A computação quântica adiabática tem sua pedra angular no teorema adiabático, cuja eficiência está relacionada tradicionalmente à proporção da variação temporal do Hamiltoniano que descreve o sistema e o gap mínimo entre o estado fundamental e o primeiro excitado. Normalmente, esse gap tende a diminuir quando aumenta o número de recursos (bit quântico: qubit) de um processador quântico, exigindo dessa maneira variações lentas do Hamiltoniano para assim garantir uma dinâmica adiabática. Entre os candidatos para a sua implementação física, estão os qubits baseados em circuitos supercondutores os quais têm um grande potencial, por causa de seu alto controle e escalabilidade promissora. No entanto, quando esses qubits são implementados, eles têm uma fonte intrínseca de ruído devido a erros de fabricação, que não podem ser desprezados. Por isso, nesta tese nós estudamos como os efeitos causados pelas flutuações dos parâmetros físicos do qubit afetam o comportamento da fidelidade da computação, realizando com esse propósito a simulação da dinâmica de cadeias de spin pequenas desordenadas. A partir do análise exaustivo desse estúdio foi possível propor uma estratégia que permite aumentar a fidelidade considerando um sistema ruidoso. Por outro lado, motivados pelo interesse de obter critérios suficientes e necessários para satisfazer uma computação quântica adiabática e pelo fato que ainda não existe uma condição de adiabaticidade geral apesar de existir inúmeras propostas, nós apresentamos um novo critério que manifesta suficiência para sistemas mais gerais e finalmente apresentamos evidências de que tal condição seria um quantificador consistente. / Adiabatic quantum computation has its cornerstone in the adiabatic theorem, whose efficiency is traditionally related to the ratio of the Hamiltonian temporal variation that describes the system and the minimum gap between the ground state and the first excited state. Usually, this gap tends to decrease when the number of quantum resources (quantum bit: qubit) of a quantum processor increases, thus it requires slow variations of the Hamiltonian to ensure an adiabatic dynamic. Among the candidates for its physical implementation are the qubits superconducting circuit-based which have great potential because of their high control and promising scalability. However, when these qubits are implemented, they have an intrinsic source of noise due to manufacturing errors that can not be despised. Therefore, in this thesis we study how the effects caused by the fluctuations of the physical parameters of the qubit affect the behavior of the fidelity of the computation, accomplishing with this purpose the simulation of the dynamics of small disordered spin chains. From the exhaustive analysis of this studio, it was possible to propose a strategy that allows to increase the fidelity considering a noisy system. On the other hand, motivated by the interest of obtaining sufficient and necessary criteria to satisfy an adiabatic quantum computation and the fact that there is still no general adiabaticity condition despite there being numerous proposals, we present a new criterion that manifests sufficiency for more general systems and we finally presented evidence that such a condition would be a consistent quantifier.

Submitted by Irene Nascimento (irene.kessia@ufpe.br) on 2016-07-12T19:37:30Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Tese_FlaviaPortela_2015.pdf: 24722134 bytes, checksum: ecaec07cae254145230ae32ad7de71e3 (MD5) / Made available in DSpace on 2016-07-12T19:37:30Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Tese_FlaviaPortela_2015.pdf: 24722134 bytes, checksum: ecaec07cae254145230ae32ad7de71e3 (MD5) Previous issue date: 2015-03-26 / CAPES / O estudo em sistemas de multicamadas supercondutoras alcan cou grande interesse na pesquisa de novos materiais, pois al em de apresentarem ampla aplicabilidade tecnol ogica, oferecem fascinantes possibilidades de observar novos fen^omenos na supercondutividade. Esta ultima caracter stica e a de maior motiva c~ao para esta tese, uma vez que pouco ainda se tem reportado a respeito da supercondutividade em multicamadas constitu das por supercondutor(S)/supercondutor(S0). Neste trabalho, estudamos tr^es novos sistemas nanoestruturados formados por supercondutores elementares, de baixa temperatura cr tica, tais como ni obio (Nb), chumbo (Pb) e estanho (Sn), a saber, a tricamada Nb(5)/Pb(500)/Nb(50) e as multicamadas Nb(100)/[Sn(50)/Nb(50)]7 e Cr(10)/Nb(100)/[Sn(50)/Nb(50)]7, onde o termo entre par^enteses indica a espessura da referida camada em nan^ometros e os colchetes indicam que a estrutura Nb/Sn e repetida 7 vezes. Para compara c~ao, tamb em s~ao analisados lmes de refer^encia de Nb, Pb e Sn. Os lmes e as multicamadas foram crescidos a temperatura ambiente por deposi c~ao via sputtering, utilizando fontes DC e RF. As amostras foram caracterizadas por difra c~ao de raios X de baixo ^angulo, microscopia eletr^onica de varredura e por microscopia de for ca at^omica. As propriedades supercondutoras foram estudadas atrav es de medidas magn eticas e de transporte el etrico. Magnetiza c~ao e resistividade foram medidas como fun c~oes da temperatura e do campo magn etico aplicado perpendicular e paralelamente as camadas. Os tr^es sistemas apresentaram transi c~oes supercondutoras abruptas, tanto na magnetiza c~ao quanto na resistividade em fun c~ao da temperatura, com os valores de TC obtidos de 7,2 K para o Nb/Pb/Nb, 5,2 K para o Nb/[Sn/Nb]7 e 3,7 K para o Cr/Nb/[Sn/Nb]7. As propriedades supercondutoras dos lmes de Pb, Nb e Sn, tais como TC, 0Hc2 e comprimentos caracter sticos apresentaram coer^encia com os valores reportados na literatura, con rmando a boa qualidade das camadas. Flux jumps s~ao observados nos loops de histerese em todos os sistemas, os quais s~ao atribu dos a instabilidades termomagn eticas. A amostra Nb/Pb/Nb apresentou uma curvatura positiva na depend^encia de 0Hc2(T), indicando que a supercondutividade ocorre preferencialmente na camada de Pb para T > T e preferencialmente na camada de Nb para T < T , de acordo com a teoria de Takahashi-Tachiki. Os comprimentos caracter sticos s~ao calculados, classi cando as multicamadas como supercondutores do tipo II. Veri camos grande in u^encia do material magn etico Cr na supercondutividade da multicamada, causando a diminui c~ao da temperatura cr tica do sistema e modi cando o comportamento dos campos cr ticos inferior e superior. A depend^encia de 0Hc1(T), em todos os sistemas, revelou um comportamento n~ao convencional, que e atribu do a uma manifesta c~ao de sistemas multicomponentes com componentes espacialmente separadas. / The study of superconducting multilayers systems has achieved much interest in the research of new materials. Besides their wide technological applicability, these structures o er fascinating possibilities to observe new phenomena in superconductivity. The latter characteristic is the fundamental motivation for this thesis, since a little has been reported about multilayers constituted of superconductor(S)/superconductor(S0). In this work, we have studied three new nanostructured systems formed by low critical temperature conventional superconductors, such as niobium (Nb), lead (Pb) and tin (Sn): the trilayer Nb(5)/Pb(500)/Nb(50) and the multilayers Nb(100)/[Sn(50)/Nb(50)]7 and Cr(10)/Nb(100)/[Sn(50)/Nb(50)]7. The term in parenthesis indicates the thickness of the layer in nanometers and the brackets that Nb/Sn structure is repeated 7 times. For comparison, the reference lms of Nb, Pb and Sn were also analyzed. The thin lms and multilayers were growth at room temperature via sputtering deposition, by using DC and RF sources. The samples were characterized for small angle X ray di raction, scanning electron microscopy (SEM), and atomic force microscopy (AFM). The superconducting properties were investigated through magnetic and electric transport measurements. The magnetization and resistivity were obtained as functions of temperature and magnetic eld applied both perpendicularly and parallel to the layers. All the multilayers systems have showed sharp superconducting transitions in the dependence of magnetization and resistivity with the temperature. The multilayers critical temperatures were found to be 7.2 K for Nb/Pb/Nb, 5.2 K for Nb/[Sn/Nb]7 and 3.7 K for Cr/Nb/[Sn/Nb]7. The superconducting properties of reference lms Pb, Nb and Sn, such as critical temperature, upper critical eld and characteristic lengths were consistent with the literature values, con rming the good quality of the samples. In the three systems ux jumps were observed, which are attributed to thermomagnetic instabilities. The Ginzburg-Landau parameter is estimated, classifying the multilayers as type II superconductors. In the case of Nb/Pb/Nb sample, it presented an upward curvature in the 0Hc2(T) diagrams, which is a signature of superconductivity nucleation in the each layer, in accordance with Takahashi-Tachiki theory for multilayered systems. We found out a noticeable in uence of the magnetic material Cr on the multilayer superconducting properties, reducing the critical temperature of the system and modifying the lower and upper critical elds behavior. The dependence of 0Hc1(T), of all the multilayers, revealed a non-conventional behavior feature, which is consistent with a multicomponent behavior with spatially separated components.

Abstract Elementary experiments of atomic physics and quantum optics can be reproduced on a circuit board using elements built of superconducting materials. Such systems can show discrete energy levels similar to those of atoms. With respect to their natural cousins, the enhanced controllability of these ‘artificial atoms’ allows the testing of the laws of physics in a novel range of parameters. Also, the study of such systems is important for their proposed use as the quantum bits (qubits) of the foreseen quantum computer. In this thesis, we have studied an artificial atom coupled with a harmonic oscillator formed by an LC-resonator. At the quantum limit, the interaction between the two can be shown to mimic that of ordinary matter and light. The properties of the system were studied by measuring the reflected signal in a capacitively coupled transmission line. In atomic physics, this has an analogy with the absorption spectrum of electromagnetic radiation. To simulate such measurements, we have derived the corresponding equations of motion using the quantum network theory and the semi-classical approximation. The calculated absorption spectrum shows a good agreement with the experimental data. By extracting the power consumption in different parts of the circuit, we have calculated the energy flow between the atom and the oscillator. It shows that, in a certain parameter range, the absorption spectrum obeys the Franck-Condon principle, and can be interpreted in terms of vibronic transitions of a diatomic molecule. A coupling with a radiation field shifts the spectral lines of an atom. In our system, the interaction between the atom and the field is nonlinear, and we have shown that a strong monochromatic driving results in energy shifts unforeseen in natural or, even, other artificial atoms. We have used the Floquet method to calculate the quasienergies of the coupled system of atom and field. The oscillator was treated as a small perturbation probing the quasienergies, and the resulting absorption spectrum agrees with the reflection measurement.

The rapid development of second generation (2G) high temperature superconducting (HTS) wires in the last decade has made it possible to wind high quality 2G HTS coils. These 2G HTS coils show promise for future applications such as magnetic resonance imaging (MRI) magnets, electrical machines, magnetic levitation trains, energy storage, etc. 2G HTS coils can be operated using either dc current or ac current. Several important issues have yet to be resolved, such as how to properly magnetise an HTS coil under dc conditions, or how to minimise losses under ac conditions. These problems should be carefully studied before the 2G HTS coils can be widely applied in scientific and industrial applications. This thesis focuses on emerging HTS flux pump technology for HTS coils operating in a dc environment. HTS flux pump technology applies a travelling magnetic wave to fully magnetise an HTS coil, which is both efficient and economical, and has in recent years been proven feasible. However, the underlying physics of this technology are so far poorly understood. In order to study the influence of a travelling magnetic wave on HTS films such as YBa2Cu3O7-δ, two types of circular-type magnetic flux pump (CTMFP) devices were proposed and built. These novel devices generate an annular-shape travelling magnetic wave. The first type was the original CTMFP magnet, which produces the longest wavelength of travelling wave. The second type was the updated CTMFP magnet, which can produce a shorter wavelength of travelling wave (1/2 of the original CTMFP magnet in the six phase connection and 1/4 in the three phase connection). A 2 inch diameter round shape YBCO thin film (200 nm thick of the YBCO layer) and a 46 mm× 46 mm square shape YBCO tape (1.0 µm thick of the YBCO layer, with a hole of Φ26 mm in the centre) were tested. When using a round shape YBCO thin film and the original CTMFP magnet, it was found that the travelling wave tends to decrease the existing critical magnetic gradient inside the YBCO film. The experiment was repeated under different conditions, such as zero-field cooling (ZFC), field cooling (FC), delta-shape trapped field, etc. A simulation based on the H-formulation using FEM software revealed that, after application of the travelling wave, the current density distribution inside the round shape YBCO sample was disturbed, becoming much lower than its critical current density JC. This discovery is interesting because the Bean model suggests that the current density inside a type-II superconductor should be equal to either +JC or - JC (the critical state model). It was found that a round shape YBCO sample follows the Bean model prediction for the homogeneous oscillating field (homogeneous in space), which suggests that the travelling wave is more efficient for transporting the magnetic flux inside YBCO film, compared to a homogeneous oscillating field. An updated CTMFP magnet was designed and built to investigate the influence of the degree of field inhomogeneity on the change of an existing critical magnetic gradient. The results were compared between the six phase connection (1/2 wavelength of the original CTMFP magnet) and the three phase connection (1/4 wavelength of the original CTMFP magnet). It was found that with a travelling wave of consistent amplitude, by shortening the wavelength, the change of magnetic gradient is made stronger. The result supports the assumption that the field inhomogeneity in space may have an important influence on the magnetisation of a YBCO sample. Additionally, in the case of a three phase connection (1/4 wavelength), by reversing the direction of the travelling wave, a different magnetisation profile was obtained, which suggests that the experiment may have detected a macroscopic “magnetic coupling” phenomenon. However, this result needs further study before it can be confirmed. The square shape YBCO sample was tested by applying a travelling wave in a dc background field under FC conditions. The square shape YBCO sample has a centre hole (Φ26 mm), which is closest to the condition of an HTS coil (single layer instead of multi-layer). However, in the experiment there was no clear change of magnetic flux inside the superconducting loop after application of the travelling wave. This might be attributed to the fact that, the field inhomogeneity is not strong enough to cause flux migration in the experiments, and the YBCO layer is relatively thicker which increases the difficulties. Moreover, the width of the superconducting region is relatively small (10 mm), in order to help magnetic flux migrate into the superconducting loop, the field inhomogeneity must be strong enough in the superconducting region, which increases the technical difficulties. However, this might be able to be accomplished by increase the amplitude of the travelling waves. Some experiments will be carried out in the future. The experimental findings in this thesis can not only aid in understanding the mechanism of HTS flux pump technology for an HTS coil, but also can help in understanding ac loss from a coil exposed to a travelling wave. As was suggested by the experimental results, the magnetisation of the YBCO film due to the travelling wave is very different from the magnetisation induced by a homogeneous oscillating field. Under operational conditions, such as inside an HTS motor, the HTS coils experience a travelling wave rather than a homogeneous oscillating field. This thesis discusses the difference in resultant ac loss from a travelling wave and a homogeneous oscillating field of the same amplitude. It was found that, for the round shape YBCO sample, the ac loss from a travelling wave is about 1/3 of the loss from a homogeneous oscillating field. The regions in which the ac loss occurred are also different between a travelling wave and a homogeneous oscillating field. These results suggest that the travelling wave cannot be equated to a homogeneous oscillating field when calculating ac loss. In conclusion, this thesis studies two novel experimental devices, built to study the magnetisation of YBCO films under the influence of a travelling wave. Several novel electromagnetic behaviours were observed in the YBCO films under the influence of a travelling wave, which may help improve understanding of HTS flux pump technology for an HTS coil, and the ac loss induced by a travelling wave.

No description available.

No description available.

Au cours des dernières décennies, le phénomène de quench a été une des problématiques les plus importantes abordées dans les conceptions d’aimants supraconducteurs. En effet, la transition de quench d’un aimant de son état supraconducteur à son état normal induit une grande quantité de l’énergie par effet Joule. Cet apport de chaleur va ensuite augmenter rapidement la température du conducteur ainsi que la pression du liquide de refroidissement à l’hélium. Le dépassement d’un certain seuil sur ces deux paramètres peut engendrer une détérioration irréversible à l’aimant et au système de refroidissement cryogénique. Afin de mettre en évidence les comportements de quench des bobines supraconductrices à champ toroïdal (TF) du Tokamak JT-60SA, nous avons réalisé des études expérimentales et numériques sur les phénomènes thermohydrauliques résultant du quench d’un aimant supraconducteur fabriqué en câble-en-conduit conducteur (CICC) et refroidi par l’écoulement forcé à l’hélium supercritique. Dans ce cadre, toutes les 18 TF bobines de JT-60SA ont été testées dans une configuration à une seule bobine dans leurs conditions de fonctionnement nominales de courant et de température (25,7 kA et 5 K). Une augmentation progressive de la température a été appliquée à l'entrée de l'hélium jusqu'à la température de quench, suivie d'une décharge rapide du courant dès que le quench est détecté pour protéger l'aimant. Les analyses expérimentales de ces tests ont d'abord permis d'identifier plusieurs phases dynamiques très différentes pendant toute la propagation de quench. Ensuite, les phénomènes physiques parcourant chacune de ces phases ont été étudiés et les plus prédominants ont été mis en évidence tels que les charges thermiques externes, les performances magnétiques des brins, les transferts thermiques conducto-convectifs entre conducteurs et hélium ou encore l'expulsion d'hélium et le reverse flow. Sur la base de ces analyses expérimentales, un modèle numérique d’une seule galette a été développé dans le code THEA afin d'analyser un phénomène physique à la fois sans construire un modèle global trop complexe de l'ensemble de l'aimant. Ce modèle d’une seule galette a été validé sur les données d'expériences de quench et a été appliqué avec succès pour faire d'autres analyses plus détaillées des phénomènes physiques ainsi que des phases dynamiques identifiées pendant la propagation de quench des TF bobines. Ce modèle numérique a même permis d'identifier certains phénomènes prépondérants qui n'ont pas pu être étudiés dans l'analyse expérimentale, tels que l'impact des instabilités des conditions de test sur la dynamique de quench. Les très bons résultats de ce modèle et sa cohérence avec les analyses physiques expérimentales en font une étape très intéressante vers la modélisation complète de toute la TF bobine de JT-60SA et l'étude de son comportement de quench dans une vraie machine Tokamak et non en conditions d'essais. / During the last decades, the quench phenomenon has been one of the most important issues addressed in the superconducting magnets designs. Indeed, the quench transition of a magnet from its superconducting state to its normal state induces a large deposition of the Joule effect energy leading to an abrupt temperature increase in the conductor as well as a large pressure rise in the helium coolant. Any excess of these two parameters can cause an irreversible damage either to the magnet or to the cryogenic system. In order to achieve a better understanding of the quench behavior of the TF coils in the superconducting Tokamak JT-60SA, we carried out both experimental and numerical studies of the thermohydraulic phenomena taking place during the quench of a superconducting magnet manufactured with Cable-In-Conduit Conductor and cooled in forced flow with supercritical helium. In this framework, all the 18 JT-60SA TF coils were tested in a single coil configuration at their nominal operating conditions of current and temperature (25.7kA and 5K). A progressive temperature increase has been applied to the helium inlet up to the quench temperature, followed by a current fast discharge as soon as the quench is detected to protect the magnet. The experimental analyses of these tests allowed first to identify several very different dynamic phases in the overall quench propagation time. Then, the physical phenomena driving each one of these phases have been studied and the most predominant ones have been highlighted such as the external heat loads, the strands magnetic performances, the conductive and convective heat transfers between conductors and helium or even the helium expulsion and reverse flow. Based on these experimental analyses, a single pancake numerical model has been developed in the THEA code in order to analyze one physical phenomenon at a time without building a too complex global model of the entire magnet. This single pancake model has been validated on the quench experiments data and has been successfully applied to make further more detailed analyses of the physical phenomena as well as the dynamic phases identified during the TF coils quench propagation. This numerical model even allowed identifying some driving physical phenomena that could not be studied in the experimental analysis, such as the impact of the testing conditions instabilities on the quench dynamics. The very good results of this model and its coherence with physical experimental analyses makes it a very interesting step towards the full modelling of the entire JT-60SA TF coil and the study of its quench behavior in real Tokamak and not test facility conditions.

Abstract In this thesis we study the interaction between radiation and matter using superconducting circuits that behave analogously with the conventional photon-atom interaction in quantum optics. The research is done with a system consisting of a waveguide resonator (radiation) strongly coupled to a transmon device (matter). We focus on the phenomena caused by strong coupling between the radiation and matter, and by driving the resonator to higher excited states with a strong monochromatic radiation. These have been studied little in the traditional radiation-matter systems. Increasing the strength of the monochromatic radiation drive, the dynamics of the system experiences a transition from the quantum to the classical regime. Also, the free-particle states of the transmon start being populated. In the weak driving limit, the transmon can be regarded as a two-state system. As a consequence, the resonator-transmon system is conventionally discussed in terms of the linear Jaynes–Cummings model. However, for strong coupling the Bloch–Siegert shift, caused by the terms neglected in the Jaynes–Cummings model, is strong and the Jaynes–Cummings model is insufficient for describing the dynamics of the system. We study the effects caused by strong coupling and the excitation of the higher transmon states instigated by the driving of the resonator. With reflection spectroscopy, we measure the absorption spectrum of the system and compare this with the spectrum calculated numerically using the Floquet–Born–Markov approach. We find that, in the region of the quantum-to-classical transition, the two-state approximation for the transmon is insufficient and the higher transmon states are necessary for accurate simulations. By calculating the average resonator occupation, we compare different numerical models: the Lindblad master equation, the Floquet–Born–Markov, and the semiclassical model. Coupling a transmon to a resonator shifts the energy levels of the resonator. This shift in the energy levels prevents the higher resonator states from being populated if the system is weakly driven with a frequency that is near the resonance frequency of the resonator. We simulate this photon blockade numerically and show that the blockade is substantially different for the two-state and multistate transmon approximations. / Original papers Original papers are not included in the electronic version of the dissertation. Pietikäinen, I., Danilin, S., Kumar, K. S., Vepsäläinen, A., Golubev, D. S., Tuorila, J., &amp; Paraoanu, G. S. (2017). Observation of the Bloch-Siegert shift in a driven quantum-to-classical transition. Physical Review B, 96(2). https://doi.org/10.1103/PhysRevB.96.020501 http://jultika.oulu.fi/Record/nbnfi-fe201803073899 Pietikäinen, I., Danilin, S., Kumar, K. S., Tuorila, J., &amp; Paraoanu, G. S. (2018). Multilevel Effects in a Driven Generalized Rabi Model. Journal of Low Temperature Physics, 191(5–6), 354–364. https://doi.org/10.1007/s10909-018-1857-8 http://jultika.oulu.fi/Record/nbnfi-fe2018061325770 Pietikäinen, I., Tuorila, J., Golubev, D. S., &amp; Paraoanu, G. S. (2019) Quantum-to-classical transition in the driven-dissipative Josephson pendulum coupled to a resonator, Manuscript. https://arxiv.org/abs/1901.05655

Les accélérateurs de particules hyperfréquences sont au cœur de projets d’envergure aux visées scientifiques (comme l’European Spallation Source) ou énergétiques (comme le réacteur hybride MYRRHA). Pour ces applications, les cavités résonnantes composant ces accélérateurs doivent atteindre des champs accélérateurs très importants. Elles ont alors besoin d’une grande puissance RF. Le coupleur de puissance doit permettre d’injecter cette puissance dans la cavité tout en garantissant une grande fiabilité. L’étude d’un coupleur à 704,4MHz a été réalisée. Les comportements radiofréquences, thermiques et mécaniques ont été modélisés pour une puissance de 50kW en réflexion toutes phases. Les performances simulées permettent d’envisager son utilisation comme coupleur pour les cavités elliptiques de l’ADS MYRRHA. L’industrialisation d’un tel coupleur a également été étudiée. Cette étude a donné lieu à la fabrication de deux prototypes. L’étude d’un coupleur à 352MHz a également été réalisée afin de répondre aux exigences de la cavité SPOKE ESS. Un nouveau processus d’assemblage, impliquant une méthode de précontrainte de la céramique, sera exposé. Ce processus diminue le nombre d’étapes de fabrication et renforce mécaniquement la fenêtre du coupleur. Cette thèse a également permis la validation de l’extension aux ondes progressives d’un logiciel de simulation 3D du multipactor : Musicc3D. Les simulations et les mesures ont été favorablement comparées pour les coupleurs SPIRAL2 et XFEL. / Nowadays, the number of projects aiming at building high intensity proton linear accelerators is increasing thanks to a large field of applications: particles & nuclear physics, spallation sources and some applications in material sciences, biology and nuclear waste reprocessing. All these linear accelerator projects are based on superconducting technology that allows high accelerating gradients in continuous mode. The RF power coupler is one of the main components of the accelerator. It is designed to transmit the radio frequency power from the waveguide at room temperature to the cavity at 4 Kelvin with high reliability. In this thesis, a study of a 704.4MHz power coupler was carried out. The RF, thermal and mechanical behaviors of the coupler were simulated for 50kW full reflection continuous wave. This power coupler could be used as RF injector for MYRRHA elliptical cavities. The fabrication process has been established and two prototypes were build. A study of a 352MHz power coupler was also carried out to fulfill ESS Spoke cavities needs. The obtained design satisfies the specifications and lowers the maximum electric field in the power coupler window compared to existing design. Then a new fabrication process was developed using prestressed ceramic. Brazing operations have halved with this new process. It was shown that the prestressed could enhance the strengths ceramic. The last study in this thesis consisted in validating an extension to progressive waves of 3D simulation software of Multipacting, Musicc3D. The results obtained were favorably compared to the measurements for SPIRAL2 and XFEL power couplers.

En microscopie IRM, la sensibilité de détection est critique pour obtenir des images avec un rapport signal sur bruit suffisant car l’intensité du signal RMN devient extrêmement faible. Une stratégie alternative à l’utilisation de champs statiques élevés consiste à améliorer les performances des antennes radiofréquences qui détectent le signal d’IRM. Plus particulièrement, la stratégie dans laquelle s’inscrit ce travail de thèse vise à exploiter la haute sensibilité des antennes miniatures supraconductrices basées sur le principe des résonateurs monolithiques à ligne de transmission avec comme objectif à terme la mise en réseau de ce type d’antennes. Le développement d’un tel réseau représente un enjeu instrumental majeur car cela permet de profiter de la haute sensibilité intrinsèque des antennes miniatures supraconductrices tout en autorisant l’observation de zones étendues ou en profondeur. Cependant, les caractéristiques géométriques de ces antennes et la nature des matériaux utilisés posent des difficultés importantes pour réaliser les opérations d’accord, d’adaptation, ou découplage mutuel lors de leur utilisation en IRM. Dans le cadre de cette thèse, nous avons conduit des travaux abordant ces différentes problématiques et permettant la mise œuvre d’un réseau supraconducteur pour l’imagerie haute résolution à champ clinique.Nous avons développé en premier lieu un système permettant de réaliser automatiquement et sans contact l’accord et l’adaptation des antennes miniatures monolithiques. Ce système utilise des techniques, basées sur le couplage électrique et magnétique, que nous avons préalablement étudiées en utilisant différentes méthodes. Les performances de ce système ont été étudiées et la faisabilité de sa mise en œuvre dans une expérience d’IRM a été établie.Nous avons abordé en second lieu le problème de découplage mutuel des éléments constituant un réseau. Pour cela, des techniques de découplage potentiellement compatibles avec les antennes miniatures supraconductrices ont été étudiées. En particulier, la technique de découplage par anneau de blindage a été investie en profondeur, par simulation numérique et expérimentalement, et nous avons développé un modèle analytique d’optimisation du niveau de découplage accessible par cette technique. Nous avons mis en œuvre et validé cette technique avec des réseaux en cuivre de quatre antennes et des premiers essais ont été conduits avec un réseau constitué de deux antennes miniatures supraconductrices. / In MRI microscopy, the sensitivity of the detection is a critical issue for acquiring images with high signal to noise ratio because the amount of NMR signal is extremely low. An alternative to the use of high field strength is to improve the performances of the radiofrequency coil that detect the NRM signal. In particular, the strategy underlying the present work aims at exploiting the high sensitivity of miniature superconducting coils based on the monolithic design of transmission line resonators, with the long term objective of implementing an array of these coils. The development of such array represents an important instrumental stake since it allows for benefiting from the intrinsically high sensitivity of miniature superconducting coils while allowing the observation of extended region of interest. However, the highly compact structure, the small size of the coil and the nature of the material used make rise important difficulties for achieving the tuning, matching and mutual decoupling when using these coils in MRI. In the frame of this PhD project, with conducted works to address these problematics and allow for implementing a superconducting array for high resolution imaging at clinical filed strength.We have firstly developed a control system which allows for automatic and contactless tuning and matching of miniature monolithic coils. This system uses techniques based on electric and magnetic coupling, that we beforehand investigated using various methods. The performances of this system were studied and the feasibility of implementing it in an MRI experiment was established.The second part of this work addresses the issue of mutual decoupling between the elements of an array elements. To this end, decoupling techniques being potentially compatible with miniature superconducting coils have been studied. In particular the decoupling technique using shielding rings has been deeply investigated, by numerical simulation and experimentally, and we have developed an analytical model for optimizing the decoupling level achievable with this technic. We have implemented and validated this technic with a four-element copper coil array and first trials were performed with an array of two miniature.superconducting coils.