Atom interferometry in a twin lattice

Abstract

Atom interferometers represent a well-proven tool for precision measurements and are used for a variety of applications ranging from geodesy and inertial navigation to fundamental questions in physics. State-of-the-art sensors typically operate with laser-cooled atoms, which exhibit a relatively large spatial and momentum width. The systematic uncertainties associated with the latter can be overcome by employing Bose-Einstein condensates (BECs). The sensitivity of an atom interferometer can be enhanced by the interrogation time as well as by a large momentum separation of the atomic wave packets. Hereby, a small atomic velocity width is crucial to achieve high-fidelity manipulation.

Within this thesis, the realization of a novel beam splitter for the transfer of large momentum on an ultra-cold atom cloud is presented. The condensate is generated in a miniaturized atom-chip based setup and its expansion rate is further reduced by delta-kick collimation. The beam splitting light field consists of two counterpropagating lattices with orthogonal polarizations. In such a twin lattice, the efficient combination of double Bragg diffraction with Bloch oscillations promises to overcome current limitations of large momentum transfer. Bloch transfer efficiencies of more than 99.9% per photon recoil (ħk) can be achieved and, thus, excellent scalability is provided. The symmetry of the twin lattice enables the suppression of systematic errors. These features allow for the realization of a symmetric Mach-Zehnder-type geometry, where contrast can be observed up to a maximum splitting of 408ħk corresponding to a total transfer of 1632ħk. To our knowledge, this represents the largest momentum separation in an atom interferometer reported so far. A detailed experimental and theoretical study reveals that the current limitations are solely caused by technical properties of the experiment. In particular, light field distortions arising due to the diffraction of the laser beam at different apertures cause a dephasing leading to a contrast decay. The results open up new routes for the miniaturization of inertial quantum sensors as well as for gravitational wave detectors. In addition, the combination of an atom-chip trap with an optical dipole trap is investigated, which serves as a pathfinder experiment for atom interferometry in optical waveguides. Waveguides allow extending the interrogation time without an increase of the interferometer region. Optimizing the spatial overlap of both traps as well as the temporal sequence, the BEC can be transferred into the dipole trap with an efficiency of over 99%. Such a setup also has applications in future atom-chip experiments featuring two atomic species, whose interactions can only be controlled in a purely optical trap.