Les gaz atomiques dégénérés représentent des systèmes modèles pour étudier la superfluidité. Ils offrent la possibilité d'explorer la physique en dimensions restreintes, profondémentdifférente par rapport au cas tridimensionnel. Nous disposons d'un gaz de Bose bidimensionnel dégénéré confiné dans un potentiel très anisotrope et dont on peut changer la géométrie dynamiquement. Une déformation contrôlée du piège permet d'exciter les modes collectifs du gaz. Nous avons fait d'abord une analyse en composantes principales du gaz, et nous avons montré que ces dernières coïncident avec les modes de Bogoliubov. Nous avons ensuite effectué une étude détaillée du mode ciseaux, dont nous nous servons pour sonder le caractère superfluide du gaz, en développant une nouvelle technique d'analyse, appelée "analyse de la moyenne locale". Enfin nous avons réalisé un piège en anneau, obtenu à l'intersection d'un piège en forme de bulle et du potentiel optique d'un faisceau qui présente un noeud d'intensité au centre, la "double nappe", et nous proposons différentes protocoles de mise en rotation des atomes dans l'anneau. / Degenerate atomic gases can be a versatile tool to study superfluidity. They also offer the possibility to explore the low-dimensions physics, which is deeply different from the three dimensional case. We prepare a degenerate Bose gas in a very anisotropic trap, dynamically adjustable. A controlled deformation of the trapping potential can excite the collective modes of the trapped cloud. First we perform a « principal components analysis » of the gas and we show that the principal components coincide with the Bogoliubov modes. We then restrain our analysis on the scissors mode, which we use to probe superfluidity of the sample, by introducing a new analysis technique, called « local average analysis ». Finally I will report on the realization of a ring trap, obtained by superposing a double sheet light beam to a bubble trap, and describe the different possibilities we planned to set atoms into rotation.

Cold bosonic atoms trapped in optical lattices formed by standing wave interference patterns of multiple laser beams constitute excellent emulators of models of strongly correlated quantum systems of bosons. In this thesis, we develop and deploy strong-coupling expansion (i.e., an expansion in terms of the ratio of the inter-site hopping amplitude of the bosons to the strength of their interactions) techniques for studying the properties of three different instances of such systems. In the first instance, we have used strong coupling expansion techniques to calculate the density pro le for bosonic atoms trapped in an optical lattice with an overall harmonic trap at finite temperatures and large on site interaction in the presence of super fluid regions. Our results match well with quantum Monte Carlo simulations at finite temperature. We present calculations for the entropy per particle as a function of temperature which can be used to calibrate the temperature in experiments. Our calculations for the scaled density in the vacuum-to-super fluid transition agree well with the experimental data for appropriate temperatures. We also discuss issues connected with the demonstration of universal quantum critical scaling in the experiments. Experimental realizations of “atomtronic" Josephson junctions have recently been created in annular traps in relative rotation with respect to potential barriers that generate the weak links. If these devices are additionally subjected to optical lattice potentials, then they can incorporate strong-coupling Mott physics within the design, which can modify the behaviour and can allow for interesting new configurations of system generated barriers and of super fluid ow patterns. we have examined theoretically the behavior of a Bose super fluid in an optical lattice in the presence of an annular trap and a barrier across the annular region which acts as a Josephson junction. As the fluid is rotated relative to the barrier, it generates circulating super-currents until, at larger speeds of rotation, it develops phase slips which are typically accompanied by vortices. We use a finite temperature strong-coupling expansion about the mean- held solution of the Bose Hubbard model to calculate various properties of the device. In addition, we discuss some of the rich behavior that can result when there are Mott regions within the system. Rubidium-Cesium dipolar molecule formation through Feshbach resonance is an area of great current interest, for, the dipolar molecules, once formed, interact via v long range dipolar forces, leading to possibilities of novel phases. Experimentalists currently make such systems mostly using trial and error, and the resulting efficiencies for molecule formation tend to be low. With a goal to assist cold-atom experimentalists to achieve higher e ciencies of molecule formation, we have estimated the trap parameters for Rb and Cs atoms in a 3D optical lattice required to create single occupancy per site Mott phase for both the species in the same regions of the trap. We thus identify the ne tuning of the external magnetic held near Rb-Cs Feshbach resonance required to achieve highest probability for creating single Rb-Cs Feshbach molecules in the system. We have used the Falicov-Kimball model to describe the relevant system and strong-coupling expansions about the mean- held solution to calculate the density pro les for both species and efficiency for molecule formation, determined by overlapping regions of single occupancy for both Rb and Cs, up to second order in the expansion. We also calculate the entropy per particle which serves as an estimation of the temperature in the experimental system