SciPost Thesis Link
|Title:||Quantum engineering of a low-entropy sample of RbCs molecules in an optical lattice|
|As Contributor:||(not claimed)|
|Degree granting institution:||University of Innsbruck|
Ultracold molecules with electric dipole moments are currently of large interest to the community of experimental and theoretical quantum physicists. Due to the long lifetime of stable ground-state molecules in combination with the strong long-range interaction of the electronic dipole moments, dipolar molecules are ideal candidates to study exotic quantum phases in optical lattices, such as the many-body lattice super solid. Further potential applications of dipolar molecules are the simulation of spin systems, or to use them as a platform for quantum computation. Most of these proposals require a low-entropy sample of dipolar molecules in a lattice. However, it is a great experimental challenge to prepare such ultracold molecular samples in an optical lattice. In the past the realization of low-entropy samples of dipolar molecules in lattices was not feasible. The molecular samples were typically created in the absence of a lattice and in some cases the samples were loaded into an optical lattice subsequently. The resulting phase-space densities of the molecular samples were too low to form low-entropy samples. In this thesis a novel method is introduced that is based on the formation of weakly bound heteronuclear molecules directly in an optical lattice. This way, a significantly lower entropy can be achieved, which presents an ideal starting point to study dipolar molecules in a lattice. This thesis is structured as follows: Initially, the preparation of 87Rb133Cs ground-state molecules, the dipolar molecules of our choice, is introduced. We start from ultracold samples of Rb and Cs atoms. The atoms are associated to weakly bound RbCs molecules by means of a magnetic Feshbach resonance. Subsequently the weakly bound molecules are transferred to their rovibrational and hyperfine ground-state via stimulated Raman adiabatic passage (STIRAP). Next, we demonstrate our method to form the weakly bound RbCs molecules directly in a three-dimensional (3D) optical lattice. We start from spatially separated Rb and Cs Bose-Einstein condensates (BECs). The BECs are loaded into a 3D lattice. The lattice is designed in a way that the Cs sample is frozen out in a single-shell Mott insulator, so that the single Cs atoms are isolated on individual lattice sites, while the Rb sample is still superfluid. The Rb sample is subsequently transported onto the Cs Mott insulator. Finally the depth of the lattice is further increased to also freeze out the Rb atoms in a Mott insulator state. In this way, Rb-Cs pairs can be prepared on individual lattice sites. These pairs are subsequently converted into RbCs Feshbach molecules. We create samples of more than 5000 RbCs molecules with a lattice filling of more than 30%, corresponding to an entropy per molecule of about 2 k_B. Finally, we characterize two important aspects that limit our molecule production efficiency: The critical transport velocity at which the superfluid transport breaks down and the stability of the Cs Mott insulator. We compare the critical transport velocity that we measure with results from other groups and map out the parameter regime in which superfluid transport can be realized. Furthermore, we discuss whether confinement induced resonances might be the reason for a decay of the single-shell Cs Mott insulator that we observe at large repulsive Cs interaction.