L’exploration expérimentale de l’intrication quantique est un domaine de recherche très actif actuellement. Les systèmes d’électrodynamique quantique en cavité permettent notamment de générer de l’intrication dans des ensembles de plusieurs dizaines de particules, grâce à l’interaction à longue portée fournie par un mode du champ électromagnétique. Mon travail de thèse a consisté à mettre en place un nouveau dispositif expérimental afin d’assurer le chargement rapide et fiable d’une cavité optique miniaturisée à l’intérieur de laquelle l’adressage et la détection d’atomes uniques viennent s’ajouter à l’interaction collective fournie par le mode de la cavité. La tomographie quantique des états intriqués requière l’acquisition d’un grand nombre de données expérimentales, un soin tout particulier doit donc être porté quand à la stabilité et à la rapidité de répétition de l’expérience. Pour satisfaire à ces critères, un système de lasers particulièrement compacts et robustes, a été conçu et fabriqué afin d’assurer le refroidissement et l’interaction avec les atomes. Pour permettre la rapidité de répétition de l’expérience, une source de rubidium est utilisée en mode impulsionnel dans l’unique cellule à vide. Elle permet de moduler temporellement la pression atomique en fonction des besoins de l’expérience. Un chargement prompt du piège magnéto-optique est alors possible sans réduire la durée de vie des atomes dans la cavité, au moment où se déroulent les expériences. Le transport des atomes entre leur position de capture et le centre de la cavité s’effectue grâce à un piège dipolaire, déplacé selon son axe fort de confinement à l’aide d’un déflecteur acousto-optique. Cela permet un déplacement rapide, de l’ordre de la centaine de millisecondes pour une distance de 1,5 cm. Grâce à cette combinaison de techniques, ce nouveau dispositif expérimental devrait donner accès à la physique riche des systèmes intriqués à plusieurs dizaines de particules. / The study of quantum entanglement is a very active research field. Cavity quantum electrodynamics systems are versatile tools allowing for instance entanglement in mesoscopic systems, that is to say with about a hundred particles. The purpose of the new experimental setup built during this thesis is to reach the single atom manipulation and detection level while working with mesoscopic ensembles, collectively coupled to the cavity mode. Toward this goal, three new experimental techniques have been developed to enable reliable and fast data acquisition rate, essential to reconstruct entangled states by quantum tomography means. First, robust extended cavity diode lasers have been constructed, allowing acquisitions that last for days. Then, a pulsed atomic source has been set up, it combines the advantages of fast magneto-optical trap loading and long lifetime in conservative traps by modulating the pressure inside a single vacuum chamber apparatus on a short timescale. Finally, to ensure the fast transport of cold atomic ensembles from the magneto-optical trap to the cavity position, a dipole trap moved with an acousto-optic deflector has been built. This allows a transport over few centimetres leaving the full optical access to the atomic cloud for other manipulations. Thanks to this new experimental setup, we hope to contribute to the understanding of the rich physics lying beyond multi-particle entangled systems.

Shortage of drinking water in most parts of the world has been a growing concern in recent times. The situation has been getting worse in underdeveloped and developing countries due to sudden explosion in population growth and the growth in the industries. The natural resources for potable water are limited and unless a feasible solution is obtained in the near future, the ‘concern situation’ may turn into a ‘panic situation’. A possible solution for the shortage in drinking water is to use water from inexhaustible sources such as oceans and seas and make it potable using desalination process. However, the process of desalination is an energy intensive process which the poor countries can not afford. In recent times, the cost of fossil fuels has been skyrocketing. With the crude oil costing more than Rs.5200 (US$120) a barrel as on September 2008, even the rich countries like USA and the countries in the European Union are feeling the pinch of the energy cost. Alternate energy sources such as solar, wind, geo-thermal, hydrogen etc., have become the order of the day. These sources are renewable and are environmental friendly. More than one third of the populations of the world live in coastal areas. These areas get abundant amount of solar energy throughout the year. Utilization of this energy in desalination process would solve drinking water problem to a very great extent. However, construction of centralized desalination plants requires large amount of capital which the poor countries can ill-afford. An alternate solution would be to construct decentralized smaller plants that would require smaller capital to construct and easier to maintain. If the energy requirement is tapped from renewable sources such as solar, then the operational cost also becomes affordable for the poor countries. By taking care of the water requirement of the coastal areas through this process, one may save large amounts of money in transporting potable water from interior areas to the coastal areas. There would be enough water for the people living in the interior areas. The water bed level in the interior areas would gradually increase, thereby reducing the drinking water concerns significantly. In this thesis, a small scale stand alone power generation system for the desalination process is proposed that is suitable to provide clean potable water from sources such as sea water or brackish water. Solar energy is proposed as a source of energy for the proposed desalination system. This source is available in plenty in arid and semi-arid areas. It is free and is also friendlier to the environment. It is proposed to use solar energy in thermal form to obtain energy equivalent of ‘latent heat of vaporization’ for the vaporization process and in electrical form for operating the dc machines and electronic control units that are integral parts of the desalination system. The proposed desalination unit can be built as small as possible even to feed a single household’s requirement and hence can be conceptualized as decentralized units. These units would require considerably less capital to build, and would require minimum maintenance. The desalination process is based on flash evaporation wherein a heated liquid is subjected to a pressure reduction by passing through a throttling device resulting in an initially superheated state. In the proposed desalination process, the traditional flash evaporator is extended to include continuous dynamic pressure modulation to obtain an optimal flow rate for a specified energy input. The cost function or the performance index for optimization is defined as the ratio of flow rate to the energy spent. The optimal flow rate occurs at a specific chamber pressure for a given inlet water temperature. By operating at optimal pressure, significant energy is saved for a specific flow rate. This principle is validated experimentally and the results are presented and discussed in the thesis. This proposed scheme along with hybrid energy input will prove to be an attractive solution for community drinking water problem. A system needs to have a mathematical representation in order to predict the dynamic behavior of the system. This thesis proposes the bond graph method of modeling the physical system wherein the energy flow across the electrical, thermal and the hydraulic domains are included. A system may comprise of several subsystems and the energy flow in each subsystem may be in a different domain. A desalination system is such a system wherein the energy flow in subsystems is in different domains such as electrical, thermal and hydraulics. The bond graph approach is best suited for modeling of such systems where the power/energy flow across domains can be easily and seamlessly integrated. The thesis proposes a fifth order dynamic model of a single stage flash evaporation with pressure modulation using the bond graph approach. The proposed model incorporates the effects of chamber pressure variation, the entropy flow from the chamber due to conduction, convection, radiation and also the thermal dynamics of the water bodies in the evaporation, condensation and collection chambers. The proposed model is simulated in MATLAB/Simulink environment. The simulation results are compared with the experimental results to validate the model. This proposed model can be used for both analysis and synthesis of a desalination system. The desalination system is a complex system wherein multiple energy domains are involved. The thesis presents a systematic process for the design of the desalination system. Design of the desalination system involves design of multi domain subsystems. The design becomes much more complex if the energy source to run the system is solar/ hybrid solar based. The energy budget has to be carefully evaluated considering the worst case conditions for the availability of solar energy. Hence, the information on the quantum of solar energy available at any location is a critical parameter needed for design of the desalination system. A generic method of developing a solar insolation model for a specific region such as the Indian sub-continent is proposed in this thesis along with the validation of the model by comparing measured value with the values that are obtained from the model. As the insolation model is dependent on the water vapor content in the vertical column at the location, the methodology is further applied to develop a model for estimating the precipitated water vapor content in a vertical column for any location. The model is validated by comparing the computed values to the measured values. The thesis further presents the design and selection of the balance of the system. The selection of the balance of the system includes sizing of solar thermal plate collectors such as flat plate for pre-heating and paraboloid for vaporization, solar PV panels for operating pumps, actuator and control units, and battery for backup source for night loads and during ‘no-sun days’, criteria of selecting centrifugal pump for circulating condensation water, vacuum pump for dynamic pressure modulation and selecting linear actuator for Sun tracking of the paraboloid concentrator. A discussion on the electronic circuits used in the control scheme is presented in this thesis. This includes the circuit for maximum power point tracking, circuit for DC-DC conversion, circuit for pressure modulation, circuit for speed control of linear actuator, and finally the circuit for water level limiter. A discussion on the life cycle costing is also presented in this thesis. This is an important parameter that refers to the accumulated worth of all the costs related to building and operating the desalination plant during its life span. It is emphasized that the objective of the design process is to minimize the life cycle cost while meeting other performance requirements. Thus, life cycle costing is an essential part of the design cycle. The design methodology and the approach used to design the desalination system are implemented in the form of a toolbox in the MATLAB environment. The various functions of the toolbox are highlighted by a detailed step by step presentation of the design modules in the thesis. The modules form the components of the design toolbox for designing the proposed desalination system.