Superconductor-Insulator-Superconductor (SIS) tunnel junction mixers are now commonly used in astronomical receivers at (sub)millimeter wavelengths because of their superb sensitivity, high dynamic range and stability of operation. Niobium SIS mixers operating at frequencies well below the super- conducting gap (∼680 GHz) have already achieved quantum limited sensitivity. Therefore to further enhance the receiver sensitivity, increasing the Intermediate Frequency (IF) bandwidth of SIS mixers has became crucial. This thesis focuses on the theoretical modeling, design and experimental verifi- cation of Nb SIS mixers operating around 230 GHz with a wide IF bandwidth of 1–15 GHz. These mixers were designed for a single baseline heterodyne interferometer (GUBBINS), which is being built to observe the Sunyaev-Zel’dovich effect in the Cosmic Microwave Background. The combination of wide IF bandwidth SIS mixers and complex analogue correlators will allow GUBBINS to feature high surface brightness sensitivity, that helps to distinguish the weak SZ effect from the background noise. The SIS mixer detector system was assembled inside the GUBBINS cryostat together with the IF electronics and RF/LO optical systems. Low noise temperatures of around 71 K were then measured in the GUBBINS system. The Nb SIS mixer we have developed uses a unilateral finline and fully integrated planar circuits deposited on a silicon substrate, to couple the electromagnetic radiation from the waveguide into the SIS junction. The finline mixer allows a broad-band RF coupling, an easy integration of the on-chip planar circuits and an easy-to-fabricate mixer block. To achieve a wide IF bandwidth, the output impedance of the SIS mixer was well matched to the input impedance of the amplifier by a multi-stage microstrip circuit. Additionally, the planar circuit of the SIS mixer was also designed to have a small lumped inductance and capacitance. The SIS mixer chip was extensively simulated by rigorous electromagnetic software (HFSS) and the S-parameter was exported to a quantum mixing package SuperMix to produce a full-wave model of the mixer. Experimental testing yielded a best noise temperature of 50 K with an average noise temperature of 75 K over an RF bandwidth of 160 GHz– 260 GHz. We have performed thorough experimental and computational investigation of the IF system in particular the constraints on the bandwidth caused by the lumped element capacitance of the mixer chip and the matching of the output impedance of the mixer to the IF amplifier. Our conclusion is that a bandwidth of 1–15 GHz could be achieved using our mixer design, subject to the performance of the amplifier. Finally, a variable temperature load system was successfully developed and tested inside the cryostat, to avoid the losses from the room-temperature optics. We have showed that the noise temperature of the SIS detector could be reduced by as much as 15 K by testing the mixer using a variable temperature load inside the cryostat.

L'étude de la planète Vénus est particulièrement intéressante car elle possède de nombreuses caractéristiques avec la Terre, des dimensions et une composition semblables. Malgré ces éléments en commun, ces planètes sont foncièrement différentes à notre époque. Vénus possède une atmosphère cent fois plus dense à sa surface que celle de la Terre. Cette atmosphère est composée essentiellement de dioxyde de carbone là où l'atmosphère terrestre est composée d'un mélange de diazote et d'oxygène moléculaire. La composition chimique de l'atmosphère vénusienne pose de nombreux problèmes aux chimistes. Vénus fait l'objet d'observations modernes depuis la Terre, de missions spatiales depuis les années 70 et l'ESA a lancé en 2005 la mission Venus Express qui est restée en orbite autour de Vénus jusqu'en décembre 2014. L'explication de la distribution et des concentrations des espèces chimiques mesurées est étudiée par une communauté de modélisateurs en parallèle des observations. Cependant, ces modélisations étaient contraintes à utiliser des modèles 1D qui contenaient de nombreuses paramétrisations. Durant ma thèse j'ai mis à jour un modèle chimique, développé au LATMOS, et développé un modèle des nuages de Vénus qui ont tout deux été introduit à un GCM 3D de Vénus développé au LMD. Ce nouveau modèle permet d'appliquer en 3D les hypothèses les plus récentes sur les cycles chimiques vénusiens, tels que la décroissance du dioxyde de soufre et de la vapeur d'eau dans la couche nuageuse, les variations en latitudes du monoxyde de carbone, la stabilité du dioxyde de carbone, la composition des gouttelettes des nuages de Vénus et la masse condensée dans les nuages. / The study of Venus is particularly interesting on many aspects. Venus shares many features with the Earth, similar size and composition. For, despite some common elements, these planets are fundamentally different in our time. Venus has an atmosphere a hundred times denser at the surface than the Earth and is composed essentially of carbon dioxide where the earth's atmosphere is composed of a mixture of molecular nitrogen and molecular oxygen. The chemical composition of the Venusian atmosphere poses many problems for chemist. Venus is the subject of modern observations from Earth and then through space missions of the 70 ESA mission launched in 2005 which remained in orbit around Venus until December 2014. the explanation of the distribution and concentrations measured chemical species is studied by a community of modelers in parallel observations. However, these models were forced to use 1D models that contained numerous parameterization. During my PhD I have updated a chemical model developed at LATMOS, and developed a model of the Venus clouds which have both been introduced to Venus 3D GCM developed at LMD. It allows to apply in 3D the latest hypothesis about the Venusian chemical cycles such as, among others, the decrease of sulfur dioxide and water vapor in the cloud layer, changes of carbon monoxide in latitudes, the stability of carbon dioxide, the composition of clouds droplets and the condensed mass in those clouds.