METASTABLE ATOM TRAP

Using a combination of circularly polarised laser beams and a quadrupole magnetic field, as shown in the right figure, atoms are confined near the zero point of the magnetic field and cooled to a temperature of about 1 mK. A real-life picture of the trap chamber is shown in the left picture. As the depth of the trap thus formed is only about 1 K, the atoms emerging from a liquid nitrogen cooled source first have to be slowed and collimated in the bright beam machine. The trapped atoms can be observed by fluorescence; however the fluorescence on the trapping transition (23S->23P transition at 1083 nm) is difficult to detect with a standard CCD camera. Therefore, we have used fluorescence on other transitions to make the atoms visible, as shown below.

Trap Parameters

Trapping beam size : 1"

Laser intensity I/I0 = 22 p. beam (I0 = Saturation intensity of the 23S -> 23P transition)

Trapping beam total power : 24 mW

Background pressure : 1x10-9 torr

Trap diameter : 2.1 mm, FWHM (typical)

Temperature : ~ 1mK (from trap expansion, see below)

Magnetic field gradient : 20 - 40 G/cm

Trap laser detuning range : D = [-10G; -30G]
 
 

Currently, experiments investigating the influence of these other light colours on trap collision experiments are underway.

Trap detection - RF Spectroscopy

Detecting the trap is DIFFICULT because a CCD video camera or any other silicon based device is very insensitive to 1083 nm (infrared) light.

Absorption of a probe laser beam is not a viable option since the amount of light absorbed by the trap is a small percentage and so, to detect efficiently the trap, a quite intense beam would be needed.
Such a beam would apply pressure upon the trap, distorting it and changing significantly the density.

An alternative is to use germanium based devices; video cameras in particular are slow and heating produces a halo which makes the image difficult to interpret.

In addition to this we had the need of detecting any change in the trap density and number of trapped atoms in real time  as some of the parameters change in a period shorter then the trap life time.
 

The phase contrast (RF) spectroscopy provides a non-destructive tool for monitoring the trap conditions.

A probe laser beam emitted by a semiconductor (diode) laser at 1083 nm passes through and electro optic crystal mounted inside a metal cavity containing a phase modulator. The phase modulator uses an electro-optic effect by applying an oscillating field to a LiNbO3 crystal and inducing phase modulation in the light wave passing through.
In frequency space this is represented by two anti phase side bands equally distant from the carrier by 860 MHz.
 
 
 

Fig. 1 - Schematic diagram of the experimental configuration
 

Fig. 2 -  Electric field vs frequency diagram. Note that the two side bands are off phase by 1800 deg.

The modulated laser is sent through the trap and the light transmitted collected onto a 1GHz AsGe photodiode, highly efficient at 1083 nm.

The signal produced is then analysed with a spectrum analyser

The photodiode detects the beat note between the carrier and the right side band + the beat note between the carrier and left side band.
With no phase shift, the two signals cancel out perfectly, but, at n slightly different from resonance, the will not.
The phase shift is the result of a change in the index of refraction-  maximum at the atomic resonance.

To avoid influencing the trap, the probe beam is detuned from resonance.
In the limit where W/g»1, W/g »1 and D/w «1  the signal from Phase Modulation (PM) is

Im = side band intensity
Il = carrier intensity
PM=Phase modulation signal
ABS=Absorption signal

For the absorption  (ABS),

Hence

 To minimize the effect on the trap, the carrier detuning is set to be ~ 1GHz from the 23S -> 23P (trapping transition, see metastable He energy levels diagram)

The RF spectroscopy is also used as a real time monitor for the trap conditions: allows fast response and great sensitivity.
Even in the case of low number of atoms trapped, we will be able to detect them. Fitting the decay curve of the expanding trap we were able to extract a preliminary measure of the trap temperature, which is close to 1 mK, in agreement with theoretical predictions.

Fig. 3 - Dc signal from the spectrum analyser as seen on the oscilloscope when the probe laser is scanning through resonance.