Ingénierie quantique de la lumière à l'aide d'atomes de Rydberg en cavité

Producing interactions between optical photons is a critical step for implementing quantum communication protocols and photonic quantum computing. Achieving this goal requires a strongly nonlinear medium. This thesis presents a setup designed to facilitate such interactions, consisting of a small atomic cloud placed inside a medium-finesse optical cavity and driven to a highly excited Rydberg state. The collective nature of the atomic ensemble enhances the coupling between the cloud and the cavity light field, while Rydberg interactions suppress double excitations within the cloud. As a result, the atomic cloud behaves as an effective single tree-level collective “superatom” coupled to the cavity mode. The first part of this thesis addresses imperfections in the ideal superatom model by accounting for multiple Rydberg excitations within the cloud. A new theoretical model is introduced to describe the effects of an imperfect Rydberg blockade, specifically in the regime where up to two Rydberg excitations are possible. This model is then validated through a series of experiments, demonstrating its applicability within the experimental framework. In the second part, the internal state of the superatom is coherently controlled and mapped onto a free-propagating light mode to generate an optical qubit encoded as a quantum superposition of 0 and 1 photons. Photon-correlation measurements reveal the single-photon nature of the emitted states, with a strong antibunching effect and only a residual 0.5% probability of emitting two photons per pulse. The photonics states are generated in the desired spatiotemporal mode with a high 60% efficiency. Using homodyne-tomography measurements, the density matrix of the output states is reconstructed, leading to the Wigner functions. In agreement with theoretical predictions, these functions are quadraturesqueezed for small qubit rotation angles, and develop a negative region when the angle approaches p. In the third part, the experimental setup is upgraded to incorporate an additional atomic cloud within the cavity. Initial experiments focus on introducing an excitation into one superatom and transferring it through the cavity to the second superatom. The process achieves an efficiency of approximately 30 %, in close agreement with t heoretical predictions. These results pave the way for future advancements in realizing more complex quantum gates and communication protocols.