Atom Cooling and Trapping Experiments

at Manchester

Page constructed by Andrew Murray

This page describes the atom trapping and cooling experiments being conducted in Manchester. Details of the AC-MOT as used for collision experiments can be found here.

Papers from the group can be found here:

1.0 Introduction

The purpose of this page is to detail the Atom trapping and Cooling experiments being conducted at Manchester.

The new apparatus which has been developed is aimed at providing a high density source of atoms for several different experiments in cold atoms. The apparatus produces a bright, cold atom source which competes with the best in the world for speed of accumulation of cold atoms into a Magneto-Optical Trap (MOT), and for the lifetime of atoms that are subsequently cooled and trapped.

In these experiments, new types of collision processes are to be studied, including electron collisions from cold atoms, BEC's and from slow atomic beams. These experiments will exploit the momentum transferred in the collision process to ascertain new information about the target interactions, and to study collisions with cold and ultra-cold targets for the first time.

2.0 Overview of Cold Atom Experiments in Manchester

The Cold Atom Facility which has been developed in Manchester is a bright source of cold atoms, and is shown in Figure 1. The apparatus consists of a recirculating oven source, which produces a quasi-collimated beam of atoms into the laser collimation stage. Initial collimation of the atomic beam is achieved in the recirculating oven source through the use of small apertures as is shown in figure 2. In the first experiments using this Facility, potassium atoms have been cooled and trapped.

FIG 1a. The cold atom facility that has been developed at Manchester, showing the various components of the apparatus and the laser beams which are used to cool and trap the atoms emitted from the source.

FIG 1b. Photograph of the actual apparatus which has now been commissioned.

Atoms emitted from the source can be laser collimated in the second stage of the experiment, which consists of a 6-way 200mm 316LN cross attached to the source chamber. Large windows are attached to the cross to allow the collimating laser beams to enter this region.

Laser collimation can be accomplished either using an optical molasses, through a 2D MOT or by exploiting the transverse momentum transferred to the atomic beam as it traverses this region using angled mirrors. In the present setup, optical molasses has been used to increase the collimation of the atomic beam prior to the Zeeman slower.

Atoms from the collimated source that enter the Zeeman slower (figure 5) are longitudinally cooled and compressed in phase space using a blue detuned laser which passes along the axis of the Zeeman slower. Atoms travelling at velocities up to ~800m/s can be captured and cooled by the laser beam within the Zeeman slower. The atoms exit the Zeeman slower all with a velocity ~30m/s - 50m/s (depending on the setup of the Zeeman slower magnetic field conditions) and are then subjected to the laser and magnetic fields in the Magneto-Optical Trap (MOT1) (figure 6). Once the atoms enter MOT1, they experience frictional force due to the laser interaction and a restoring force due to the magnetic field, so that they rapidly cool to a temperature ~250 micro-Kelvin around the centre of the trap.

The design of the apparatus in Manchester allows atoms to be cooled, trapped and accumulated very rapidly from the source, which operates continuously.

It has been demonstrated that the trap density reaches saturation (where no more atoms can be trapped due to radiation trapping and other restrictions) within 1 second. The density of trapped atoms in MOT1 saturates at  ~ 1010 atoms/cm3, and the atoms can be held in the trap for more than 400 seconds before collisions with the background gas (typical pressures in the MOT1 chamber are ~7 x 10-11 torr) eliminate them from the trap.

The MOT1 chamber has been designed with numerous CF-70 ports to allow additional experiments to be connected to the chamber. The aim of this design is to facilitate rapid trapping of the cold atoms, then transfer of these cold atoms to different experiments that are connected to the MOT1 chamber via these ports. In this way, atoms accumulated in MOT1 can be transferred to several different experiments where they can be recaptured for further study, or experiments can be directly conducted on the trapped atoms in MOT1.

Since the ratio of accumulation time to trapping time is 400:1, the apparatus will be able to continuously load different cold atom experiments from the main source while maintaining the maximum trapping density which can be achieved in each of these experiments. In this way, experiments which have low cross sections (such as collisions experiments with electrons, neutrals or ions) will be possible in the near future.

3.0 The Atomic Beam Source

One of the goals in the design of this apparatus was to produce a well-collimated beam of atoms from the atomic beam source prior to laser collimation (see figure 1). By ensuring this criteria is met, the vacuum chambers are far less polluted by unwanted atoms that are emitted from the source.

There are several different types of atomic beam oven that can be used in these types of experiments, ranging from a simple crucible that directs atoms through a hypodermic needle (and hence produces a beam that is quite divergent) through capillary array sources (which are unsuitable for metal vapour targets as needed here), on through passively collimated sources as is used in Manchester for super-elastic scattering experiments from calcium and finally on to the most sophisticated (and most complex) sources which use a recirculating arrangement for collimation and recycling of the material under study.

It is the recirculating type of oven that was chosen for these studies.

Recirculating oven sources have several advantages over other sources (if they are suitable).

Firstly, since the source only produces atoms which enter the atomic beam, there is very little waste of material. This means that the source can be operated for very long periods of time without the need to refill the crucible.

As an example, the potassium source used in these cold atom experiments in Manchester has been operating for over 2 years without the need to refill the crucible.

Secondly, the use of a condensation chamber surrounding the emitting nozzle means that vapour from the oven is re-condensed onto the inner walls of this chamber, to be fed back to the crucible. The physical size of the condensation crucible and the output aperture then defines the collimation of the atomic beam that exits and therefore becomes the emerging atomic beam. Since no atoms are emitted apart from those in the atomic beam, contamination of the vacuum chamber is virtually eliminated.

FIG 2. The recirculating Oven source used in the Cold Atom Experiments in Manchester.

Figure 2 shows the recirculating oven that has been designed and commissioned in Manchester. The crucible is loaded with potassium (melting temperature ~337K) and the temperature of different components of the source are closely controlled. In this way, atoms are emitted from the nozzle into the condensation chamber, where they are re-condensed onto the walls of this chamber.

The condensation chamber is held at a temperature ~340K so that the potassium on the walls is in a liquid form. Surface tension ensures that the liquid potassium runs down the walls of the condensation chamber and is gravity fed back to the crucible (which operates at a temperature of ~650K).

Potassium vapour that is directed through the output aperture of the condensation chamber becomes the atomic beam which is used in the experiments.

The estimated density of the emerging beam is ~1011 - 1012 atoms/cm3, whereas the angular divergence of this beam is ~ 2 degrees.

Close control of the recirculating oven temperature is achieved by using several constant current supplies which feed the tungsten heaters as shown in figure 2. The temperature of the condensation chamber is accurately controlled by passing a heated high temperature silicon oil through a sealed chamber within the walls of the chamber as shown. In this way, the complex temperature conditions which are essential for correct operation of this source can be maintained.

Although not yet tested with Rb, it is fully expected that this atomic beam oven can be used with equal success using Rubidium (melting temperature 312K) as the chosen source.

Rubidium will be used for the production of Bose Einstein Condensates (BEC's) and for other experiments in the near future.

Details of this recirculating beam source are currently being written as a paper for submission to an appropriate journal.

4.0 The Zeeman Slower

Following laser collimation using an optical molasses (figure 1), the atomic beam enters the Zeeman slower. Since atoms emerging from the oven have a wide range of different velocities (as dictated by the Maxwell Boltzmann distribution), it is necessary to slow these atoms down so that they can be captured by the MOT. Further, it is important to also compress the velocity distribution of these atoms so that they have a much narrower distribution at the exit of the slowing tube.

The effective way to slow atoms down using lasers is to adopt the process of laser cooling where the spontaneous emission process is used as a mechanism to provide a dissipative force.

FIG 3. The cooling process used to slow atoms down.

Figure 3 shows the process which is used.

An atom initially travelling with momentum pA = MAv is subjected to laser radiation directed in the opposite direction, where MA is the mass of the atom and v is the velocity of the atom. The laser frequency (wavelength) is set to be resonant with the atomic transition so that there is a high probability of absorption of a photon from the laser field by the atom, which is then promoted to an excited state.

The linear momentum of the photon is given by p = h/lambda, where lambda is the wavelength of the incident laser radiation, and h is Planck's constant.

For potassium, lambda ~ 766nm, whereas h = 6.626 x 10-34 m2 Kg/s.

Since linear momentum must be conserved, absorption of a photon from the field means that the atom must slow down.

Once the atom is in an excited state, two possible processes can occur.

Firstly it is possible for a second photon from the laser field to stimulate the atom to decay back to the ground state. In this case (stimulated emission process) the photon from the atom emerges in the same direction as the photon in the laser field. In this case, the momentum of the atom will return back to its original value pA, and there is NO slowing of the atom by the laser beam.

Secondly, the atom may relax back to the ground state spontaneously (spontaneous emission process). In this case, the atom has no 'memory' of how it was excited, and so does not know the direction of the incident photon. Hence the spontaneously emitted photon can be emitted in ANY direction.

IF the photon is emitted in the same direction as the incident photon, then the momentum of the atom will return back to it's initial value prior to the interaction pA = MAv, and so the atom velocity does not change.

IF the photon is emitted in the OPPOSITE direction to the incident photon, then the momentum of the atom will be reduced further, so that after the interaction the atom momentum will be pA = MAv - 2h/lambda. The atom will SLOW DOWN.

Since the process of spontaneous emission is completely random, this means that ON AVERAGE for each photon that is absorbed, there is a NETT change of linear momentum of the atom given by h/lambda.

By repeating this process many thousands of times, the atoms will slow down until they are stopped.

The process of compression of the velocity distribution occurs since atoms with different velocities will only come into resonance with the laser beam at certain times during the cycling process. By ensuring that either the laser frequency or the atomic transition frequency is time (or space) dependent, compression can occur.

There are several methods to slow and compress the velocity distribution of these atoms. One method that is used is to 'chirp' the slowing laser beam (figure 1) so that as the atoms slow down they remain in resonance with the laser beam.

This process can be used to produce a pulsed source of slow atoms.

The method adopted here changes the internal states of the atoms by application of a magnetic field, which perturbs the atomic transitions which are coupled together by the laser radiation field. As the atoms slow down, the magnetic field is then adjusted so that the atoms remain in resonance with the laser beam (or they will Doppler shift out of resonance).

This process produces a CONTINUOUS beam of slow atoms from the Zeeman slower.

FIG 4. The magnetic field profile required for the Zeeman slower.

Figure 4 shows the type of magnetic field that is required to ensure the atoms can be effectively slowed and cooled. A set of 'disruption coils' are used at the exit of the slower to ensure atoms can escape into the MOT chamber.

FIG 5. Picture of the Zeeman slower shown with the cooling coils surrounding the current carrying coils.

Figure 5 and figure 1 shows Zeeman slower that has been designed for the experiments in Manchester.

3,500m of rectangular section aluminium wire is used to provide the field, wound onto the flight tube as 5 individual coils.

These coils are operated by 5 x 20A constant current supplies whose current can be tailored so as to produce the required field profile along the flight tube.

Four additional coils are located at the exit of the Zeeman slower which disrupt the field so as to release atoms into the MOT chamber. These coils are also supplied by 20A constant current supplies.

The Zeeman slower dissipates ~5,000W when operated at full current, and cooling of the slower is implemented by the external copper coils and by an internal water jacket next to the flight tube.

5.0 The MOT Chamber.

Atoms from the Zeeman slower enter the MOT chamber as seen in Figure 1. Inside this chamber is a twin set of coils setup to produce zero field at the centre of the coils (anti-Helmholtz configuration), as shown in figure 6.

FIG 6. The anti-Helmholtz coils setup in the MOT chamber at Manchester.

Six laser beams of appropriate polarization are directed at the centre of the coils (the trapping region) which act as an OPTICAL MOLASSES to the atoms which enter this region, as shown in Figure 7.

The laser interaction acts as the velocity dependent frictional force that slows the atoms down, in the same way as inside the Zeeman Slower.

FIG 7. Setup for the MOT showing the coils and laser beams that are adopted.

Since the molasses component is velocity dependent, as the atoms slow down they will gradually stop moving (within these approximations).

The magnetic field from the anti-Helmholtz coils in combination with the laser beams ensures that the force acting on the atoms also has a restoring component that drives the atoms back to the centre of the trapping region (this force depends on the distance from the centre of the trap).

Since the restoring component is distance dependent, atoms subjected to this component will perform simple Harmonic motion (within these approximations).

The combination of restoring force and frictional force means that the motion of the atoms will be damped, leading to the atoms all stopping in the centre of the trap (with velocity = 0 m/s).

This does not happen, as the atoms may still absorb and emit photons from the laser field, leading to a minimum velocity that is given by the recoil momentum of the atoms undergoing these absorption and emission processes (the quantum limit).

For potassium atoms in the experiment in Manchester, the temperature of the atoms in the trap has been measured to be ~250 micro-Kelvin.

FIG 8. Picture of the atoms trapped by the laser and magnetic fields inside the MOT chamber

FIG 9. Trapped atoms showing the direction of atoms as they enter the MOT from the Zeeman Slower.

FIG 10. Movie of the atoms being trapped in the MOT from the Zeeman slower (first successful trapping with an accumulation time ~1sec, and a  lifetime ~10 seconds).

6.0 Collision Experiments.

The success of the apparatus in producing a bright rapidly accumulating source of cold atoms means that new experiments now become possible.

The first experiments to be attempted will scatter low energy electrons from the cold atoms either produced from the Zeeman slower, or inside the MOT. To carry out these experiments we have invented a new type of atom trap, the AC-MOT. For details, see the AC-MOT web page.

For these experiments, a pulsed electron gun has been constructed and installed in the AC-MOT chamber to produce a low energy electron beam of up to 10 microamps current at energies from 2eV to 100eV.

Electrons from this source are directed at the trapped or slow atoms, so that they are ionized by the electron collision as shown in figure 11.

FIG 11. Collisional ionization of atoms from the Zeeman slower or within the MOT trap, showing the setup of the experiment which detects the recoiling ions from the interaction.

In this case, momentum of incident electron beam is known, as is momentum of cooled or trapped atoms.

FIG 12. The collision for ionization of cold atoms from Zeeman slower and those trapped in the MOT.

By measuring the recoiling ions, it is then possible to measure differential cross sections for ionization of these cold targets over the complete scattering geometry as shown in Figure 12.

The graph shows the momentum which is transferred from the electron to the ionized potassium target as a function of the momentum of the neutral target, where both outgoing electrons are scattered symmetrically wrt the incident beam direction. The incident energy is 30eV, and the curves show the effect of the ionization process on the final ion momenta (velocity and angle) as a function of the scattering angle.

By detecting one electron in coincidence with the ion momentum, it is possible to completely determine the cross section for all scattering geometries.

This technique is also used in COLTRIMS experiments which exploit supersonically prepared target atoms, however by using cold atoms the sensitivity of the technique is enhanced many times.

Further, since the momentum of the initial target can be controlled by the Zeeman slower, it is possible to take sensitive measurements at different angles so as to ensure accuracy in the data.

This is not possible in the COLTRIMS (reaction microscope) technique.

Other experiments being considered include super-elastic scattering of electrons from trapped atoms, study of interaction of electrons with BEC's, and interaction of fermions with ultra-cold ensembles.

Watch this space for new results and developments in these collision experiments, including details of the AC-MOT!

7.0 Conclusions & Future work.

The Cold Atom Facility developed in Manchester has potential to deliver a high density of cold atoms to different experiments so as to allow multiple experiments to be performed from the one source. These experiments may be far ranging, depending on the apparatus that is connected to the source.

Examples of experiments that might be performed include:

        Experiments involving optical lattices

        Studies of BEC processes

        Studies of ultra-cold Fermionic ensembles (eg potassium 40)

        Studies of cold and ultra-cold atoms including measurements of fundamental constants

        Studies of ultra-cold molecules produced by photo-association

        Studies of ultra-cold atoms produced by sympathetic cooling

        Combined laser and electron studies of cold targets

and many other experiments where cold atoms are required at high density.