Le couplage de guides d'ondes nanoscopiques et d'atomes froids a récemment ouvert de nouvelles voies de recherche. Le guide d'onde dans notre cas est une nanofibre qui confine la lumière transversalement à une échelle inférieure à la longueur d'onde. La lumière guidée présente un fort champ évanescent permettant une interaction atome-photon exaltée au voisinage de la nanofibre. Dans notre expérience, un nuage atomique froid est d'abord superposé à une nanofibre optique. Puis, en utilisant un piège dipolaire via le champ évanescent de la nanofibre, les atomes froids sont piégés à proximité de sa surface. Avec cette plateforme, nous avons obtenu des épaisseurs optiques élevées OD ~ 100 et de longues durées de vie ~ 25 ms en utilisant un schéma de piégeage qui préserve les propriétés internes des atomes. Une direction intéressante est alors d'explorer les effets collectifs résultant de l'ordre spatial des atomes. Lorsque la période du réseau est proche de la longueur d'onde de résonance, une réflexion de Bragg aussi élevée que 75% est observée. Cette réflexion dépend de la polarisation de la sonde par rapport aux réseaux atomiques - une signature de la chiralité dans les systèmes à guide d'ondes nanoscopiques. La possibilité de contrôler le transport de photons dans les guides d'ondes couplés à des systèmes de spin permettrait de nouvelles fonctionnalités pour les réseaux quantiques et l'étude d'effets collectifs résultant d'interactions à longue distance. / The coupling of cold atoms to 1D nanoscale waveguides have opened new avenues of research. The waveguide in our case is a nanofiber, which confines light transversally to a subwavelength scale. The guided light exhibits a strong evanescent field allowing enhanced atom-photon interaction in the vicinity of nanofiber. In our experiment, a cold atomic cloud is first interfaced with an optical nanofiber. By using an optical lattice in the evanescent field, the atoms are then trapped in 1D atomic arrays close to the nanofiber. In this platform, we reach high optical depth OD ~ 100 and long lifetimes ~ 25 ms by using a dual-color compensated trapping scheme that preserves the internal properties of atoms. In this thesis, we explore collective effects emerging from the spatial ordering of atoms. When the period of the lattice is made close to commensurate with the resonant wavelength, Bragg reflection, as high as 75%, is observed. The reflection shows dependency on orientation of the probe polarization relative to the atomic arrays - a chiral signature in nanoscale waveguide-QED systems. The ability to control photon transport in 1D waveguides coupled to spin systems would enable novel quantum networking capabilities and the study of many-body effects arising from long-range interactions.

Diese Doktorarbeit beschäftigt sich mit theoretischen Aspekten der Wellenleiterelektrodynamik (WQED), also mit der Wechselwirkung von Materie und Licht, welches nur in einer Dimension propagieren kann. Dieses Forschungsfeld erfreut sich seit seiner Entstehung in den 1990er Jahren wachsender Beliebtheit, der Grund hierfür sind die mannigfaltigen Anwendungsmöglichkeiten, beispielsweise bei der Konstruktion von Quantencomputern als auch von klassischen Computern. Auch Vorschläge für sogenannte Pump-Probe-Experimente auf der Basis der WQED sind Gegenstand der aktuellen Forschung.\\ All diese Gebiete sind darauf angewiesen, die zugrunde liegenden Prinzipien zu verstehen, diese Arbeit soll einen Beitrag dazu leisten. Hierzu haben wir einen Formalismus entwickelt, der auf Feynman-Diagrammen fußt. Das erste physikalische Modellsystem, welches hiermit untersucht wurde, besteht aus einem 1D-Wellenleiter und einem daran gekoppelten Zwei-Nievau-Atom (ZNA). Dies erlaubte uns, bekannte Rechnungen physikalisch transparenter und mathematisch kompakter zu reproduzieren und auf beliebige Disperisonsrelationen zu erweitern. Wir nachweisen, dass die Näherung einer linearen Dispersion in vielen Fällen unzureichend ist, um bestimmte interessante Effekte (beispielsweise gebundene Atom-Photon-Zustände) zu verstehen. Im zweiten Teil der Arbeit wurde das System um ein zweites ZNA erweitert, was zum Auftreten von Fluktuationskräften führt. Diese wurden anhand des Beispiels der Förster Energie untersucht, welche den strahlungsfreien Anteil des Energietransfers beschreibt. Es wurde nachgewiesen, dass dies für unser Modellsystem im Rahmen der RWA der einzig relevante Anteil ist und ausserdem nur für beschränkte Dispersionsrelationen existiert. Wir konnten zeigen, dass sowohl die Stärke als auch die Form der zugehörigen Potentiale stark vom Anfangszustand des Systems abhängt. Dies eröffnet interessante Perspektiven für die Erzeugung maßgeschneiderter Kraftprofile zwischen beiden Atomen. / This PhD Thesis deals with the theoretical aspects of the so called waveguide quantum electrodynamics (WQED). This part of physics deals with the interaction of matter and light which is confined to just one spatial dimension. This area of science experiences growing importance since its formation in the 1990s. The main reason for this are the diverse application possibilities such as the construction of quantum computers as well as classical computers on an optical basis. Furthermore pump-probe experiments using WQED are a promising direction of current research. All this topics are relying on a exact understanding of the underlying physical processes and this thesis shall make a contribution to this. For this purpose we developed a formalism, which relies on Feynman diagrams. The first model system which was investigated in this context consists of a 1D optical waveguide coupled to a two level system (TLS). We where able to reproduce many known results in a physically more transparent and mathematically more compact fashion. Furthermore we generalized this results to arbitrary dispersion relation and showed that the approximation of a linear dispersion is insufficient to describe many physical effects, like atom-photon bound states for example.\\ In the second part of this work we generalized the model system by adding an additional TLS, which supports the occurrence of fluctuation forces. Those where investigated in great detail at the example of the Förster energy, which describes the radiationless part of energy fluctuations. It was shown that this is the only relevant contribution as long as the RWA is valid and only occurs for bounded dispersion relations. We proved that the strength as well as the shape of the corresponding potential strongly depends on the initial state of the system, which opens interesting perspectives for the creation of tailored force profiles between both atoms. All calculations where done analytically as well as numerically.

<p>Strong coupling between light and matter has been demonstrated both in classical</p><p>cavity quantum electrodynamics (QED) systems and in more recent circuit-QED</p><p>experiments. This enables the generation of strong nonlinear photon-photon interactions</p><p>at the single-photon level, which is of great interest for the observation</p><p>of quantum nonlinear optical phenomena, the control of light quanta in quantum</p><p>information protocols such as quantum networking, as well as the study of</p><p>strongly correlated quantum many-body systems using light. Recently, strong</p><p>coupling has also been realized in a variety of one-dimensional (1D) waveguide-</p><p>QED experimental systems, which in turn makes them promising candidates for</p><p>quantum information processing. Compared to cavity-QED systems, there are</p><p>two new features in waveguide-QED: the existence of a continuum of states and</p><p>the restricted 1D phase space, which together bring in new physical effects, such</p><p>as the bound-state effects. This thesis consists of two parts: 1) understanding the</p><p>fundamental interaction between local quantum objects, such as two-level systems</p><p>and four-level systems, and photons confined in the waveguide; 2) exploring</p><p>its implications in quantum information processing, in particular photonic</p><p>quantum computation and quantum key distribution.</p><p>First, we demonstrate that by coupling a two-level system (TLS) or three/fourlevel</p><p>system to a 1D continuum, strongly-correlated photons can be generated</p><p>inside the waveguide. Photon-photon bound states, which decay exponentially as a function of the relative coordinates of photons, appear in multiphoton scattering</p><p>processes. As a result, photon bunching and antibunching can be observed</p><p>in the photon-photon correlation function, and nonclassical light source can be</p><p>generated on demand. In the case of an N-type four-level system, we show</p><p>that the effective photon-photon interaction mediated by the four-level system,</p><p>gives rise to a variety of nonlinear optical phenomena, including photon blockade,</p><p>photon-induced tunneling, and creation of single-photon states and photon</p><p>pairs with a high degree of spectral entanglement, all in the absence of a cavity.</p><p>However, to enable greater quantum networking potential using waveguide-</p><p>QED, it is important to study systems having more than just one TLS/qubit.</p><p>We develop a numerical Green function method to study cooperative effects in</p><p>a system of two qubits coupled to a 1D waveguide. Quantum beats emerge in</p><p>photon-photon correlations, and persist to much longer time scales because of</p><p>non-Markovian processes. In addition, this system can be used to generate a</p><p>high-degree of long-distance entanglement when one of the two qubits is driven</p><p>by an on-resonance laser, further paving the way toward waveguide-QED-based</p><p>quantum networks.</p><p>Furthermore, based on our study of light-matter interactions in waveguide-</p><p>QED, we investigate its implications in quantum information processing. First,</p><p>we study quantum key distribution using the sub-Possonian single photon source</p><p>obtained by scattering a coherent state off a two-level system. The rate for key</p><p>generation is found to be twice as large as for other sources. Second, we propose</p><p>a new scheme for scalable quantum computation using flying qubits--propagating</p><p>photons in a one-dimensional waveguide--interacting with matter qubits. Photonphoton</p><p>interactions are mediated by the coupling to a three- or four-level system,</p><p>based on which photon-photon -phase gates (Controlled-NOT) can be implemented for universal quantum computation. We show that high gate fidelity is</p><p>possible given recent dramatic experimental progress in superconducting circuits</p><p>and photonic-crystal waveguides. The proposed system can be an important</p><p>building block for future on-chip quantum networks.</p> / Dissertation