From studies of fundamental physics to quantum technologies the production of ultracold molecules will have a huge impact across a range of applications. For many years laser cooling, which became an invaluable tool in cold atomic physics, was deemed to be too impractical for application to molecules. Nevertheless, laser cooling has now been demonstrated for a few molecular species. Using a frequency-chirped laser slowing technique, the velocity distribution of a pulse of CaF molecules is compressed and slowed from 180 m/s to about 10 m/s. These slow molecules are then captured in a magneto-optical trap. I present measurements that show how the number of molecules, the photon scattering rate, the oscillation frequency, damping constant, temperature, cloud size and lifetime depend on the key parameters of the magneto-optical trap, especially the intensity and detuning of the main cooling laser. The trap contains up to 2*10^4 molecules, the maximum photon scattering rate is 2.5*10^6 s-1 per molecule, the maximum oscillation frequency is 100 Hz, the maximum damping constant is 500 s^-1, and the minimum rms radius of the trapped cloud is 1.5 mm. A minimum temperature of 730 microkelvin is obtained by ramping down the laser intensity to lower values. To reach lower temperatures, the cloud is loaded into a blue-detuned optical molasses, which cools the molecules to 55 microkelvin, well below the Doppler-limiting temperature. I characterise the cooling process and suggest the sub-Doppler mechanisms responsible. These ultracold molecules are the optically pumped into a single quantum state, and coherently transferred between selected hyperfine components of the ground and first-excited rotational states. Finally, the ultracold, state-selected molecules are loaded into a magnetic trap that has a lifetime of about 1 s.
From studies of fundamental physics to quantum technologies the production of ultracold molecules will have a huge impact across a range of applications. For many years laser cooling, which became an invaluable tool in cold atomic physics, was deemed to be too impractical for application to molecules. Nevertheless, laser cooling has now been demonstrated for a few molecular species. Using a frequency-chirped laser slowing technique, the velocity distribution of a pulse of CaF molecules is compressed and slowed from 180 m/s to about 10 m/s. These slow molecules are then captured in a magneto-optical trap. I present measurements that show how the number of molecules, the photon scattering rate, the oscillation frequency, damping constant, temperature, cloud size and lifetime depend on the key parameters of the magneto-optical trap, especially the intensity and detuning of the main cooling laser. The trap contains up to 210^4 molecules, the maximum photon scattering rate is 2.510^6 s-1 per molecule, the maximum oscillation frequency is 100 Hz, the maximum damping constant is 500 s^-1, and the minimum rms radius of the trapped cloud is 1.5 mm. A minimum temperature of 730 microkelvin is obtained by ramping down the laser intensity to lower values. To reach lower temperatures, the cloud is loaded into a blue-detuned optical molasses, which cools the molecules to 55 microkelvin, well below the Doppler-limiting temperature. I characterise the cooling process and suggest the sub-Doppler mechanisms responsible. These ultracold molecules are the optically pumped into a single quantum state, and coherently transferred between selected hyperfine components of the ground and first-excited rotational states. Finally, the ultracold, state-selected molecules are loaded into a magnetic trap that has a lifetime of about 1 s.
Ultracold molecules can be used for a diverse set of exciting applications including controlled quantum chemistry, probes of physics beyond the standard model, simulation of many-body quantum systems and quantum information processing. However their rich internal structure makes the required control of their motional and internal states difficult. In the last decade, the field has made rapid progress on developing techniques for producing, cooling and trapping molecules and, whilst challenges remain, many of these applications are now being realised. This thesis presents my work on laser cooling and coherent control of calcium monofluoride (CaF) molecules. We load molecules into a three-dimensional magneto-optical trap (MOT), the first such trap for CaF and the second for any molecule. The MOT contains around 2×10⁴ molecules with an initial temperature of 12 mK. Lowering the laser intensity brings the temperature of the MOT down to 960 μK. To further cool the molecules we load them into a blue-detuned optical molasses, cooling them to 5.4(7) μK. This temperature is ~40 times below the Doppler limit and corresponds to a typical molecule moving at only a few times the single-photon recoil velocity. We optically pump the molecules into a single Zeeman sub-level and coherently transfer them between rotational states using microwave radiation. We load the molecules into a magnetic trap where they remain confined for about 5 s, the lifetime being limited by vibrational excitation due to room-temperature blackbody radiation. We show how to prepare the trapped molecules in a coherent superposition of two rotational states, and how to maintain this coherence for long periods, up to 6.4(8) ms. Detailed simulations of the remaining decoherence mechanisms suggest that it is possible to extend this to more than a second by compressing the sample to a smaller size, and using a magnetic trap geometry where the direction of the field is nearly uniform across the trapping volume. Finally we explore theoretically how Raman sideband cooling can be applied to molecules in optical tweezer traps, potentially providing a route to full quantum control.
Laser cooling and trapping in a magneto-optical trap (MOT) have been essential to the success of cold atom physics in the last decades. Recently, the application of the same techniques to molecules has begun. The complexity of even a simple diatomic molecule makes laser cooling difficult, but promises new applications in many areas of research. In this thesis I describe the development of the first three-dimensional MOT of calcium fluoride (CaF) molecules. First, a cryogenic buffer gas source was set up, producing a pulsed beam of 9.3*10^10 molecules per steradian per pulse with forward velocities around 170 m/s. A similar source for very large molecules was set up during a 5 month internship at the University of Vienna. Next, the molecular pulse was slowed down to the capture velocity of a MOT using chirped laser slowing, resulting in about 7*10^5 CaF molecules passing through the typical MOT volume of 1 cm^3 at velocities of 15+-5 m/s. A new deceleration method, called Zeeman-Sisyphus deceleration, was also investigated. In this method molecules move through a spatially varying magnetic field and are optically pumped between low- and high-field seeking states in such as a way that they are always losing kinetic energy. The method promises to deliver more slow molecules because the molecules are guided transversely as they are decelerated. A small prototype was built and the optical pumping step was tested successfully. Finally, 7.6*10^3 CaF molecules were trapped in a MOT and cooled to a temperature of 8.5 mK. The radial trap frequency is 2 pi*130 Hz and the damping constant is beta=9.5*10^2 s^-1. The lifetime is about 100 ms and depends strongly on the scattering rate. This MOT is an an ideal starting point for a wide range of new experiments with ultracold molecules.
Cold and ultracold molecules are highly desirable for a diverse range of applications in physics and chemistry such as precision measurements, tests of fundamental physics, quantum simulation and information processing, quantum chemistry, and the physics of strongly correlated quantum matter. Laser cooling is usually infeasible in molecules because their rotational and vibrational transitions make is difficult to come up with a closed scattering cycle. Recently, a narrow range of diatomic molecules, one of which is CaF, has been shown to possess a convenient electronic structure and a highly-diagonal Franck-Condon matrix and thus be amenable to laser cooling. This thesis describes experiments on laser cooling of CaF radicals produced in a supersonic source. We first investigate the increased fluorescence when multi-frequency resonant light excites the molecules from the four hyperfine levels of the ground X²∑+(N = 1, v = 0) state to the first excited A²π½ (J’ = 1=2; v’ = 0) state. The number of photons scattered per molecule increases significantly from one or two in the single frequency case to more than 50 before the molecules get pumped into the X²∑+(N = 1; v = 1) state. We demonstrate laser cooling and slowing of CaF using counter-propagating laser light which causes the molecules to scatter more than a thousand photons on the X (N = 1, v = 0, 1) <->A (J’ = 1=2; v’ = 0) transition. The effect of the laser cooling is to slow a group of molecules from 600 ms-1 to about 580 ms-1 and to narrow their velocity distribution from an initial temperature of 3 K down to 300 mK. In addition, chirping the frequency of the cooling light to keep it on resonance with the decelerating molecules doubles the deceleration and further compresses the velocity distribution. The effect of the laser cooling is limited by the optical pumping of molecules in the X (N = 1, v = 2) state.
We present the properties of a magneto-optical trap (MOT) of CaF molecules. We study the process of loading the MOT from a decelerated buffer-gas-cooled beam, and how best to slow this molecular beam in order to capture the most molecules. We determine how the number of molecules, the photon scattering rate, the oscillation frequency, damping constant, temperature, cloud size and lifetime depend on the key parameters of the MOT, especially the intensity and detuning of the main cooling laser. We compare our results to analytical and numerical models, to the properties of standard atomic MOTs, and to MOTs of SrF molecules. We load up to molecules, and measure a maximum scattering rate of s −1 per molecule, a maximum oscillation frequency of 100 Hz, a maximum damping constant of 500 s −1 , and a minimum MOT rms radius of 1.5 mm. A minimum temperature of 730 μ K is obtained by ramping down the laser intensity to low values. The lifetime, typically about 100 ms, is consistent with a leak out of the cooling cycle with a branching ratio of about . The MOT has a capture velocity of about 11 m s −1 .
Diatomic molecules are considered good candidates for the study of strongly correlated systems and precision measurement searches due to their combination of complex internal states and strong long-range interactions. Cooling molecules down to ultracold temperatures is often a necessary step for fully utilizing the power of the molecule. This requires a trap for molecules and the ability to cool molecules to the mK regime and below. A magneto-optical trap (MOT) is a good tool for achieving mK temperatures. However, extra care is needed for molecules to form the necessary quasi-closed cycling transitions due to molecule's complicated energy structure. In our work with CaF, we use two repump lasers to block vibrational leakage and selection rules for the rotational degree of freedom to achieve about 10^{5} photon cycles.\n\n\n\nThe two-stage buffer gas beam source is a general method to generate a cold and slow beam of molecules with a forward velocity of about 50 m/s. The compatibility of the buffer-gas source with a MOT is studied and we confirm that such beams can be nicely compatible with MOTs using various atomic species. \n\n\n\nIn order to load molecules into a MOT from even such a slow beam, additional slowing is required due to the low capture velocity of a molecular MOT (< 10 m/s). We apply a frequency-broadened “white-light” slowing on CaF from a two-stage source, demonstrating slowing of CaF below 10 m/s. An AC MOT, which provides active remixing of dark substates, is also developed and Li atoms are slowed and trapped. These are crucial ingredients for co-trapping CaF molecules and Li atoms and study their collisional properties, which would lead to sympathetic cooling of molecules down to ultracold temperatures. \n\n\n\nThe achievement of slowing and development of this system allowed for the detailed study of the CaF laser cooling system, as well as physical processes involved with AC MOTs and the proposed MOT for CaF. Crucial knowledge of this archetypal system provides significant progress toward manipulation and control of molecules similar to what has been achieved with atoms and what is necessary for searches for new physics with ultracold molecules.
Potential wide-ranging scientific applications, spanning fundamental physics to quantum engineering, has led to significant efforts in controlling molecules at the quantum level. The rich internal structure of molecules, that gives rise to these desirable properties also complicates the task of controlling such species. Over the past decade, our ability to produce, control, and detect molecules has advanced tremendously with direct laser-cooling and magneto-optical trapping of diatomic molecules realized by several groups around the world. In this thesis, we describe the creation of the first RF MOT of CaF, which at the time of the writing of this thesis, remains the largest and densest molecular MOT. New laser cooling techniques for molecules are demonstrated, allowing laser cooling of molecules to 4 uK, 50 times colder than the Doppler limit. Optical trapping of these molecules is also achieved, which, in combination with in-trap laser cooling increased the density of trapped molecules by five orders of magnitude compared to the density in the MOT. We developed methods to cool and detect single molecules with high fidelity, which aided us in the creation of an optical tweezer array of single, ultracold CaF molecules. The densities of molecules (10^12 cm^-3) attained inside the tweezer traps also enabled us to observe both light-assisted and ground state collisions of laser cooled molecules, for the first time. The phase space density within the merged tweezers reached 5x10^-4, 10 orders of magnitude larger than the initial MOT. We implemented internal quantum state control of the molecules and dynamical control of the tweezers to build a platform for exploring state-selective ultracold quantum chemistry.