The bombardment of surfaces by ions can lead to the spontaneous formation of nano-structures. Depending on the irradiation conditions, smoothening or roughening mechanisms can be the leading order in pattern formation which can result in the creation of dots, ripples or ultra-smoothening effects. Because ion bombardment is already ubiquitous in industrial settings, and is relatively inexpensive compared to other surface processing techniques, self-organized patterning by ion bombardment could enable a simple, economical means of inducing well-defined nanoscale structures in a variety of settings. Understanding the fundamental behavior of surfaces during ion bombardment is therefore a vital goal; however, a complete understanding of physical processes governing surface pattern formation has not been reached yet. In order to address this issue, my thesis research has utilized three primary approaches. First, I have done real-time non-coherent X-ray scattering experiments at Cornell High Energy Synchrotron Source (CHESS) for studying kinetics of structure formation of Silicon undergoing Ar⁺ bombardment over a range of wavenumbers 4-5 times larger than has previously been obtained. From our data, we were able to extract values of the angle-dependent thickness of the amorphous layer that forms under ion bombardment, the ion-enhanced fluidity within that film, the magnitude of the stress being generated by the ion beam, and the strength of prompt atomic displacement mechanisms. Second, to further deepen our knowledge of surface dynamics, I have performed coherent X-ray studies of Ar⁺ bombardment of SiO₂ at the Advanced Photon Source (APS) for investigating the dynamics more profoundly than can be done with traditional time-resolved experiments. When using a focused ion beam, an inhomogeneous ripple motion was generated, this phenomenon reflected as an oscillatory behavior in the two-time and corresponding g₂(t) correlation functions. By fitting the oscillations in the g₂(t) correlation function, we have determined the surface ripple velocity on SiO₂ driven by Ar⁺ sputter erosion. Finally, to support the results of coherent X-ray experiments, simulations of growth models such as linear Kuramoto-Sivashinsky (KS) and Kardar-Parisi-Zhang (KPZ) have been carried out in order to compare the simulated temporal correlation functions of the scattered intensity with those obtained from the coherent x-ray scattering experiments.

Nous présentons dans ce manuscrit le travail visant à faire croitre des nanofils de silicium dans une matrice d'alumine nanoporeuse par la méthode VLS.<br />Nous avons développé un procédé de réalisation d'alumine nanoporeuse sur silicium. Celui-ci permet maintenant de contrôler les propriétés géométriques de l'alumine fabriquée (diamètre des pores, période, épaisseur de la couche). L'organisation latérale est très bonne même s'il est envisageable de l'améliorer.<br />Nous avons aussi étudié différentes méthodes pour déposer de l'or au fond des pores ainsi obtenus. Les avantages et les inconvénients de chacune des méthodes ont été détaillés et nous avons pu ainsi répondre au besoin de départ (l'obtention de catalyseurs d'or). Ces méthodes de dépôt sont utilisables sur substrat massif et présentent des avantages intéressants en termes de souplesse et de simplicité d'utilisation notamment.<br />Enfin, nous avons fait croitre avec succès des nanofils sur substrat massif et dans les pores de l'alumine. Nous avons pu en étudier quelques caractéristiques, notamment structurales. L'utilisation de l'alumine pour positionner les nanofils est donc envisageable et l'effet de guidage des nanofils par les pores pendant la croissance a été démontré. De plus, cet environnement confiné ne fait pas augmenter significativement le nombre de défauts structuraux observés dans les nanofils comparé à des nanofils crûs sur substrat massif. Nous avons réalisé quelques mesures préliminaires des tendances que suit la conduction électrique dans les nanofils. La prise de contacts s'est révélée très facile grâce à la géométrie d'encapsulation des nanofils dans les pores.

La diffusion centrale des rayons X en incidence rasante a été appliquée à l'étude in situ de la croissance de nano-objets supportés. En s'appuyant sur une analyse quantitative des mesures expérimentales, la forme, la taille, l'orientation et la distance entre les nanostructures ont été déterminées. Nous avons appliqué cette technique à la caractérisation de la croissance d'agrégats de palladium sur MgO(001). La confrontation des résultats obtenus avec ceux déduits des clichés de microscopie électronique à transmission a permis de valider l'analyse. Plus particulièrement l'énergie d'interface entre les agrégats de palladium et MgO(001) a été déterminée. D'autre part nous avons étudié la croissance auto-organisée de plots de cobalt sur différents substrats nanostructurés. Tout d'abord sur un système modèle, la surface d'Au(111), qui démontre la sensibilité de cette technique aux tous premier stades de l'organisation. Puis sur une surface d'Au(677) crantée et enfin sur un film mince d'argent sur MgO(001) nanostructuré par un réseau de dislocations enterré. Dans chaque cas, la morphologie des plots et le degré d'ordre à longue distance ont été obtenus. Les interférences entre les ondes diffusées par la structuration de la surface et celles diffusées par les plots de cobalt mettent en évidence le phénomène d'auto-organisation et permettent de localiser le site de nucléation-croissance.

I started research using synchrotron radiation (at LURE-Orsay and SRS-Daresbury) during my PhD-Thesis, between 1985 and 1988, mostly using X-Ray Absorption Fine Structure (XAS-EXAFS) and Diffuse Scattering, to investigate the structural properties of alloys, in parallel with laboratory physical metallurgy studies2. I next spent one and a half year as a postdoc at Bell-Labs, where I learnt the bases of Grazing Incidence X-Rays Scattering/Diffraction (GIXS/GIXD), both at NSLS (Brokhaven)3,4 and PEP (Stanford)5. I was employed at the CEA in 1990 to develop synchrotron radiations studies of surfaces and thin films. Between 1990 and 1992, I first built a beamline (in collaboration) and a surface diffractometer6 at LURE, on the supraconducting wiggler, in which I performed first surface diffraction (SRXD) experiments. At that time, I choose to investigate the atomic structure of metal-oxide surfaces by SXRD7,8, as well as the in situ growth of metal on oxide by GIXS and GIXD9, performing also the first in situ Grazing Incidence Small Angle X-Ray Scattering (GISAXS) measurements with a punctual detector. Then I have been quickly involved in the development of the French Interface CRF/BM32 beamline at the ESRF, where I co-developed (with Robert Baudoing-Savois) a new surface diffractometer10, associated with a large UHV chamber equipped with Molecular Beam Epitaxy (MBE). This instrument went in operation in 1996. I continued the study of oxide surfaces11 and metal-oxide interfaces12-15, there and on different ESRF beamlines, combining SXRD, GIXS, GIXD and 0D-GISAXS. This was a fast developing and very gratifying period, which was partly summarized in 1998 in a review paper13. In 1999, I developed a complete instrument on the ID32 ESRF beamline to perform GISAXS using a 2-dimensional detector, in UHV, in situ during growth, without any window before the sample, the UHV chamber being directly hooked to the sychrotron ring16,17. This yielded GISAXS measurements of exceptional quality which opened a new area of diverse studies, during which I progressively shifted from metal on oxide growth to the organized growth of metal dots on different patterned substrates1,16,18-23. Around 2001, I developed in situ GISAXS, combined with SXRD/GIXS/GIXD, on the MBE X-ray chamber of the BM32 beamline. Between 2000 and 2005, both instruments have been intensively exploited to investigate growing metal on oxides and (nanostructured) metal surfaces, together with first studies of germanium growth on nanostructured Si surfaces. I had the chance to collaborate with Remi Lazzari, a postdoc, Frederic Leroy, a PhD-student, and Christine Revenant, forming a team who developed the quantitative analysis of our GISAXS data, trying to go beyond simple analyses, and to understand in detail the diffuse scattering in GISAXS16,18,19,21,24. In parallel, thanks to the venue in our laboratory of specialists of anomalous scattering (Vincent Favre-Nicolin, Hubert Renevier, Johann Coraux), we developed the technique of Multiple Anomalous GISAXS (MAD-GISAXS)22. During all these years, we performed GISAXS measurements during growth, probing the morphology, in parallel with scattering measurements (SXRD, GIXS, GIXD), sometimes 2 anomalous, to probe the structure and composition of growing islands. Because covering all these results would be too wide a field , I have chosen in the first part of this manuscript to concentrate on the GISAXS part of our studies, corresponding to about 6 years of work, between 1999 and 2005. This work also led to the writing of an invited review25. Since then, my research has evolved toward in situ studies with all these x-ray techniques, of the growth of semiconductor islands (Ge on Si) and semiconductor surface structure26, with a very recent extension of our MBE system to perform Chemical beam Epitaxy (CBE or UHV-CVD, Chemical Beam Epitaxy) in parallel with MBE. One of the objective of this combined MBE/CVD system is to allow for in situ structural investigations of growing semiconductor nanowires, which we already started by studying the fusion/solidification processes of AuSi eutectic catalysers27. This last work is the object of a very brief perspective chapter at the end of this manuscript. During all my research work, I have tried to develop and use synchrotron radiation scattering techniques to get the most accurate structural or morphological information during growth, with the goal of going beyond technical studies, to try to understand the basic mechanisms of the processes involved, such as the 3D growth of metal on oxide, or the intermixing between the Si substrate and the Ge deposit. I hope that this permanent concern is also revealed in this manuscript.

Ce manuscrit se concentre sur l'analyse du GISAXS d'îlots sur un substrat. Les données GISAXS doivent être analysées de façon quantitative afin d'obtenir des paramètres morphologiques précis (courbes de croissance, forme d'équilibre de l'îlot et énergie interfaciale) pour le processus d'élaboration. L'accent est mis sur le facteur de forme de l'îlot, c'est-à-dire la transformée de Fourier de la forme de l'îlot. On montre que la forme de l'îlot et la taille peuvent être obtenues à partir de la symétrie de l'îlot, la présence de facettes de l''îlot, le comportement asymptotique loin dans l'espace réciproque pour une grande polydispersité et les zéros ou les minima de l'intensité pour une faible polydispersité. Une comparaison approfondie entre l'approximation de Born et l'approximation plus précise de l'onde distordue (DWBA) met en évidence la spécificité apportée par la géométrie en incidence rasante. L'analyse quantitative est illustrée pour des images GISAXS acquises in situ pendant l'épitaxie par jet moléculaire de nano-îlots Ag ou Pd sur MgO(001) pour différentes épaisseurs et températures. Les paramètres morphologiques obtenus sont en très bon accord avec des résultats de microscopie électronique à transmission. Finalement, la diffusion incohérente a été mise en évidence en GISAXS et a pour origine des corrélations entre les îlots.

In this work, the crystallographic structure and its changes under thermal treatment of different systems consisting of metal oxide nanoparticles is studied. The principal method used throughout the thesis is x-ray powder diffraction enriched with grazing incidence small angle x-ray scattering when the nanoparticles form an ordered structure or with x-ray absorption spectroscopy when additional information on local crystallographic structure is required. For all the systems the preparation conditions were optimized according to the crystallographic data for further applications.

The development of InGaN quantum dots (QDs) is both scientifically challenging and promising for applications in visible spectrum LEDs, lasers, detectors, electroabsorption modulators and photovoltaics. Such QDs are typically grown using the Stranski-Krastanov (SK) growth mode, in which accumulated in-plane compressive strain induces a transition from 2D to 3D growth. This method has a number of inherent limitations, including the unavoidable formation of a 2D wetting layer and the difficulty of controlling the composition, areal density, and size of the dots. In this research, I have developed InGaN QDs by two methods using a plasma-assisted molecular beam epitaxy reactor. In the first method, InGaN QDs were formed by SK growth mode on (0001) GaN/sapphire. In the second, I have addressed the limitations of the SK growth of InGaN QDs by developing a novel alternative method, which was utilized to grow on both (0001) GaN/sapphire and AlN/sapphire. This method relies upon the ability to form thermodynamically stable In-Ga liquid solutions throughout the entire compositional range at relatively low temperatures. Upon simultaneous or sequential deposition of In and Ga on a substrate, the adatoms form a liquid solution, whose composition is controlled by the ratio of the fluxes of the two constituents FIn/(FIn+FGa). Depending on the interfacial free energy between the liquid deposit and substrate, the liquid deposit and vapor, and the vapor and substrate, the liquid deposit forms Inx-Ga1−x nano-droplets on the substrate. These nano-droplets convert into InxGa1−xN QDs upon exposure to nitrogen RF plasma. InGaN QDs produced by both methods were investigated in-situ by reflection high-energy electron diffraction and ex-situ by atomic force microscopy, field emission scanning electron microscopy, transmission electron microscopy, high resolution x-ray diffraction, and grazing incidence small angle x-ray scattering. The optical activity and device potential of the QDs were investigated by photoluminescence measurements and the formation and evaluation of PIN devices (in which the intrinsic region contains QDs embedded within a higher bandgap matrix). InGaN QDs with areal densities ranging from 109 to 1011 cm−2 and diameters ranging from 11 to 39 nm were achieved.

Le travail qui est présenté dans ce manuscrit est le résultat d'une série d'expériences qui ont été effectuées à la fois à l'Université du Maine (IMMM Le Mans) et aux lignes de lumière lD10 et lD02 de l'ESRF (Grenoble) où j'ai passé la moitié de mon temps. Le projet que j'ai travaillé pendant trois ans a été principalement orienté sur l'étude des nanomatériaux qui ont été exposés au CO2 supercritique en utilisant de diffusion des rayons X. Par conséquent une partie de ce travail a été consacrée à la description des propriétés de ce fluide supercritique et comment il interagit avec des matériaux tels que les polymères par exemple. Les matériaux analysés ont été le polystyrène sous forme de film et des ilots, ensuite des matériaux meso-structures avec des tensioactifs fluorés et finalement le carbonate de calcium. Due à la taille nanométrique de tous ces matériaux, les techniques de rayons X qui ont été largement utilisés dans ce travail étaient Small angle X-ray scattering (SAXS), Grazing Small angle X-ray scattering ( (GISAXS) et X-ray reflectivity (XRR). / The work that is presented in this manuscript is the result of a series of experiments that were performed both at the Université du Maine (IMMM Le Mans) and at the ID10 and ID02 beam lines of the ESRF (Grenoble) where I have equally spent half of my time. The project I have been working on for three years was mostly oriented on the study by means of X-ray scattering probes of nanomaterials that were exposed to supercritical CO2. As a result another part of this work will be also dedicated to describing the properties of this supercritical fluid and how it interacts with materials such as polymers for instance. The analyzed materials were thin film and small island of polystyrene, then materials mesostructures using fluoro-surfactants and finally calcium carbonate. Due to the nanoscale of these materials, the X-ray probes that were extensively used in this work were Small Angle X-ray scattering (SAXS), Grazing Incidence Small Angle X-ray Scattering (GISAXS) and X-ray Reflectivity (XRR).

No description available.

The techniques of small-angle x-ray scattering (SAXS) and grazing-incidence small-angle x-ray scattering (GISAXS) have successfully been used for many years in the analysis of nanostructures in non-crystalline samples, e.g., polymers, metallic alloys, ceramics, and glasses. In many specimens, however, the nanostructure is not distributed homogeneously, but instead varies as a function of position in the sample. Conventional SAXS or GISAXS measurements on such heterogeneous samples merely yield an averaged scattering pattern of all the different structures present along the x-ray beam path. In this thesis, scanning tomography is combined with SAXS and GISAXS, revealing the individual local scattering cross section at each position on a virtual section through the sample. The technique thereby offers unique analytical possibilities in heterogeneous specimens. A brief review of the physics of x rays and x-ray scattering is given, before the methods of tomographic SAXS and GISAXS are introduced. Experimental requirements and limitations of both methods are discussed, including aspects of sampling, local rotational invariance and x-ray beam coherence. Experiments performed at the beamline BW4 at HASYLAB at DESY, Hamburg, Germany are described, illustrating the capabilities of the method. Finally, an outlook on possible future developments in tomographic small-angle x-ray scattering is given. / Die Methoden der Röntgenkleinwinkelstreuung (SAXS) und Röntgenkleinwinkelstreuung unter streifendem Einfall (GISAXS) werden seit vielen Jahren erfolgreich eingesetzt zur Analyse von Nanostrukturen in nicht-kristallinen Proben, z.B. Polymeren, metallischen Legierungen, Keramiken und Gläsern. In vielen Proben ist die Nanostruktur allerdings nicht homogen verteilt, sondern variiert als Funktion des Ortes in der Probe. Konventionelle SAXS- oder GISAXS-Messungen an solch heterogenen Proben liefern lediglich ein über alle unterschiedlichen Strukturen entlang des Röntgenstrahls gemitteltes Streubild. In dieser Arbeit wird Rastertomographie mit SAXS und GISAXS kombiniert und so der lokale Streuquerschnitt an jedem Ort auf einem virtuellen Schnitt durch die Probe gewonnen. Diese Technik bietet so einzigartige Analysemöglichkeiten von heterogenen Proben. Es wird zunächst ein kurzer Überblick über die Physik der Röntgenstrahlung und Röntgenstreuung gegeben, bevor die Methoden der SAXS- und GISAXS-Tomographie eingeführt werden. Die experimentellen Anforderungen und Grenzen beider Methoden werden besprochen, wobei Aspekte der Abtastung, der lokalen Rotationsinvarianz und der Kohärenz im Röntgenstrahl eine Rolle spielen. Experimente, die an der Messstrecke BW4 am HASYLAB bei DESY, Hamburg, durchgeführt wurden, werden beschrieben, um die Möglichkeiten der Methode zu illustrieren. Schließlich wird ein Ausblick auf mögliche zukünftige Entwicklungen der Kleinwinkelstreutomographie gegeben.