which the clock transition frequency is only minimally affected by the trapping light fields. This will allow longer interrogation times and has the immediate advantage of eliminating the Doppler effect by holding the atoms in Lamb-Dicke confinement.