Au CEA sont actuellement réalisées des études de conception pour un démonstrateur de SFR, le réacteur ASTRID (Advanced Sodium Technological Reactor for Industrial Demonstration). Ce travail implique de développer et valider des outils de calcul scientifique afin de créer un dossier de sûreté à transmettre à l’ASN. Notamment, l’utilisation de codes neutroniques doit permettre de calculer les caractéristiques de cœurs de réacteur avec des précisions bien maitrisées. Les données nucléaires, qui sont les paramètres d’entrée de ces codes, constituent la principale source d'incertitude dans ces calculs. Le but de cette thèse est de réduire les incertitudes dues aux données nucléaires et donc de mieux prédire les caractéristiques du cœur d’ASTRID en utilisant l’assimilation de données intégrales. Basée sur l'inférence bayésienne-laplace appliquée sur des valeurs «a priori» (bibliothèque JEFF-3.1.1 et matrices COMAC), cette méthode consiste à mettre à jour nos connaissances sur les données nucléaires par ajustement de leurs valeurs centrales et incertitudes associées en utilisant des mesures intégrales. Les résultats de ce travail ont été utilisés pour quantifier les biais et les incertitudes réduites associées aux caractéristiques du cœur d'ASTRID (masse critique, coefficient de vide et de Doppler, antiréactivité des barres de contrôle ...). / The design of an advanced SFR demonstrator, the ASTRID reactor (Advanced Sodium Technological Reactor for Industrial Demonstration) at CEA implies the development and validation of scientific calculation tools, in order to create a safety dossier. Notably, the use of neutronic codes aims at defining the characteristics of reactor cores with well-mastered accuracies. Nuclear data, the input parameters of these codes, constitute the main source of uncertainty in neutronic calculations. The purpose of this PhD is to reduce uncertainties associated to nuclear data, and hence better predict the characteristics of the ASTRID core, using Integral Data Assimilation. This method, based on Bayesian-Laplace Inference, consists in using integral data C/E (calculation-to-experiment ratio) to perform adjustments on the central value and uncertainties of nuclear data. The modifications on nuclear data suggested by assimilation results have been used to quantify the bias and the reduced uncertainties associated to the ASTRID core main characteristics.

Dans cette thèse, j'adresse le sujet fascinant et attirant du transport de charge électrique et de chaleur dans les systèmes Hall quantiques, qui sont parmi l'exemple le plus célèbre des phases topologiques de la matière, en présence de potentiels électriques dépendantes du temps. L'effet Hall se produit dans des systèmes électroniques bidimensionnels dans la limite de forts champs magnétiques perpendiculaires. Le cachet de systèmes de Hall quantiques est l'apparition d'états de bord métalliques unidimensionnels sur les frontières du système.La longueur de cohérence assurée par la protection topologique garantit d’avoir accès à la nature ondulatoire des électrons. Ces propriétés ont inspiré un nouveau domaine de la recherche, connu comme la l'optique quantique électronique. Une source d’électrons individuels peut être réalisée en s'appliquant à un système de Hall quantique impulsions Lorentzian. En considérant l'application d'un train périodique d'impulsions Lorentzian à un système Hall quantique, j'examine la densité de charge d'un état composé par beaucoup de levitons dans le régime de Hall quantique fractionnaire, constatant ainsi qu'il est réarrangé dans une configuration réguliere de sommets et des vallées. Alors, j'analyse les propriétés de transport de chaleur des levitons dans les systèmes Hall quantiques, qui représente un nouveau point de vue sur l'optique quantique électronique, étendant et généralisant les résultats obtenus dans le transport de charge. / In this thesis, I address the intriguing and appealing topic of charge and heat transport in quantum Hall systems, which are among the most famous example of topological phases of matter, in presence of external time-dependent voltages. Quantum Hall effect occurs in two-dimensional electron systems in the limit of strong perpendicular magnetic fields. The hallmark of quantum Hall systems is the emergence of one-dimensional metallic edge states on the boundary. Along these edge states particles propagate with a definite direction. The coherence length ensured by topological protection guarantees to access wave-like nature of electrons. This properties inspired a new field of research, known as electron quantum optic. Single-electron source can be realized by applying to a quantum Hall system a periodic train of Lorentzian-shaped pulses.Plateaus of the Hall resistance appear also at fractional values of the resistance quantum. The physical explanation of fractional quantum Hall effect cannot neglect the correlation between electrons and this phase of matter is inherently strongly-correlated. By considering the application of a periodic train of Lorentzian pulses to a quantum Hall system, I investigate the charge density of a state composed by many levitons in the fractional quantum Hall regime, thus finding that it is re-arranged into a regular pattern of peaks and valleys, reminiscent of Wigner crystallization in strongly-interacting electronic systems. Then, I analyze heat transport properties of levitons in quantum Hall systems, which represent a new point of view on electron quantum optics, extending and generalizing the results obtained in the charge domain.

Ring Diagram Analysis (RDA) is one of the cornerstones of local helioseismology. The technique is used to observe f - and p-mode oscillations generated in the solar interior. RDA has been used to study plasma flows inside the solar convection zone, and the subphotospheric structure of active regions. A limitation of RDA, however, is the temporal resolution at which flow maps can be produced. This work presents the Temporal Ring Diagram Analysis (TRDA) data analysis pipeline as a novel addition to RDA. TRDA employs a sliding window technique to vastly increase the temporal resolution from 512 to 10 minutes. Sun-quakes are seismic waves triggered by solar flares and CMEs. They appear suddenly, penetrate the solar interior, and can last for up to an hour before dissipating. Until now, it has not been possible to analyse sub-photospheric plasma flows during a quake event. TRDA is used to study these flows for two Sun-quake events. One- and two-dimensional flow maps have been created for the quake of 15th January 2005. Three-dimensional flow velocities for a region of quiet-Sun and a second quake, which occurred on 14th August 2004. Quiet-Sun flows are typically 50-70 m s−1, while up-flows under the active region reach range between 120 and 150 m s−1. Also, low-speed flow flow is observed in the north-eastern corner of the region. Flow vorticity maps are also presented for further analysis. A novel data processing pipeline for Hinode Extreme Ultraviolet Imaging Spectrometer (EIS) data is also presented. The EIS Multiple Wavelength Analysis (EMWA) pipeline has been designed to simultaneously process numerous spectral wavelengths, and arrange the results in ascending order of ion formation temperature. EMWA has been used to analyse atmospheric plasma flows surrounding the foot-point regions of magnetic loops. The flow results for three active regions are presented. They show that strong outflows in the close to the foot-point regions have velocities 10 km s−1, and these flows expand with increasing temperature. Further analysis has revealed a low velocity but continuous supply of plasma is required to sustain these flows.

A variety of optical and electrical properties of ZnSiP2 have been studied. Single crystals of this material were grown from solution in zinc or tin, using an accelerated crucible rotation technique. Its lattice parameters have been measured between room temperature and 1100 C and the anisotropy of the thermal expansion coefficients interpreted in terms of the bonding of the crystal. The absorption edges of the best crystals showed well defined structure due to pseudodirect transitions between the three crystal field and spin-orbit split valence bands and the lowest conduction band minimum. The lowest band gap, after taking account of the exciton binding energy, was found to be 2.082 eV at room temperature. The valence band splittings and the relative strengths of the transitions were interpreted in terms of the quasicubic model. The absorption due to the two strongest excitons was fitted to a simple model, and the exciton binding energy estimated to be 22 meV. From the magnitude of the absorption coefficient, the matrix element for pseudodirect transitions was estimated to be about a thousand times weaker than that for direct transitions in the III V compounds. An absorption peak found only in low resistivity n-type material in the near infra-red was attributed to transitions between the Gamma3 and Gamma2 conduction bands (x 1 - x3 in zinc blende) at about 0.7 eV. The electrical properties of the crystals were measured between liquid nitrogen temperature and 200 C. Crystals grown from Sn, and some of those from Zn were fairly low resistivity n-type ( .5 - 20 ohm cm) while others grown from Zn were high resistivity (~108 ohm cm) p-type. Electron mobilities up to 160 cm 2 V-1 S-1, and hole mobilities up to 22 cm2 V-1 S-1 were measured. The dominant scattering mechanism was probably due to ionised impurities. Measurements of photoconductivity and photovoltaic effect at a rectifying metal contact had spectral responses substantially in agreement with the absorption measurements, although the structure due to pseudo-direct transitions was not resolved. Strong trapping of excess carriers was observed even in the best optical material.

We study the spatial modulation of the wave function in bilayer and trilayer graphene systems originating from two underlying mechanisms: quantum interference phenomena (QIP) and quantum confinement. We also take a bottom-up approach to tailoring surface potential distributions at the atomic scale to influence/control electron behaviour, by utilising the interaction between graphene layers and nanostructured, atomically flat insulating ionic surfaces. Quantum interference phenomena were explored at bilayer-trilayer armchair interfaces in multilayer graphene with various stacking orders by using scanning tunnelling microscopy and with support from theoretical simulations. Effects of various types of edges, which terminate the stacks abruptly or appear at lateral interfaces within the multistack, were revealed and correlated with scattering mechanisms, while a taxonomy of interference patterns was established based on stacking order. The effect of extra sources of scattering was also studied to understand the origin of the well-known (√3×√3)퐑퐑 30° superstructure in graphene systems, and a new explanation was proposed based on decomposing defects into armchair contours, able to provide multiple sources of scattering. The energy dependency of the (√3×√3)퐑퐑 30° superstructure and its motifs was quantitatively explored in bilayer graphene. Finally, bilayer and trilayer graphene were overlaid on atomically flat insulating surfaces decorated with nanostructures such as step edges and closed contours, and able to induce sizeable local electrostatic potential distribution within the graphene overlayers. A well-defined, rectangular subsurface potential distribution, akin to a nanoscale quantum box applied to a physically unconfined graphene overlayer, produced state localization in the local density of states of a trilayer, while resonances were also observed in bilayer graphene around irregular subsurface features within the ionic substrate. A 1D periodic superstructure also emerged from the interaction of bilayer graphene with large area flat regions of these ionic substrates with square symmetry.

No description available.

No description available.

Computer simulations are invaluable for the study of ultrafast phenomena, as it is not possible to directly access the electronic and nuclear dynamics in experiments. We present an efficient method for simulating the time-dependent coupled electron-ion dynamics within the Ehrenfest picture in molecules under the influence of time-dependent electric fields, based on an extension of the density-functional tight binding model. We consider self-consistency in a self-multipole-consistent framework, expanding the electron density in terms of radial Gaussian and angular real cubic harmonic functions. This enables the efficient computation of the electrostatic interaction energy while retaining a physical description of charge transfer and ionic polarisation. We show that this Gaussian tight binding method produces molecular polarisabilities, time-dependent dipole moments, and electron densities in strong agreement with density-functional theory, but at a small fraction of the cost. This efficiency enables high-throughput ultrafast studies on molecules, which we demonstrate on the example of transient core-spectroscopy on polythiophene fragments.

All physical quantum systems are in contact with the external world which is inevitably a source of noise. It is therefore necessary to take into account this dissipation when designing controls that accomplish useful tasks in quantum information processing. This thesis is on that overlap of open quantum systems and control theory; it looks at what dynamics can happen with the use of external driving, and how this driving can be chosen to accomplish a desired goal. The first result looks at how a given quantum dissipator can be manipulated, using coherent controls, into replicating the action of a different type of noise. This results in no-go theorems for how noise can be transformed based on its isotropy, with applications in simulating open systems. Another way of doing simulations is to use only unitary dynamics over the system and a finite dimensional ancilla, and it is proved that there always exists a dilation Hamiltonian that replicates the noisy dynamics continuously in time. This also highlights the fact that adding controls on noise can result in a different evolution than if the controls were done on the under- lying system-environment level. A conjecture is introduced and studied which states that both approaches are equivalent in the special case of the controls commuting with the Lindbladian. A way around this difficulty is to use a quantum system to compute in situ which controls are best for achieving a desired task on itself. This problem is studied in the context of reaching entangling gates on a quantum simulator in order to upgrade it into a quantum computer. The experimental cost of doing so is found to be polynomial in the number of qubits in simulations. The same underlying principle is also used to find error correcting codes tailored to the dissipation in a system.

The discovery of neutrino oscillations implies that neutrinos are massive, and therefore is a concrete evidence that the standard model (SM), which forbids the mass of neutrinos, is not complete. As a consequence, completing the knowledge of neutrino oscillations extends our understanding of new physics. We are entering the age of precision measurement of neutrino oscillations, with the preparation for the upcoming Long Baseline experiments (LBL) — Deep Underground Neutrino Experiment (DUNE) and Tokai to Hyper-Kamiokande (T2HK). In this thesis, we firstly study how DUNE, T2HK and the combination solve the remaining problems of the standard neutrino oscillation — octant and mass ordering degeneracy problems, if CP violates, and what the value of CP phase δ is. In the following, we study how Littlest Seesaw Models (LS) can be tested by DUNE, T2HK together with short- and medium-baseline reactor experiments, after fitting these models with the current global results. In the next half of this thesis, we extend our discussion to allow external interactions — nonstandard interactions (NSIs) in matter for DUNE. After reviewing current studies on the precision of NSI-parameter measurement, we discuss the exclusion ability of DUNE to the SM prediction over the possible scenarios. Considering NSIs are flavour-dependent, we demonstrate the possible correlations between or among NSI effects under flavour symmetries A4 and Z2. Based on these correlations, we present how DUNE can test flavour symmetries A4 and Z2 through NSIs. Our results show the experimental properties of DUNE and T2HK, and how they perform for the theory of flavour symmetry.