We demonstrate quantum control over both internal and external quantum degrees of freedom in a high number of identical "chemical reactions," carried out in an array of microtraps in a 3D optical lattice. Starting from a Mott insulating phase of an ultracold atomic quantum gas, we use two-photon Raman transitions to create molecules on lattice sites occupied by two atoms. In the atom-molecule conversion process, we can control both the internal rovibronic and external center of mass quantum state of the molecules. The lattice isolates the microscopic chemical reactions from each other, thereby allowing photoassociation spectra without collisional broadening even at high densities of up to 2 x 10(15) cm(-3).
We demonstrate quantum control over both internal and external quantum degrees of freedom in a high number of identical "chemical reactions," carried out in an array of microtraps in a 3D optical lattice. Starting from a Mott insulating phase of an ultracold atomic quantum gas, we use two-photon Raman transitions to create molecules on lattice sites occupied by two atoms. In the atom-molecule conversion process, we can control both the internal rovibronic and external center of mass quantum state of the molecules. The lattice isolates the microscopic chemical reactions from each other, thereby allowing photoassociation spectra without collisional broadening even at high densities of up to 2 x 10(15) cm(-3).
The creation of a gas of ultracold polar molecules with a high phase space density brings new possibilities beyond experiments with ultracold atomic gases. In particular, long-range, anisotropic, and tunable dipole-dipole interactions open the way for novel quantum gases, with applications including strongly correlated many-body systems, and ultracold chemistry. This thesis will present the final steps to complete control over both internal and external degrees of freedom of the molecule which allows us to control, and even completely suppress, the chemical reactions between molecules. First, the control over internal states has been achieved through coherent state transfer to the ro-vibronic ground state and coherent manipulations of the hyperfine and rotational states with microwave radiation. Second, external degrees of freedom are controlled by loading the gas into an optical lattice. With the molecules loaded into a one-dimensional lattice, the orientation of the molecular collisions is controlled by manipulating both internal (hyperfine states) and external (motional states in the direction of tight confinement) degrees of freedom. Most striking is that by preparing the molecules all in the lowest band of the lattice in the same internal state, the molecular collisions can only occur in a side-by-side" orientation, where the chemical reaction rate is suppressed by the repulsive dipole-dipole interactions. The chemical reaction can be suppressed completely by further constraining the motion in the trap in a strong 3D lattice. Here we see lifetimes longer than 20 s, limited by o-resonant light scattering. Finally, the ac polarizability of the molecules is explored and controlled. The different rotational states of the molecule have different polarizabilities and will experience a different trapping force in both the optical dipole trap or lattice. We show that there is a magic angle" between the quantization axis and the polarization of the trapping laser at which the polarizabilities of two different rotational states can be matched, eliminating dephasing and allowing for coherent manipulations between rotational states.
We report the successful production of subradiant states of a two-atom system in a three-dimensional optical lattice starting from doubly occupied sites in a Mott insulator phase of a quantum gas of atomic ytterbium. We can selectively produce either a subradiant 1(g) state or a superradiant 0(u) state by choosing the excitation laser frequency. The inherent weak excitation rate for the subradiant 1(g) state is overcome by the increased atomic density due to the tight confinement in a three-dimensional optical lattice. Our experimental measurements of binding energies, linewidth, and Zeeman shift confirm the observation of subradiant levels of the 1(g) state of the Yb(2) molecule.
Recent years have seen tremendous progress in the field of cold and ultracold molecules. A central goal in the field is currently the realization of stable rovibronic ground-state molecular samples in the regime of quantum degeneracy, e.g. in the form of molecular Bose-Einstein condensates, molecular degenerate Fermi gases, or, when an optical lattice is present, molecular Mott-insulator phases. However, molecular samples are not readily cooled to the extremely low temperatures at which quantum degeneracy occurs. In particular, laser cooling, the 'workhorse' for the field of atomic quantum gases, is generally not applicable to molecular samples. Here we take an important step beyond previous work 1 and provide details on the realization of an ultracold quantum gas of ground-state dimer molecules trapped in an optical lattice as recently reported in Ref. 2. We demonstrate full control over all internal and external quantum degrees of freedom for the ground-state molecules by deterministically preparing the molecules in a single quantum state, i.e. in a specific hyperfine sublevel of the rovibronic ground state, while the molecules are trapped in the motional ground state of the individual lattice wells. We circumvent the problem of cooling by associating weakly-bound molecules out of a zero-temperature atomic Mott-insulator state and by transferring these to the absolute ground state in a four-photon STIRAP process. Our preparation procedure directly leads to a long-lived, lattice-trapped molecular many-body state, which we expect to form the platform for many of the envisioned future experiments with molecular quantum gases, e.g. on precision molecular spectroscopy, quantum information science, and dipolar quantum systems.
This thesis describes the results of several experiments that studied \nthe photoassociation of an ultracold atomic Rb gas. In the first experiment, \nwe produced ultracold diatomic molecules from an atomic gas via single-color \nphotoassociation. The molecules were detected with resonance-enhanced multiphoton \nionization. Trapping of these molecules in a quadrupole magnetic \ntrap, with lifetimes up to 20 seconds, was also demonstrated. In addition, the \nrate constant for inelastic collisions between the trapped molecules and atoms \nwas determined from measurements of the atomic density dependence of the \ndecay rate of the trapped molecules. \nIn another experiment, stimulated Raman photoassociation of Rb atoms \nin a Mott insulator state was studied. A Bose-Einstein condensate (BEC) of \n87Rb atoms was loaded into a three-dimensional optical lattice formed by the \ninterference pattern of three orthogonal standing wave laser fields. This system \nconstitutes a very good realization of the Bose-Hubbard model, which \npredicts a quantum phase transition between a superfluid state and a Mott \ninsulator state at a particular lattice height. A time-of-flight imaging method \nwas used to study the state of the atomic gas, and the quantum phase transition \nwas observed at the predicted lattice height. The signature of the phase \ntransition was the disappearance and reappearance of peaks in the image that \narose from the interference of atoms originating from different lattice sites. \nTwo coherent laser fields were applied to the gas in its Mott insulating state, \nand tuned close to a Raman photoassociation resonance, and this resulted in \nan observable loss of atoms due to the formation of molecules. This transition \nexhibited a double-peaked spectrum, with one of the peaks arising from photoassociation \nof atoms in sites containing only two atoms, and the other from \nsites containing three atoms. Also, the loss of atoms vs. the duration of the \nRaman photoassociation period was studied, with the lasers tuned to the peak \ncorresponding to two atoms per site. It was found that a central core of the \ngas, containing about 40 percent of the atoms, exhibited a coherent oscillation \nbetween an atomic and molecular quantum gas.
We consider the transfer of a two-species Bose-Einstein condensate (BEC) into an optical lattice with a density such that a Mott-insulator state with one atom per species per lattice site is obtained in a deep lattice regime. Depending on collisional and transition parameters, the result could be either a ``mixed'' or a ``separated'' Mott-insulator phase. Such a mixed two-species insulator would be well suited for the formation of dipolar molecules via photoassociation. The resulting array of dipolar molecules could then be used for the formation of a dipolar molecular condensate or for computation. For the case of a ${}^{87}\mathrm{Rb}{\ensuremath{-}}^{41}\mathrm{K}$ two-species BEC, however, the large interspecies scattering length makes it difficult to obtain the desired mixed Mott-insulator phase. To overcome this difficulty, we investigate the effect of varying the lattice frequency on the mean-field interaction and propose a favorable parameter regime, under which a lattice of dipolar molecules could be generated.
Ultracold polar molecules, because of their long-range, spatially anisotropic interactions, are a new quantum system in which to study novel many-body phenomena. In our lab, we have produced the first quantum gas of 40K87Rb polar molecules. These molecules were found to undergo exothermic chemical reactions, and this led to interesting studies of chemistry near absolute zero. By creating the molecules at individual sites of a 3D optical lattice, we completely suppress these chemical reactions, and the polar molecule gas becomes stable and lives for tens of seconds. This thesis documents our efforts to explore coherent, many-body phenomena resulting from long-range dipolar interactions in the lattice. By encoding a spin-1=2 system in the rotational states of the molecules, we were able to realize spin-exchange interactions based on a spin Hamiltonian, which is one of the first steps in studying quantum magnetism with polar molecules. While this study was the first realization of such coherent dipolar interactions with polar molecules in a lattice, its full potential was limited by the low lattice filling fractions. Using our ability to exquisitely control the initial atomic gas mixture, we loaded a Mott insulator of Rb and a band insulator of K into the lattice. This quantum synthesis approach led to significantly higher molecular filling fractions and represents the first fully connected system of polar molecules in an optical lattice. This low-entropy quantum gas of polar molecules opens the door to interesting quantum simulations, which should be attainable in the next generation of the experiment.