This dissertation reports on the design, construction, and investigation of a pulsed optical dipole force trap which uses laser light to confine ultracold rubidium (Rb) atoms. Because the laser frequency is detuned far from the atomic resonance frequency, the optical dipole force trap is also called a "far-off-resonance trap" (FORT). The use of pulsed laser light to create an optical trap may find application in expanding the number of atomic species which can be confined. The experiments reported here are principally aimed, however, at understanding the physics of pulsed FORT dynamics in anticipation of using the free electron laser (FEL) at Jefferson Lab (Jlab) to construct a FORT. The Jlab FEL will provide a tunable, high power laser light source enabling the realization of spatially large and/or energetically deep traps which are not presently accessible with table-top laser sources. Here, a mode-locked Nd:YAG laser is used as a pulsed laser source. Since the conservative optical dipole force trap does not cool Rb atoms, ultracold atoms must be loaded into the FORT. The optical dipole force trap is formed at the focus of a Gaussian, 1.06 μm Nd:YAG laser beam which is located at the center of a high vacuum chamber and superimposed onto the center of a pre-cooled Rb atom cloud that has been previously accumulated in a magneto-optic trap (MOT). The performance of the pulsed FORT is compared to a continuous wave mode (cw) FORT which is built using the same laser beam; the operation of the two kinds of FORTs can be switched easily without disturbing the experimental alignment. The dependencies of FORT loading efficiency to FORT and MOT parameters such as FORT laser power, loading time, storage/holding time, detuning of the primary MOT laser frequency, and repump laser intensity are investigated. There are about 1.5×107 ultracold rubidium atoms in the MOT. At 7 Watts FORT laser power, about 8% of the atoms are loaded successfully into the cw FORT and about 5% into the pulsed FORT under similar, but not identical, conditions. In most respects both the cw and pulsed FORTs show comparable behavior. The behavior of both FORTs depends strongly on FORT laser power. As expected, at higher power, more atoms can be loaded into the FORT. Both FORTs also depend on the loading time and holding time.
The aim of this research is to pave the way for making a trap for cold neutral atoms based on the force generated by pulses of a mode-locked laser. As an onset towards such a trap we decided to use a far-off-resonance trap (FORT) loaded with cold rubidium atoms from a magneto-optical trap (MOT). Our FORT setup consists of a titanium-sapphire laser that can operate both in continuous wave mode and in mode-locked mode. Our MOT setup uses a magnetic field gradient, and two diode lasers whose frequency is locked very accurately to the atomic transition using polarisation spectroscopy. Since the measurements we do require very precise timing of the order of less than one millisecond between the various phases of the experiment, a time-sequencing program was developed. Using this sophisticated setup, several characteristics of FORTs made with continuous and pulsed lasers were analysed. For example, we have investigated the loading behaviour of the traps and found that under general circumstances there is no significant difference. However, under special conditions it is possible to load more atoms into a pulsed trap. The lifetime of the number of atoms in the trap shows no dependence on the power of the trap laser in the continuous case. For the pulsed trap, however, the increased scattering due to the high peak intensities of the pulses does limit the achievable lifetimes. Moreover, due to the presence of a large number of photoassociation lines, i.e. lines were loosely bound molecules are formed by the interaction of two atoms an a photon, the two-particle losses depend heavily on the wavelength of the trap laser. Furthermore we have done measurements of the temperature of the trapped atoms, which revealed that the temperature is a constant fraction of the trap depth. We also looked at the effects of parametric excitation of the atoms in the FORT. The anharmonicity of the trap formed by the titanium-sapphire laser results in interesting physics, including a change in the average temperature of the atoms, depending on the modulation frequency. We also investigated the trapping behaviour of the FORT at low pulse repetition rates by switching the laser intensity on and off on a kHz scale. We investigated the dependence of the number of atoms on the polarisation of the FORT laser light, both for continuous and pulsed traps. Atoms can be trapped at elliptical polarisations as well, although the number of trapped atoms is less. As the laser is detuned further to the red of the rubidium D line, this effect becomes less and less pronounced, as predicted by our model.
We theoretically study the loading of cold atoms into a crossed optical dipole trap (ODT). We compared the loading efficiency dependent on the trap geometry for Nd:YAG and CO2 laser ODTs. By crossing two Nd:YAG laser beams at a small angle, such as 10°, we can shape the geometry of the effective potential in the intersection region and achieve comparable performance to that of a single-beam CO2 laser ODT. We also calculate the optimal experimental parameters for these two laser wavelengths for laser power per beam ranging from 5 W to 100 W. We further experimentally studied the loading of laser-cooled Rb87 atoms into a small-angle crossed Nd:YAG ODT and found that the experimental data agreed with the simulation results.
The interaction of light with atoms and molecules is of fundamental interest in many branches of science. In atomic physics, this interaction can be used to cool and spatially confine (trap) atoms. These traps can be used as the starting point for other experiments, but the dynamics of the cooling and trapping processes is itself of interest. In order to better understand the physics of trapping atoms in an optical dipole force trap, we have conducted a series of spectroscopic measurements of ultracold rubidium atoms in such a trap. The trap was created at the focus of a Nd:YAG laser beam with wavelength 1064nm and nearly Gaussian spatial mode. For rubidium, the trap light is red-detuned and the atoms are confined to the highest intensity in the beam. In order to probe the atoms confined in the trap, we have performed spectroscopy on the 5S1/2 → 5P3/2 transition. Because the polarizability of the ground and excited states is not the same at the trap wavelength, the spectra exhibit both a shift and inhomogeneous broadening. We have investigated the spectra for both linear and circular polarized traps. We also studied the application of a second laser to couple the excited state (5P3/2) to another higher excited state (5D5/2) to reduce the inhomogeneous broadening of the 5S1/2 → 5P 3/2 transition. Finally, two-photon spectroscopy was performed on atoms in the dipole force trap and compared to spectra taken in a magneto optical trap (MOT). Autler-Townes splitting was clearly observed in the MOT and appears to have been observed in the dipole force trap also.
We demonstrate an efficient scheme for continuous trap loading based upon spatially selective optical pumping. We discuss the case of
1S0 calcium atoms in an optical dipole trap (ODT), however, similar strategies should be applicable to a wide range of atomic species, which do not permit efficient conventional trap loading. Our starting point is a moderately cold $(\ensuremath{\approx}300\phantom{\rule{0.3em}{0ex}}\ensuremath{\mu}\mathrm{K})$ and dense $(\ensuremath{\approx}9\ifmmode\times\else\texttimes\fi{}{10}^{9}\phantom{\rule{0.3em}{0ex}}{\mathrm{cm}}^{\ensuremath{-}3})$ reservoir of metastable
3P2 atoms in a magneto-optic trap (triplet MOT). A focused
532nm laser beam produces a strongly elongated optical potential for
1S0 atoms with up to $350\phantom{\rule{0.3em}{0ex}}\ensuremath{\mu}\mathrm{K}$ well depth. A weak focused laser beam at
430nm, carefully superimposed upon the ODT beam, selectively pumps the
3P2 atoms inside the capture volume to the singlet state, where they are confined by the ODT. The triplet MOT perpetually refills the capture volume with
3P2 atoms, thus providing a continuous stream of cold atoms into the ODT at a rate of ${10}^{7}\phantom{\rule{0.3em}{0ex}}{\mathrm{s}}^{\ensuremath{-}1}$. Limited by evaporation loss, in
200ms we typically load $5\ifmmode\times\else\texttimes\fi{}{10}^{5}$ atoms with $3.5\ifmmode\times\else\texttimes\fi{}{10}^{10}\phantom{\rule{0.3em}{0ex}}{\mathrm{cm}}^{\ensuremath{-}3}$ peak density and axial and radial temperatures of 300 and $85\phantom{\rule{0.3em}{0ex}}\ensuremath{\mu}\mathrm{K}$, respectively. After terminating the loading we observe evaporation during
50ms, typically leaving us with
105 atoms with nearly equal radial and axial temperatures close to $40\phantom{\rule{0.3em}{0ex}}\ensuremath{\mu}\mathrm{K}$ and a peak density of $5\ifmmode\times\else\texttimes\fi{}{10}^{11}\phantom{\rule{0.3em}{0ex}}{\mathrm{cm}}^{\ensuremath{-}3}$. A simple model is presented, which identifies the relevant physical mechanisms of the loading and decay dynamics. We point out that a comparable scheme could be employed to load a dipole trap with
3P0 atoms.