The work of this dissertation falls into two broad categories. In the first part, I describe loading a single atom from a reservoir into a blue-detuned crossed vortex bottle beam trap using a dynamic 1D optical lattice. The lattice beams are frequency chirped using acousto-optic modulators, which causes the lattice to move along its axial direction and behave like an optical conveyor belt. A stationary lattice is initially loaded with approximately 6000 atoms from a reservoir, and the conveyor belt transports them 1.1 mm from the reservoir to a bottle beam trap, where a single atom is loaded via light-assisted collisions. Photon counting data confirm that an atom can be delivered and loaded into the bottle beam trap 13.1% of the time. In part II, I describe a theory and experiment in the field of atomtronics that displays a coherent gain mechanism for a triple-well matterwave transistor oscillator. I start with a well-established semi-classical description of an atomtronic transistor but model the system using a many-body formalism. The quantum model predicts interesting physics when the atoms flowing through the transistor have sufficiently low enough temperatures such that the motional state of a dipole oscillating BEC, placed in the transistor itself, couples atom transitions between high lying transistor energy eigenstates. In this regime, the coupling gives rise to a new gain mechanism that increases the flux of matterwaves flowing out of the transistor system, compared to when the coupling is absent. Our experiments suggest that the gain mechanism is coherent and increases the spread of matterwave momenta that flows out of the transistor.
The work of this dissertation falls into two broad categories. In the first part, I describe loading a single atom from a reservoir into a blue-detuned crossed vortex bottle beam trap using a dynamic 1D optical lattice. The lattice beams are frequency chirped using acousto-optic modulators, which causes the lattice to move along its axial direction and behave like an optical conveyor belt. A stationary lattice is initially loaded with approximately 6000 atoms from a reservoir, and the conveyor belt transports them 1.1 mm from the reservoir to a bottle beam trap, where a single atom is loaded via light-assisted collisions. Photon counting data confirm that an atom can be delivered and loaded into the bottle beam trap 13.1% of the time. In part II, I describe a theory and experiment in the field of atomtronics that displays a coherent gain mechanism for a triple-well matterwave transistor oscillator. I start with a well-established semi-classical description of an atomtronic transistor but model the system using a many-body formalism. The quantum model predicts interesting physics when the atoms flowing through the transistor have sufficiently low enough temperatures such that the motional state of a dipole oscillating BEC, placed in the transistor itself, couples atom transitions between high lying transistor energy eigenstates. In this regime, the coupling gives rise to a new gain mechanism that increases the flux of matterwaves flowing out of the transistor system, compared to when the coupling is absent. Our experiments suggest that the gain mechanism is coherent and increases the spread of matterwave momenta that flows out of the transistor.
This thesis presents novel techniques to realize controllable quantum systems of neutral atoms. Besides the preparation of the exact number of atoms, we manipulate all physical degrees of freedom of the trapped particles. The first part (Chapters 2, 3) reports on a deterministic source of single atoms, which overcomes the limitations imposed by statistical arrival in conventional sources. Single cold cesium atoms prepared in a magneto-optical trap are transferred into a standing wave optical dipole trap, made of two counter-propagating red-detuned Nd:YAG laser beams. Mutual detuning of the laser beam frequencies moves the standing wave pattern, allowing us to accelerate and stop an atom at preselected points along the standing wave. This "optical conveyor belt" can transport one atom over a maximum distance of 10~mm. The second part (Chapters 4, 5) reports on the investigation of the coherence times of the Cs hyperfine ground states using microwave transitions. Using Ramsey spectroscopy techniques we measured coherence times of up to 100 ms. The limiting dephasing effects are experimentally identified and are of technical rather than fundamental nature. We present an analytical model of the reversible and irreversible dephasing mechanisms. Finally, we prove that controlled transport by the optical conveyor belt over macroscopic distances preserves the atomic coherence with slight reduction of coherence time.
A self-consistent theoretical treatment of a triple-well atomtronic transistor circuit reveals the mechanism of gain, conditions of oscillation, and properties of the subsequent coherent matter waves emitted by the circuit. A Bose-condensed reservoir of atoms in a large source well provides a chemical potential that drives circuit dynamics. The theory is based on the ansatz that a condensate arises in the transistor gate well as a displaced ground state, that is, one that undergoes dipole oscillation in the well. That gate atoms remain condensed and oscillating is shown to be a consequence of the cooling induced by the emission of a matter wave into the vacuum. Key circuit parameters such as the transistor transconductance and output current are derived by transitioning to a classical equivalent circuit model. Voltage-like and current-like matter-wave circuit wave fields are introduced in analogy with microwave circuits, as well as an impedance relationship between the two. This leads to a notion of a classically coherent matter wave that is the dual of a coherent electromagnetic wave and which is distinct from a de Broglie matter wave associated with cold atoms. Subjecting the emitted atom flux to an atomic potential that will reduce the de Broglie wavelength, for example, will increase the classical matter-wave wavelength. Quantization of the classical matter-wave fields leads to the dual of the photon that is identified not as an atom but as something else, which is here dubbed a ``matteron.''
This dissertation presents a theoretical analysis of methods to manipulate and control the momentum state of coherent matter waves. Of particular interest is the coherent acceleration of a quantum-degenerate atomic system, which, as a direct consequence of the form of the de Broglie wavelength, results in tunable source of matter waves. Such sources are of considerable importance for a number of potential applications in the field of atom optics, including the development of highly sensitive gyroscopes, accelerometers, gravity gradiometers or atom lithography and holography, as well as for potential uses in integrated atom optics. Our basic setup consists of a Bose-Einstein condensate in a moving optical lattice created by a pair of frequency-chirped counterpropagating laser beams acting as a "conveyor belt'' for ultracold atoms. Whereas the acceleration of ultracold but non-condensed atoms in such a lattice was demonstrated earlier, we extend this scheme to the case of Bose-Einstein condensates. As a first step, we investigate the acceleration efficiency for various acceleration rates and nonlinear interaction strengths. We find parameter regimes where efficient acceleration is possible, i.e. all atoms are accelerated to the same velocity and the initially sharp momentum distribution and thus its monochromaticity is preserved. However, in general we identify switch-on effects of the lattice, dynamical loss and nonlinear effects to be responsible for deterioration of the monochromaticity of the condensate: On the one hand, switch-on effects and dynamical loss induce a coupling of the initially populated momentum mode to other modes, thereby distributing the momentum over several modes. On the other hand, the nonlinear release of mean field energy during the acceleration process causes the mode profile itself to broaden, also leading to a contamination of the initial monochromaticity. As a second step, we discuss ways to improve this scheme by removing the restriction of constant accelerations. We employ genetic algorithms to optimize the time-dependent motion of the lattice. We show that with such flexibility, it is possible to achieve a fast and highly efficient coherent acceleration of condensates, even when mean-field effects cannot be neglected. The same scheme also enables the creation of arbitrary coherent superposition states in momentum space. The technique is thus suitable for building highly efficient momentum state beam splitters. In addition to simply accelerating condensates, it is desirable for many potential applications to transport atomic wave packets without dispersion over large distances. This can be achieved by launching bright atomic solitons, where the effects of the nonlinearity counterbalance the dispersion. Placing a Bose-Einstein condensate with repulsive interactions in a lattice, one can create a negative effective mass. Under these circumstances bright and stable soliton solutions exist, so-called gap solitons. After a careful analysis of the soliton properties, we use the tools we developed for condensate acceleration and demonstrate two feasible schemes to excite the solitons. Experimental data released after publication of our results demonstrating the acceleration of Bose-Einstein condensates in moving lattices and the very recent observation of atomic gap solitons indicates that our theoretical analysis was timely and indeed experimentally feasible. As an outlook, we briefly comment on a new direction in the field of atom optics that holds promise for future applications: the use of quantum degenerate Fermi gases. In atomic as well as in optical physics one often encounters situations where there exists a coupling between several modes of a system. Here, we illustrate the "toy model'' of a fermionic coupler where transitions between two internal states are induced by Raman coupling. Due to Fermi statistics and interatomic interactions, this is a simple example of a nonlinear multimode coupler. Investigation of this system consisting of only a few fermions already clearly illustrates the basic differences between bosonic and fermionic dynamics and sheds light on the role of two-body collisions. Understanding the basic mechanisms of this system is a first step towards more sophisticated coherent control of fermionic systems.
Research in the field of atomtronics aims to develop a new paradigm for the use of ultracold atomic systems in a manner that mimics the functionality of electronic circuits and devices. Given the ubiquity of the electronic transistor and its application to a vast array of signal processing tasks, the development of its atomtronic counterpart is of significant interest. This dissertation presents the experimental studies of two atomtronic circuit elements: a battery and transistor. Experiments are conducted in an atom-chip-based apparatus utilizing hybrid magnetic and optical trapping techniques that enable one to "pattern" atomtronic circuit elements. An atomtronic battery is realized in a double-well trapping potential in which a finite-temperature Bose-Einstein condensate is prepared in a non-equilibrium state to generate thermodynamic gradients that drive atom current flow. Powered by the atomtronic battery, a triple-well atomtronic transistor is demonstrated, and quasi-steady-state behavior of the device is characterized. Results are found to be in agreement with a semiclassical model of the transistor that is also used to study the active properties of the device, including current gain. Based on these results, future directions regarding signal processing operations are proposed.
An atomtronic transistor circuit is used to realize a driven matterwave oscillator. The transistor consists of Source and Drain regions separated by a narrow Gate well. Quasi-steady-state behavior is determined from a thermodynamic model, which reveals two oscillation threshold regimes. One is due to the onset of Bose-Einstein condensation in the Gate well, the other is due to the appearance of a negative transresistance regime of the transistor. The thresholds of oscillation are shown to be primarily dependent on the potential energy height difference between Gate-Drain and Gate-Source barriers. The transistor potential is established with a combination of magnetic and optical fields using a compound glass and silicon substrate atom chip. The onset of oscillation and the output matterwave are observed through in-trap imaging. Time-of-flight absorption imaging is used to determine the time dependence of the Source well thermal and chemical energies as well as to estimate the value of the closed-loop ohmic Gate resistance, which is negative and is observed to cause cooling of Source atoms.
This thesis describes and characterizes a setup to conduct experiments combining the fields of cold atoms and nanophotonics. Furthermore, a series of numerical simulations is employed to deepen of atom transport under experimentally realistic conditions. The study of cold atoms is usually fundamental in nature and a very clean approach to investigate the laws of physics. High purity samples of atoms are cooled down to millikelvins or microkelvins. The potential landscape for the atoms can be carefully controlled using lasers and/or magnetic fields. Since the atoms' responses to electric, magnetic or electromagnetic fields are known with very high fidelity, any deviations to the state the atoms are prepared in can be precisely measured. Nanophotonics is the study of light close to or beyond the diffraction limit. In this region the field-like nature of light starts to play a role. By confining light in sub-wavelength-sized structures, high peak intensities can be achieved, giving rise to exciting non-linear optics phenomena. The strong field gradient that can be created allows, for instance, precise spectroscopy measurements, phase-sensitive sub-wavelength microscopy, as well as producing highly receptive sensing devices. Combining cold atoms and nanophotonics creates strong synergy effects. It gives access to strong coupling of light and matter. The scalability of nanophotonics now extends to atomic physics. Atoms can be trapped, coupled and manipulated at very small length scales, and in large numbers. Moreover, the potential landscape for atoms is highly customizable. The rubidium isotope 87 is used in this setup. A cloud of rubidium atoms is gathered and cooled in a two dimensional magneto-optical trap (MOT). Since the atoms are only cooled in two dimensions, they are free to escape in the third dimension and load a following three dimensional MOT. The atoms are now cooled further, compressed and loaded into an optical dipole trap. This dipole trap transports the atoms close to a sample inside the vacuum chamber. A moving optical lattice, or optical conveyor belt, brings the atoms then close to the sample surface. Now experiments can be conducted, studying the interaction of cold atoms and strongly confined light. The atom transport down to the sample surface shows some unexpected behaviour: the atoms seem to spread through the optical lattice during transport. A combination of microscopic and macroscopic numerical simulations, including genetic algorithms, are used to deepen our understanding of the underlying physics. This also allows us to get access to experimental variables that cannot be directly obtained.