Nesta tese estudamos dois problemas diferentes na área de átomos ultra frios: Ondas de choque em condensados de Bose-Einstein e Espalhamento inelástico de átomos em um potencial de dois poços. No primeiro problema, estudamos o fluxo supersônico de um condensado de Bose-Einstein (BEC) através de um obstáculo macroscópico impenetrável delgado no sistema da equação de Schrödinger não-linear (NLS) bidimensional. Assumindo-se que a velocidade do fluxo é suficientemente alta, reduzimos assintoticamente o problema bidimensional original de valor de contorno para o fluxo estacionário através de um obstáculo alongado ao problema do pistão dispersivo unidimensional descrito pela NLS 1D dependente do tempo, no qual a coordenada original x reescalonada faz o papel de tempo e o movimento do pistão está vinculado à geometria do obstáculo. Duas ondas de choque dispersivas (DSWs) são geradas no fluxo, cada uma sendo formada em uma extremidade (frontal e traseira) do obstáculo. A DSW frontal é descrita analiticamente construindo-se soluções de modulação exatas para as equações de Whitham e a para a DSW traseira, empregamos a regra de quantização de Bohr-Sommerfeld generalizada para descrever a distribuição dos sólitons escuros. Propomos uma extensão da solução de modulação tradicional, a fim de incluir o padrão de ship-wave linear formado fora da região da DSW frontal. Realizamos simulações numéricas 2D completas e verificamos a validade das previsões analíticas. Os resultados deste problema podem ser relevantes para experimentos recentes sobre o fluxo de BECs através de obstáculos. No segundo problema, estudamos uma mistura atômica de dois átomos fermiônicos leves de spin 1/2 e dois átomos pesados em um potencial de dois poços. Processos de espalhamento inelástico entre ambas as espécies atômicas excitam os átomos pesados e renormalizam a taxa de tunelamento e a interação entre os átomos leves (efeito polarônico). A interação efetiva dos átomos leves muda de sinal e se torna atrativa quando o espalhamento inelástico é forte. Observamos também o cruzamento de níveis de energia, de um estado onde cada poço contém apenas um férmion (espalhamento inelástico fraco) para um estado onde um poço contém um par de férmions e ou outra está vazio (espalhamento inelástico forte). Identificamos o efeito polarônico e o cruzamento dos níveis de energia estudando-se a dinâmica quântica do sistema. / In this thesis we study two different problems in the field of ultracold atoms: Shock waves in Bose-Einstein condensates and Inelastic scattering of atoms in a double well. In the first problem, we study the supersonic flow of a Bose-Einstein condensate (BEC) past a slender impenetrable macroscopic obstacle in the framework of the twodimensional (2D) defocusing nonlinear Schr¨odinger equation (NLS). Assuming the oncoming flow speed sufficiently high, we asymptotically reduce the original boundary-value problem for a steady flow past a slender body to the one-dimensional dispersive piston problem described by the nonstationary NLS equation, in which the role of time is played by the stretched x-coordinate and the piston motion curve is defined by the spatial body profile. Two steady oblique spatial dispersive shock waves (DSWs) spreading from the pointed ends of the body are generated in both half-planes. These are described analytically by constructing appropriate exact solutions of the Whitham modulation equations for the front DSW and by using a generalized Bohr-Sommerfeld quantization rule for the oblique dark soliton fan in the rear DSW. We propose an extension of the traditional modulation description of DSWs to include the linear ship-wave pattern forming outside the nonlinear modulation region of the front DSW. Our analytic results are supported by direct 2D unsteady numerical simulations and are relevant to recent experiments on Bose-Einstein condensates freely expanding past obstacles. In the second problem, we study a mixture of two light spin-1/2 fermionic atoms and two heavy atoms in a double well potential. Inelastic scattering processes between both atomic species excite the heavy atoms and renormalize the tunneling rate and the interaction of the light atoms (polaron effect). The effective interaction of the light atoms changes its sign and becomes attractive for strong inelastic scattering. This is accompanied by a crossing of the energy levels from singly occupied sites at weak inelastic scattering to a doubly occupied and an empty site for stronger inelastic scattering. We are able to identify the polaron effect and the level crossing in the quantum dynamics.

In der vorliegenden Arbeit wird die erstmalige Erzeugung eines<br />Bose-Einstein-Kondensates in einer Mikrochip-Falle beschrieben; dies ist eine Magnetfalle für Neutralatome, die mithilfe stromführender Leiterbahnen auf einem Chipsubstrat gebildet wird. Die Eigenschaften dieser Chipfallen, speziell die hohen<br />Magnetfeldgradienten und -krümmungen, haben es ermöglicht, die<br />Bose-Einstein-Kondensation in weniger als einer Sekunde Verdampfungskühlzeit zu erreichen, was rund eine Größenordnung schneller als in bisher verwendeten Magnetfallen ist<br />und ein Faktor drei schneller als auf dem bisher schnellsten Weg in einer optischen Dipolfalle. Damit verbunden sind die Ansprüche<br />an den experimentellen Aufbau, insbesondere das Vakuumsystem und<br />den Laseraufbau, deutlich gesunken.<br /><br />Weiterhin wird der zerstörungsfreie Transport des Bose-Einstein-Kondensats entlang der Chipoberfläche über makroskopische Distanzen demonstriert wie auch erstmalig die Aufspaltung eines Kondensates in zwei getrennte Kondensate mit rein magnetischen Mitteln.<br /><br />Diese Resultate, nämlich kohärente Materie in einem integrierten<br />atomoptischen System manipulieren zu können, lassen hoffen, daß in<br />naher Zukunft Anwendungen wie Atominterferometrie, Untersuchungen<br />zu niederdimensionalen Quantengasen und<br />Quanteninformationsverarbeitung 'on-chip' verwirklicht werden<br />können.

Die Dynamik anfänglich aus dem Gleichgewicht gebrachter wechselwirkender Quantenvielteilchensysteme wirft aktuell noch spannende Fragen auf. In Bezug auf die Thermalisierung ist z.B. nach wie vor ungeklärt, in welcher Form sie überhaupt stattfindet und in welchen Observablen bzw. auf welcher Zeitskala sie zu beobachten ist. Eine ideale Grundlage zur Erforschung von Relaxationsdynamiken in wechselwirkenden Vielteilchensystemen bieten ultrakalte Quantengase aufgrund ihrer guten Kontrollier- und Variierbarkeit. Ein allgemeiner theoretischer Rahmen, auf dessen Basis solche Prozesse zu untersuchen sind, steht jedoch infolge der großen Anzahl der beteiligten Freiheitsgrade bisher nicht zur Verfügung. Für ultrakalte bosonische Gase stellt die Gross-Pitaevskii-Gleichung eines der wichtigsten theoretischen Werkzeuge dar, eine klassische Feldgleichung für die Kondensatwellenfunktion in Molekularfeldnäherung. Die ihr zugrunde liegende Näherung erlaubt jedoch keine nicht-trivialen Aussagen über den vollen N-Teilchenzustand, dessen Kenntnis für die Untersuchung einer möglichen Relaxationsdynamik unabdingbar ist. Um der theoretischen Beschreibung des vollen bosonischen Feldes einen Schritt näher zu kommen, untersucht die vorliegende Arbeit die Anwendung semiklassischer Methoden auf ultrakalte Bosegase. Diese sind in der Regel dann sehr genau, wenn die beteiligten Wirkungen groß gegenüber dem Planckschen Wirkungsquantum sind. Für bosonische Felder wird dieser Grenzfall durch die Bedingung einer großen Teilchenzahl ersetzt. Die immense Anzahl an Teilchen in den hier behandelten Vielteilchensystemen macht die Anwendung semiklassischer Methoden auf diesem Gebiet also vielversprechend. Als zentrales Modellsystem wird ein anfänglich aus dem Gleichgewicht gebrachtes ultrakaltes bosonisches Doppelmuldensystem betrachtet, das eine hochinteressante Dynamik aufweist, die auf das Wechselspiel der Tunneldynamik einerseits und der Wechselwirkung der Teilchen untereinander andererseits zurückzuführen ist. Als Referenz lassen sich aufgrund der speziellen Fallengeometrie im Rahmen der Zwei-Moden-Näherung die Ergebnisse einer numerisch exakten Untersuchung heranziehen. Durch den Einsatz der namhaften WKB-Quantisierung und des besonders aus der Molekülphysik bekannten Reflexionsprinzips wird hier ein geschlossener analytischer Ausdruck für die sogenannte Populationsdifferenz im Doppelminimum hergeleitet, der ausschließlich von den wenigen relevanten Systemparametern abhängt. Diese mächtige Formel erlaubt es nun zum ersten Mal, in quantitativer Weise die charakteristische Sequenz aus Oszillationen, Kollapsen und Revivals in Abhängigkeit der vorausgesetzten Parameter zu untersuchen. Nach dieser ersten erfolgreichen Anwendung semiklassischer Methoden im Modellsystem wird über die reduzierte Dynamik der Populationsdifferenz hinausgegangen. Mithilfe des semiklassischen Herman-Kluk-Propagators lässt sich selbst der volle N-Teilchenzustand untersuchen. Da es letztlich um die Beschreibung ultrakalter Bosonen in beliebigen Potentialen gehen soll, wird zunächst der Herman-Kluk-Propagator für eine Feldtheorie vorgestellt. Im Doppelmuldensystem zeigt sich dann in der Anwendung die semiklassische Propagation in der Lage, für alle untersuchten Parameterregime gute Übereinstimmung mit den numerisch exakten Ergebnissen zu liefern. Zusätzlich findet ein Abgleich der Resultate mit der Truncated Wigner Approximation statt, auf die im Forschungsgebiet ultrakalter Bosonen häufig zurück gegriffen wird. Diese beschreibt die Zeitentwicklung einer Wignerverteilung unter Aussparung der Quanteninterferenzen. In der vorliegenden Arbeit wird gezeigt, dass die Herman-Kluk-Propagation unter Berücksichtigung der Phasen weit über die Truncated Wigner Approximation hinausgeht: Sie gibt alle wichtigen Charakteristika der Dynamik im Doppelmuldensystem wieder. Um die Semiklassik auf ihre Aussagefähigkeit in Bezug auf eine noch komplexere Dynamik zu untersuchen, wird zum Abschluss das Drei-Topf-System betrachtet, das zusätzlich chaotische Regionen im Phasenraum aufweist. Auch hier zeigt sich, dass die semiklassische Berücksichtigung der Phasen die Truncated Wigner Approximation in den Schatten stellt. Allerdings ergeben sich durch die Instabilität der Trajektorien für stark chaotische Regime numerische Probleme, die es in der Zukunft zu lösen gilt. / The dynamics of initially non equilibrium interacting quantum many body systems is an ongoing and interesting field of research. It is still an open question in which form relaxation occurs in such systems, and in which observables and on which timescales a possible thermalization might appear. A perfect playground for the investigations of relaxation dynamics in interacting many body schemes is provided by ultracold quantum gases, which are easily to be controlled and varied in experiments. However, a general theoretical framework for the investigation of such processes is still missing, due to the huge amount of involved degrees of freedom. One of the main theoretical tools in the field of ultracold bosonic gases represents the famous Gross-Pitaevskii equation, a field equation for the Bose-Einstein condensate wave function in terms of a mean-field approximation. However, the underlying approximation prevents the possibility to draw non-trivial conclusions about the full N-particle state, the information of which is necessary for the analysis of relaxation processes. To gain the theoretical description of the full bosonic field, the present thesis deals with the application of semiclassical methods to ultracold boson gases. Those techniques become in general exact, as long as the involved actions are large compared to Planck's constant. For many body systems it turns out that semiclassics are expected to give good results also for the condition of high particle numbers, which is precisely fulfilled in these schemes, making the semiclassical approaches promising. As an essential model system an initially out of equilibrium ultracold bosonic double-well system is investigated. This configuration provides highly interesting dynamics due to the interplay of the tunneling dynamics on the one hand and the interaction amongst the particles on the other. The special trap geometry makes exact numerical calculations in the framework of the two-mode approximation available, which serve in the following as reference data. By applying the common semiclassical WKB approximation and the reflection principle known from molecule physics, a closed analytical expression for the so-called population imbalance of the bosons in the double-well is derived, depending only on the few relevant system parameters. This mighty formula allows for the first time the quantitative investigation of the characteristic sequence consisting of oscillations, collapse and revivals in dependence on the parameters of the system. Since the semiclassical approaches succeeded for the double-well model so far the so-called Herman-Kluk propagator is adopted, to go beyond the reduced dynamics of the population imbalance. The propagator provides the possibility to treat the full N-particle state theoretically and is introduced for the most general case of a bosonic quantum field. Its application to the double-well system yields for all investigated parameter regimes very good agreement with the numerical exact results. Furthermore the outcomes are compared to the Truncated Wigner approximation, which is frequently used in the research field of ultracold bosons. This approach pictures the time evolution of a Wigner distribution, without taking into account the quantum interferences. In the present thesis it is shown that the Herman-Kluk propagation goes clearly beyond the truncated Wigner approach by considering in addition the quantum phases: The propagator is able to reproduce all of the distinctive features of the double-well dynamics. In order to test the performance of semiclassical methods in matters of even more complex systems, the ultracold bosonic triple-well model is finally considered, which exhibits unlike the double-well scheme chaotic regions in phase space. It turns out that the semiclassical propagation outplays again the truncated Wigner approximation. On the other hand the instability of the highly chaotic trajectories causes numerical problems, which have to be solved in the future.

Bose-Einstein condensation (BEC) is a phenomenon in which identical bosons occupy the same quantum state below a certain critical temperature. A hallmark of BEC is the coherence between particles every particle shares the same quantum wavefunction and phase. This coherence has been demonstrated for the external (motional) degrees of freedom of the atomic condensates by interfering two condensates. In this thesis, the coherence is shown to extend to the internal spin degrees of freedom of a spin-1 Bose gas evidenced by the observed coherent and reversible spin-changing collisions. The observed coherent dynamics are analogous to Josephson oscillations in weakly connected superconductors and represent a type of matter-wave four-wave mixing. Control of the coherent evolution of the system using magnetic fields is also demonstrated. The studies on spinor condensates begin by creating spinor condensates directly using all-optical approaches that were first developed in our laboratory. All-optical formation of Bose-Einstein condensates (BEC) in 1D optical lattice and single focus trap geometries are developed and presented. These techniques offer considerable flexibility and speed compared to magnetic trap approaches, and the trapping potential can be essentially spin-independent and are ideally suited for studying spinor condensates. Using condensates with well-defined initial non-equilibrium spin configuration, spin mixing of F = 1 and F = 2 spinor condensates of rubidium-87 atoms confined in an optical trap is observed. The equilibrium spin configuration in the F = 1 manifold confirms that 87Rb is ferromagnetic. The coherent spinor dynamics are demonstrated by initiating spin mixing deterministically with a non-stationary spin population configuration. Finally, the interplay between the coherent spin mixing and spatial dynamics in spin-1 condensates with ferromagnetic interactions is investigated.

Ultracold atoms provide a powerful tool for studying quantum control of interacting many-body systems with well-characterized and controllable Hamiltonians. In this thesis, we demonstrate quantum control of a many-body system consisting of a ferromagnetic spin-1 Bose-Einstein condensate (BEC). By tuning the Hamiltonian of the system, we can generate either a phase space with an unstable hyperbolic fixed point or a phase space with an elliptical fixed point. A classical pendulum with a stable oscillation about the "down" position and an inverted pendulum with unstable non-equilibrium dynamics about the "up" position are classical analogs of the quantum spin dynamics we investigate in this thesis. In one experiment, we dynamically stabilize the system about an unstable hyperbolic fixed point, which is similar to stabilizing an inverted pendulum. In a second experiment, we parametrically excite the system by modulating the quadratic Zeeman energy. In addition, we demonstrate rectifier phase control as a new method to manipulate the quantum states of the many-body system. This is similar to parametric excitation and manipulation of the oscillation angle of a classical pendulum. These experiments demonstrate the ability to control a quantum system realized in a spinor BEC, and they also can be applied to other quantum systems. In addition, we extend our studies to atoms above the Bose-Einstein transition temperature, and we present results on thermal spin relaxation processes and equilibrium spin populations.

My dissertation treats the physics of ultracold gases, in particular of Bose-Einstein condensates with long-ranged interactions induced by admixing a small fraction of a Rydberg state to the atomic ground state. The resulting interaction leads to the emergence of supersolid states and to the self-trapping of a Bose-Einstein condensate.

The quantum dynamics and grand canonical thermodynamics of many-mode (one-, two-, and three-dimensional) interacting Bose gases are simulated from first principles. The model uses a lattice Hamiltonian based on a continuum second-quantized model with two-particle interactions, external potential, and interactions with an environment, with no further approximations. The interparticle potential can be either an (effective) delta function as in Bose-Hubbard models, or extended with a shape resolved by the lattice. Simulations are of a set of stochastic equations that in the limit of many realizations correspond exactly to the full quantum evolution of the many-body systems. These equations describe the evolution of samples of the gauge P distribution of the quantum state, details of which are developed. Conditions under which general quantum phase-space representations can be used to derive stochastic simulation methods are investigated in detail, given the criteria: 1) The simulation corresponds exactly to quantum mechanics in the limit of many trajectories. 2) The number of equations scales linearly with system size, to allow the possibility of efficient first-principles quantum mesoscopic simulations. 3) All observables can be calculated from one simulation. 4) Each stochastic realization is independent to allow straightforward use of parallel algorithms. Special emphasis is placed on allowing for simulation of open systems. In contrast to typical Monte Carlo techniques based on path integrals, the phase-space representation approach can also be used for dynamical calculations. Two major (and related) known technical stumbling blocks with such stochastic simulations are instabilities in the stochastic equations, and pathological trajectory distributions as the boundaries of phase space are approached. These can (and often do) lead to systematic biases in the calculated observables. The nature of these problems are investigated in detail. Many phase-space distributions have, however, more phase-space freedoms than the minimum required for exact correspondence to quantum mechanics, and these freedoms can in many cases be exploited to overcome the instability and boundary term problems, recovering an unbiased simulation. The stochastic gauge technique, which achieves this in a systematic way, is derived and heuristic guidelines for its use are developed. The gauge P representation is an extension of the positive P distribution, which uses coherent basis states, but allows a variety of useful stochastic gauges that are used to overcome the stability problems. Its properties are investigated, and the resulting equations to be simulated for the open interacting Bose gas system are derived. The dynamics of the following many-mode systems are simulated as examples: 1) Uniform one-dimensional and two-dimensional Bose gases after the rapid appearance of significant two-body collisions (e.g. after entering a Feshbach resonance). 2) Trapped bosons, where the size of the trap is of the same order as the range of the interparticle potential. 3) Stimulated Bose enhancement of scattered atom modes during the collision of two Bose-Einstein condensates. The grand canonical thermodynamics of uniform one-dimensional Bose gases is also calculated for a variety of temperatures and collision strengths. Observables calculated include first to third order spatial correlation functions (including at finite interparticle separation) and momentum distributions. The predicted phenomena are discussed. Improvements over the positive P distribution and other methods are discussed, and simulation times are analyzed for Bose-Hubbard lattice models from a general perspective. To understand the behavior of the equations, and subsequently optimize the gauges for the interacting Bose gas, single- and coupled two-mode dynamical and thermodynamical models of interacting Bose gases are investigated in detail. Directions in which future progress can be expected are considered. Lastly, safeguards are necessary to avoid biased averages when exponentials of Gaussian-like trajectory distributions are used (as here), and these are investigated.

The quantum dynamics and grand canonical thermodynamics of many-mode (one-, two-, and three-dimensional) interacting Bose gases are simulated from first principles. The model uses a lattice Hamiltonian based on a continuum second-quantized model with two-particle interactions, external potential, and interactions with an environment, with no further approximations. The interparticle potential can be either an (effective) delta function as in Bose-Hubbard models, or extended with a shape resolved by the lattice. Simulations are of a set of stochastic equations that in the limit of many realizations correspond exactly to the full quantum evolution of the many-body systems. These equations describe the evolution of samples of the gauge P distribution of the quantum state, details of which are developed. Conditions under which general quantum phase-space representations can be used to derive stochastic simulation methods are investigated in detail, given the criteria: 1) The simulation corresponds exactly to quantum mechanics in the limit of many trajectories. 2) The number of equations scales linearly with system size, to allow the possibility of efficient first-principles quantum mesoscopic simulations. 3) All observables can be calculated from one simulation. 4) Each stochastic realization is independent to allow straightforward use of parallel algorithms. Special emphasis is placed on allowing for simulation of open systems. In contrast to typical Monte Carlo techniques based on path integrals, the phase-space representation approach can also be used for dynamical calculations. Two major (and related) known technical stumbling blocks with such stochastic simulations are instabilities in the stochastic equations, and pathological trajectory distributions as the boundaries of phase space are approached. These can (and often do) lead to systematic biases in the calculated observables. The nature of these problems are investigated in detail. Many phase-space distributions have, however, more phase-space freedoms than the minimum required for exact correspondence to quantum mechanics, and these freedoms can in many cases be exploited to overcome the instability and boundary term problems, recovering an unbiased simulation. The stochastic gauge technique, which achieves this in a systematic way, is derived and heuristic guidelines for its use are developed. The gauge P representation is an extension of the positive P distribution, which uses coherent basis states, but allows a variety of useful stochastic gauges that are used to overcome the stability problems. Its properties are investigated, and the resulting equations to be simulated for the open interacting Bose gas system are derived. The dynamics of the following many-mode systems are simulated as examples: 1) Uniform one-dimensional and two-dimensional Bose gases after the rapid appearance of significant two-body collisions (e.g. after entering a Feshbach resonance). 2) Trapped bosons, where the size of the trap is of the same order as the range of the interparticle potential. 3) Stimulated Bose enhancement of scattered atom modes during the collision of two Bose-Einstein condensates. The grand canonical thermodynamics of uniform one-dimensional Bose gases is also calculated for a variety of temperatures and collision strengths. Observables calculated include first to third order spatial correlation functions (including at finite interparticle separation) and momentum distributions. The predicted phenomena are discussed. Improvements over the positive P distribution and other methods are discussed, and simulation times are analyzed for Bose-Hubbard lattice models from a general perspective. To understand the behavior of the equations, and subsequently optimize the gauges for the interacting Bose gas, single- and coupled two-mode dynamical and thermodynamical models of interacting Bose gases are investigated in detail. Directions in which future progress can be expected are considered. Lastly, safeguards are necessary to avoid biased averages when exponentials of Gaussian-like trajectory distributions are used (as here), and these are investigated.

Dans ce manuscrit, nous présentons une étude expérimentale d’un gaz de Bose de spin-1 avec des interactions antiferromagnétiques, réalisée pour des atomes de sodium ultra-froids dans l’état hyperfin F=1. Gr au refroidissement évaporatif, nous obtenons un condensat de Bose-Einstein (CBE) spineur, soit dans un piège très confinant (« piège 0D »), soit sous la forme d’un quasi-condensat quasi-unidimensionnel dans un piège très allongé. Les deux systèmes présentent un ordre magnétique a très basse température, qui résulte de la compétition entre les interactions d’échange et l’énergie Zeeman quadratique q dans un champ magnétique externe. Nous étudions dans un premier temps l’ordre magnétique se forme dans le piège 0D. À très bassetempérature deux phases magnétiques sont possible : une phase dite « antiferromagnétique » pour q < Us, ou une phase dite « à aimantation transverse » dans le cas inverse. Dans ce travail, nous nous plaçons près de la température critique. Nous mesurons plusieurs scénarios de condensation séquentielles en changeant la magnétisation et le champ magnétique externe, ou une composante Zeeman condense toujours en premier et ou l’ordre magnétique n’apparait qu’à une seconde température de condensation. Les résultats expérimentaux pour les températures critiques sont bien décrits par une théorie d’Hartree-Fock simplifiée dans les cas ou une seule composante Zeeman est condensée. Dans un second temps, nous étudions l’ordre magnétique du système quasi-unidimensionnel a basse température. On observe la formation de domaines de spin ou les composantes Zeeman se sépare spontanément en domaines disjoints en l’absence de force extérieure (par exemple, un gradient de champ magnétique). On étudie l’état d’équilibre du système en fonction de la magnétisation et du champ magnétique. On observe une transition de phase entre une phase miscible et une phase immiscible ou la composante Zeeman mF = 0 forme un domaine séparé de mF = ±1 dans le centre du piège. L’équation d’état d’un nuage polarisé (atomes dans l’état mF = +1) est utilisée pourmesurer la température du système. Enfin, nous mesurons la réponse mécanique a une force magnétique appliquée pour un système binaire mF = 0, +1. Nous mesures une exaltation de la réponse par rapport a l’attente na basée sur l’effet Zeeman habituel, d’un facteur qui peut varier de plusieurs dizaines a environ cent. La configuration spatiale des domaines est ainsi sensible a de très faibles gradients de champ magnétique inférieurs au mG/cm. / In this thesis we present the experimental study of a spin-1 Bose gas of ultra-cold Na atoms with antiferromagnetic interactions in the F=1 manifold. Thanks to evaporative cooling in optical traps we obtain, depending on the trap geometry, quasi-pure spinor Bose-Einstein condensates (BEC) in 0D traps and quasi-condensates in quasi-1D traps. The quantum-statistical Bose enhancement, typical of BEC, allows inter-component interactions (between the different Zeeman components) to order the system just below the Bose-Einstein condensation temperature. The magnetic ordering of the system is set: by contact interactions, that do not change the Zeeman populations, by spin-exchange interactions (U_s spin-exchange energy), that do, and by the quadratic Zeeman energy q. In particular, for q < U_s the system is in the antiferromagnetic phase while, for q > U_s, is in the transverse magnetised phase. We study first in which order the magnetic ordering appears, in the 0D trap, near to the critical temperature for BEC. We experimentally study different condensations scenarii varying q and magnetisation. The condensation of the different components is sequential and strongly influenced by interactions. We find a good agreement between the experimental data and a simplified Hartree-Fock model.Then we study the magnetic ordering, at T=0, in a quasi-1D trap. The system presents the formation of spin domains. We study the ground state of the system varying magnetisation and q. We observe a transition from the miscible to the immiscible phase, associated with the transition from the antiferromagnetic to the transverse magnetised phase. This is due to the relative strengths of inter-species contact interaction. To measure the temperature of the system, we measure the equation of state for a polarised cloud (all atoms in m_F=+1). Finally, we prepare the system in the immiscible phase m_F=0,+1 and we measure the spin-dipole polarisability of the system.

In recent years, improved control of the motional and internal quantum states of ultracold neutral atoms and ions has opened intriguing possibilities for quantum simulation and quantum computation. Many-body effects have been explored with hundreds of thousands of quantum-degenerate neutral atoms and coherent light-matter interfaces have been built. Systems of single or a few trapped ions have been used to demonstrate universal quantum computing algorithms and to detect variations of fundamental constants in precision atomic clocks. Now in our experiment we investigate how the two systems can be advantageously combined. We immerse a single trapped Yb+ ion in a Bose-Einstein condensate of Rb atoms. Our hybrid setup consists of a linear RF-Paul trap which is overlapped with a magnetic trap and an optical dipole trap for the neutral atoms. A first synergetic effect is the sympathetic cooling of the trapped ions to very low temperatures through collisions with the ultracold neutral gas and thus without applying laser light to the ions. We observe the dynamics of this effect by measuring the mean ion energy after having an initially hot ion immersed into the condensate for various interaction times, while at the same time monitoring the effects of the collisions on the condensate. The observed ion cooling effect calls for further research into the possibility of using such hybrid systems for the continuous cooling of quantum computers. To this end a good understanding of the fundamental interaction processes between the ion and the neutrals is essential. We investigate the energy dependent elastic scattering properties by measuring neutral atom losses and temperature increase from an ultracold thermal cloud of Rb. By comparison with a Monte-Carlo simulation we gain a deeper understanding of how the different parameters affect the collisional effects. Additionally, we observe charge exchange reactions at the single particle level and measure the energy-independent reaction rate constants. The reaction products are identified by in-trap mass spectrometry, revealing the branching ratio between radiative and non-radiative charge exchange processes.