Les superalliages monocristallins à base de nickel sont utilisés en aéronautique pour les aubes de turbines. Cette étude est consacrée à la modélisation du comportement en fluage du superalliage monocristallin AM1 après mise en radeaux, au cours d’essais isothermes à contrainte variable. Des diffractogrammes (200) ont été obtenus in situ par diffractométrie trois axes en rayonnement synchrotron, à haute température (950-1150°C) pour des paliers de contrainte entre 0 et 300MPa. Pour chaque phase, les déformations élastiques se déduisent de la position des pics et les contraintes, déformations plastiques et vitesses de déformation sont obtenues par la mesure du désaccord paramétrique, en utilisant un modèle composite en série. Ces résultats sont combinés à une caractérisation post mortem en microscopie électronique : morphologie des phases, densité et nature des dislocations. La mesure in situ du désaccord paramétrique donne accès à la densité instantanée de dislocations aux interfaces y/y’. Dans la phase ylors d’incréments de la contrainte appliquée, la contrainte de Von Mises augmente, puis se relaxe jusqu’à une contrainte seuil. Cette contrainte est en accord avec la contrainte d’Orowan et les largeurs des couloirs mesurées post mortem. La déformation plastique de la phase y’ est produite par montée de dislocations de vecteur de Burgers perpendiculaire à l’axe de traction sous l’action de la seule contrainte transverse et contrôlée par l’entrée de dislocations depuis les interfaces. Une simulation des pics de diffraction permet de reproduire l’évolution de leur largeur en fonction de la nature et de la répartition des dislocations aux interfaces et dans la phase y' / Nickel-based superalloys are used in aeronautics for turbine blades. This study aims at modelling the creep behaviour of single-crystalline AM1 superalloy with a rafted γ/γ’ microstructure during isothermal tests under variable applied stresses. (200) diffraction profiles are obtained with a triple crystal diffractometer and high energy synchrotron radiation at high temperature (950-1150°C) with an applied stress varying between 0 and 300 MPa. For each phase, the elastic strain is deduced from the peaks’ positions and the stress, plastic strain rate from the lattice mismatch, assuming a model lamellar composite material. Post mortem characterizations by electronic microscopy completes the results: morphology of each phase, dislocations densities and nature. The measurement of lattice mismatch leads to an in situ estimation of the dislocations’ density at the γ/γ’ interfaces. For the γ phase, during successive jumps of the applied stress, the Von Mises stress increases and then relaxes down to a threshold stress. This stress is in agreement with Orowan stress deduced from the post mortem measurements of the γ channels’ width. Plastic strain of the γ' phase is produced by climb of dislocations with Burgers’ vectors perpendicular to the tensile axis under the mere transversal stress and is controlled by the entrance of dislocations into the rafts from the interfaces. The distribution of elastic strains was simulated by assuming two main contributions: dislocations at the γ/γ’ interfaces and within the γ’ rafts. The simulation reproduces the absolute magnitude of the peaks’ width, as well as their increase with dislocation densities

Silicon technologyhas been seekingfor a monolithic solution for a chip where data processing and data communication is performed in the CMOS part and the photonic component, respectively. Traditionally, silicon has been widely considered for electronic applications but not for photonic applications due to its indirect bandgap nature. However, band structure engineering and manipulation through alloying Si with Ge and Sn has opened new possibilities. Theoretical calculations show that it is possible to achieve direct transitions from Ge ifit is alloyed with Sn. Therefore, a GeSn system is a choice to get a direct bandgap. Extending to ternary GeSnSi and quaternary GeSnSiCstructures grown on Si wafers not only makes the bandgap engineering possible but also allowsgrowing lattice matched systems with different strain and bandgaps located in the infrared region. Different heterostructures can be designed and fabricated for detecting lightas photonic sensing oremitting the light as lasers. Alloying not only makes engineering possible but it also induces strain which plays an important role for electronic applications. Theoretical and experimental works show that tensile strain could increase the mobility, which is promising for electronic devices where high mobility channels for high performance MOSFETs is needed to speed up the switching rate. On the other hand, high n-doping in tensile strains in p-i-n structures makesΓ band transitions most probable which is promising for detection and emission of the light. As another benefit of tensile strain, the direct bandgap part shrinks faster than the indirect one if the strain amount is increased. To get both electronic and photonic applications of GeSn-based structures, two heterostructures (Ge/GeSn(Si)/GeSi/Ge/Si and Ge/GeSn/Si systems), including relaxed and compressive strained layers used to produce tensile strained layers, were designed in this thesis. The low temperature growth is of key importance in this work because the synthesis of GeSn-based hetrostructures on Si wafers requires low thermal conditions due tothe large lattice mismatch which makes them metastable. RPCVD was used to synthesize theseheterostructures because not only it offers a low temperature growth but also because it is compatible with CMOS technology. For utilization of these structures in devices, n-type and p-type doping of relaxed and compressive strained layers were developed. HRRLMs, HRTEM, RBS, SIMS, and FPP techniques were employed to evaluatestrain, quality, Sn content and composition profile of the heterostructures. The application of GeSn-based heterostructures is not restricted to electronics and photonics. Another application investigated in this work is photovoltaics. In competition with Si-based solar cells, which have, or areexpected to have,high stability and efficiency, thirdgeneration solar cells offer the use of low cost materials and production and can therefore be an alternative for future light energy conversion technology. Particularly, quantum dot sensitized solar cells are associated with favorable properties such as high extrinsic coefficients, size dependent bandgaps and multiple exciton generation and with a theoretical efficiencyof 44%. In this work, two categories of QDs, Cd-free and Cd-based QDs were employed as sensitizers in quantum dot sensitized solar cells (QDSSCs). Cd-based QDs have attracted large interest due to high quantum yield,however, toxicityremains still totheir disadvantage. Mn doping as a bandgap engineering tool for Cd-based type IIZnSe/CdS QDs wasemployed to boostthe solar cell efficiency. Theoretical and experimental investigations show that compared to single coreQDSSCs,typeII core-shells offer higher electron-hole separation due to efficient band alignment where the photogenerated electrons and holes are located in the conduction band of the shell and valence band of the core, respectively. This electron-hole separation suppresses recombination and by carefully designing the band alignment in the deviceit can increase the electron injection and consequently the power conversion efficiency of the device. Considering eco-friendly and commercialization aspects, three different “green” colloidal nanostructures having special band alignments, which are compatible for sensitized solar cells, were designed and fabricated by the hot injection method. Cu2GeS3-InP QDs not only can harvest light energy up to the infraredregion but can also be usedastypeII QDs. ZnS-coating was employed as a strategy to passivate the surface of InP QDs from interaction with air and electrolyte. In addition, ZnS-coating and hybrid passivation was applied for CuInS2QDs to eliminate surface traps. / <p>QC 20151125</p>