While the quest for the cosmological parameters has met spectacular progress, understanding the processes regulating galaxy formation is still a major open issue in observational cosmology. We now have a fairly detailed census of the diverse galaxy populations in the distant Universe, hence the time is ripe for fundamental advances in understanding galaxy formation and evolution in the crucial first few billion years since the big bang. This requires going beyond simply counting and classifying galaxies, and to observationally constrain and clarify the physical processes that operated at those early epochs. I have contributed to this effort by pioneering a new selection technique that is allowing us to build representative, unbiased samples of galaxies in the redshift interval 1,4 < z < 2,5; i.e. at the peak of star-formation, mass assembly and AGN activity. By applying this "BzK technique" I have been leading research projects that have now provided major new results on high redshift z~2 galaxies. These include the widespread presence of previously unknown Compton-thick AGNs inside massive galaxies; the major phase of star formation at very high rates; and the existence of evolved galaxy clusters containing X-ray emitting gas already at z~2. Building on the legacy of these discoveries and critical results, I ask for support to fund the establishment of a new research team to lead research aimed at exploring thephysics of galaxy formation in the distant Universe. With three postdocs each year for a total of 5 years, we will pave new avenues towards understanding the relation between black holes and galaxies at the time of their major mass growth and assembly. In a full multiwavelength approach, by obtaining and using data from all major observational facilities (both in space and on the ground) we will aim to clarify the physical trigger of downsizing, catch AGN feedback in action and assess its role in galaxy transformations, along with the effects of the environment, gas accretion, star formation and merging in driving galaxy formation.

L'objectif du projet SURTRAIN est le développement d'une plate-forme de surveillance basée sur l'image et le son capable d'aider le travail des opérateurs chargés de la surveillance grâce à une continuité de service entre les équipements embarqués et les infrastructures au sol. Tout cela, de manière à améliorer la sécurité des passagers et la protection des infrastructures._x000D_ L'innovation principale du projet est le développement et la mise en œuvre conjointe d'algorithmes d'analyse d'image et du son en temps réel en environnement mobile et permettant la détection de situations à risque pour les passagers et la mise en œuvre rapide des mesures conservatoires par un opérateur présent dans la boucle et supervisant le système de surveillance._x000D_ La plate-forme embarquée présente une architecture à ressources distribuées entre les capteurs vidéo et audio, les unités d'analyse, celles de stockage et le serveur d'application. Elle s'appuie sur un réseau bord de communication haut débit adapté au contrainte de l'environnement. Une passerelle de radiocommunication Bord/Sol est mise en œuvre permettant notamment un travail coopératif entre les ressources embarquées et celles en gare.

In cosmology, the growth of gravitational instability in a dark matter fluid is not fully understood yet. However, with the advent of a new generation of observational cosmology projects - let us mention ground-based or space large-scale cosmic shear surveys or the Square-Kilometer-Array - it is necessary to be able to compute with a great precision the density fluctuation spectra as they develop in the matter fluid. N-body simulations can provide us with some answers but only for a limited number of cosmological parameters. The scientific objectives of those projects, and among them the determination of the dark energy property, make however largely use of our ability to perform such calculations for a large variety of models. It is thus necessary to refine our theoretical knowledge of the behavior of the gravitational instabilities. The aims of this project is to develop computation tools of the density spectrum, its bispectrum (three-point correlation function) for a set a various cosmological models. More precisely we wish - to have at our disposal tools for computing those quantities in a manner that would be fully analytical and predictive in the quasilinear regime (e.g. in a regime that extends the purely linear regime in a way that remains to be defined); - to develop models allowing the description of the matter-halo relations or to tackle the nonlinear regime with well controlled approximations. The fulfillment of the first item of this scientific program takes advantage of recent resummation techniques (initially developed by R. Scoccimarro and his collaborators). They allow the derivation of spectra, and in principle correlation functions of any order, in a fully predictive way without introducing divergences. Those analytical results have still to be extended to various cosmologies and to quantities beyond spectra and tested against numerical simulations. The second item of our program is more speculative. There are however favorable indications.

Bien que des progrès considérables aient été réalisés depuis 10 ans dans l'analyse - des macromolécules, peu d'informations structurales peuvent être aujourd'hui obtenues pour - les systèmes biologiques multimériques alors qu'ils sont le siège de nombreux processus - importants (la RMN en phase liquide est limitée aux complexes de poids inférieurs à 35 kDa). - De plus, les systèmes qui s'agrègent ou de faible solubilité dans l'eau ne sont pas facilement - analysables par RMN ou par diffraction des rayons X. De telles limitations empêchent l'étude - de problèmes biologiques importants comme : - (i) La détermination de la structure bioactive de petites molécules en interaction - avec une macromolécule insoluble. - (ii) Les changements de conformation de protéines induits lors de la formation - d'architectures multimoléculaires. On dispose maintenant d'indications - claires montrant que certaines protéines subissent des réorganisations - structurales in vivo afin d'acquérir leurs propriétés fonctionnelles. - (iii) Les modifications structurales qui ont lieu dans les régions d'interface des - cellules (par exemple proche de la membrane cellulaire). Il est maintenant - bien établi que l'interaction des peptides avec les membranes est importante - dans de nombreux processus biologiques et peut s'accompagner de - changements structuraux pour lesquels on ne dispose que de peu - d'information. - Dans ces trois exemples les changements conformationnels peuvent être suivis par des - mesures de variation de distance entre atomes (rapporteurs). Pour de telles études la RMN du - solide est devenue un outil indispensable. Aujourd'hui, les plus grandes distances que l'on - peut mesurer, à l'aide de marquage 13C-15N et 13C-13C, sont de l'ordre de 5-6 Å. Pour accéder - à des distances supérieures une stratégie efficace consiste à changer la nature des atomes - rapporteurs, par exemple en utilisant 1H ou 19F. Cependant l'utilisation de 1H nécessite de - marquer par du deutérium l'ensemble du milieu biologique (protéine, lipides...) à l'exception - des deux positions dont on veut mesurer la distance. L'obtention d'un tel milieu marqué est - rarement possible. L'utilisation du 19F n'est pas meilleure puisque cet atome peut modifier - significativement le système à étudier. - Dans le cadre de ce projet, nous envisageons de développer une approche totalement - nouvelle qui permettra de mesurer des distances jusqu'à 16 Å : la RMN du tritium en milieu - solide. Cette approche sera ensuite appliquée à deux problèmes biologiques importants : - - la détermination de la conformation bioactive d'une molécule anticancéreuse, le - Taxotère, lorsqu'elle interagit avec la tubuline. - - L'étude de H-NS, une petite protéine bactérienne liant l'ADN et modulant - l'expression de nombreux gènes. H-NS fonctionne en polymérisant sur l'ADN pour - conduire à un complexe de structure inconnue. Notre approche servira à préciser - l'organisation du complexe et à établir un lien entre la structure et les différentes - fonctions de H-NS. - La RMN du tritium en milieu solide permettra d'obtenir des informations structurales - importantes sur des assemblages moléculaires complexes qui ne peuvent que difficilement - être étudiés par d'autres méthodes. Elle devrait devenir rapidement un outil de choix pour - l'étude de nombreux processus biologiques. -

Organic coatings on surfaces are present everywhere in today’s life, from paints used in car industry to medical devices. Among the different coating techniques, electrografting which is based on the electro-induced polymerization of dissolved electro-active monomers on metallic or semiconducting surfaces leads to coatings strongly attached to the surface. Itis a powerful and versatile technique which finds applications in various fields including biocompatibility, corrosion, lubrication, soldering, and adhesion. These last decades, due to an increasing technological demand for miniaturization, there has been a growing interest for modified surfaces containing functionalities with micrometer or nanometer spatial definitions. Scanning Electrochemical Microscopy (SECM), often used as an analytical technique, has also been shown to be an interesting tool for local surface modification. In particular, using SECM, localized electrografting of vinylic monomers and diazonium salts on various substrates has been demonstrated. Unfortunately, the specific conditions that are necessary for the process to occur with vinylic monomers are extreme in comparison to conventional surface transformation with SECM. Among others, combination of high current densities and high electric potentials generates bubbles, leading to turbulences and temporary insulating layers at the electrode-solution interface. In addition both the local electric potential and the chemical content of the solution underneath the probe are observed to play a role in the transformation. Up to now, the contribution of each is still not well identified, making a complete control of the procedure still out of reach.With the present project, we propose to face the challenge of understanding the local electrografting situation arising in the case of a highly polarized microelectrode moved closer to a substrate. Combining theoretical and experimental work, and starting from very simple situations to increase the complexity gradually, we propose to construct a realistic theoretical model for high voltage localized electrografting by SECM. It will be made extensive use of numerical simulations with Comsol Multiphysics, a commercially available user friendly software, especially devoted to multiphysics coupling. The work will be divided into five main steps. First, a suitable model for the conventional SECM direct mode of surface transformation under mild conditions will be provided. Secondly, the complex chemical reactivity associated with the presence of diazonium salts and vinylic monomers in solution will be studied. In a third step, the specific impact on mass transport of bubbles formed in the vicinity of a microelectrode will be investigated. Models and results developed in those three preliminary steps will then be brought together to propose a general model for high voltage local electrografting, taking into account the entire complexity of the problem. After a necessary step of validation by confrontation against experiment, the model proposed will be exploited for several purposes, some more technical some more fundamental. Among other, it will represent a convenient framework to optimize the experimental protocol for the synthesis of diazonium salts and vinylic polymers grafted films whose experimental feasibility by localized electrografting has already been established in the literature. But it will also be exploited to investigate the feasibility of new coatings. It is also expected that the model can be re-invested in more general situations than localized electrografting, for example in the context of corrosion.