Ce manuscrit présente les travaux réalisés sur l’évolution physico-chimique d’un revêtement protecteur développé par la société Messier-Bugatti-Dowty. Ce revêtement est utilisé pour protéger de l’oxydation des freins d’avions en matériaux composites C/C. La protection développée est de type multi-couches et résulte de l’association entre les différents éléments que sont : (i) une solution de phosphate métallique, (ii) un verre borosilicaté, (iii) une solution de silice colloïdale et (iv) un borure métallique. Si la formulation donne de très bons résultats en anti-oxydation, il existe peu d’information sur l’évolution physico-chimique des constituants et sur leurs interactions. Le travail de thèse a donc consisté en une caractérisation multi-techniques (DRX, RMN des solides, microscopie, TOF-SIMS) de systèmes simulant le revêtement. Les premières études ont été réalisées sur un système simplifié puis l’impact de chaque constituant a été mis en évidence en le complexifiant progressivement pour aboutir à la formulation complète. L’ensemble des résultats a permis de confirmer le rôle de chaque constituant dans la formulation, de caractériser le vieillissement de la protection en température et en présence d’eau et a également apporté de nouvelles informations sur la réactivité entre les différents constituants. Les conclusions de cette étude physico-chimique délimitent les avantages et limitations de la protection et nous proposons quelques pistes d’améliorations pour le revêtement. / The manuscript presents our work on the physico-chemical evolution of a protective coating developed by the Messier-Bugatti-Dowty company. This coatong is used to protect aircraft brakes against oxidation, and they are manufactured from C/Ccomposites. The developed protection is multi-layer type and is based on the association between different elements: (i) a metal phosphate solution, (ii) a borosilicate glass, (iii) a solution of colloidal silica and (iv) a metal boride. If the formulation gives very good results as anto-oxidationcoating, there is little information on the physico-chemical evolution of its components and their interactions. The thesis work has consisted of a multi-technique characterization (XRD, solid-state NMR, microscopy, TOF-SIMS) of systems simulating the coating. The first studies were carried out on anextremely simplified system and the impact of each component has been highlighted by adding each component separately, leading finally to the complete formulation. The overall results confirmed the role of each component in the formulation. We characterized the aging of the protection with the temperature and in the presence of water, which provided new information on the reactivity between the various components. This physicochemical study enabled to define the advantages and limitations of the protection, and we could propose new improvements to enhance the efficiency of the coating.

This thesis presents the results of a single-crystal electron paramagnetic resonance (EPR) spectroscopic investigation of defect centers in selected borates and borosilicates (i.e., datolite, danburite, and jeremejevite). The research brings new complementary data to the current understanding of defect structures in minerals, which are not only important to Earth Sciences but also directly relevant to environmental applications (e.g., nuclear waste disposal) and materials science. Single-crystal EPR spectra of a gamma-ray-irradiated datolite from Bergen Hill, New Jersey, USA, reveal the presence of a boron-oxygen hole center (BOHC). Spin-Hamiltonian parameters obtained from single-crystal EPR spectra and radiation-dose-dependence experiments allow us to confirm the BOHC center in datolite as the [BO4]0 type, involving hole trapping on the hydroxyl oxygen atom after the removal of the hydrogen atom: via a reaction O3BOH --> O3BO• + H0, where • denotes the unpaired electron. Density functional theory (DFT) calculations support the proposed structural model, and the calculated 11B hyperfine coupling constants are in excellent agreement with the experimental results. Also, isochronal and isothermal annealing experiments provide information about the thermal stability and decay kinetics of the [BO4]0 center in datolite. The confirmation of the [BO4]0 center and its formation from the O3BOH precursor in datolite are compared with other BOHCs in minerals and are discussed with relevance to the implications for not only understanding of BOHCs in alkali borosilicate glasses but also their applications to nuclear waste disposal. A combined study by use of synchrotron X-ray absorption spectroscopy (XAS), single-crystal EPR and pulse electron spin echo envelope modulation (ESEEM) spectroscopy provides compelling evidence for lattice-bound arsenic in danburite from Charcas, San Luis Potosi, Mexico. Arsenic K-edge X-ray absorption near-edge (XANES) spectra show that the dominant oxidation state is +3, and modeling of the extended X-ray absorption fine structure (EXAFS) spectra suggests that As3+ mainly occupies the Si site. Detailed single-crystal EPR spectra, measured before and after gamma-ray irradiation, reveal three arsenic-associated paramagnetic electron centers (I, II and III). Centers I and II are varieties of the [AsO2]2 radicals, formed from electron trapping on a substitutional As3+ ion at the Si site. This model is also supported by the 11B superhyperfine structures determined by ESEEM spectra at 80 K. Center III is the [AsO3]2 radical, originated from electron trapping on a [AsO4]3¬ group after removal of the O4 atom during gamma-ray irradiation. Therefore, arsenic in danburite is present in both the +3 and +5 oxidation states and preferentially occupies the Si site. Single-crystal EPR spectra of jeremejevite from Cape Cross, Namibia, reveal an S = 1/2 hole center characterized by a hyperfine structure arising from interaction with two equivalent 27Al nuclei. Our results suggest that this aluminum-associated oxygen hole center represents hole trapping on a hydroxyl oxygen atom linked to two equivalent octahedral Al3+ ions, after the removal of the proton (i.e., a VIAl−O−−VIAl center). Periodic ab initio UHF and DFT calculations confirmed the experimental 27Al hyperfine coupling constants and directions, supporting the proposed structural model. Also, isochronal annealing experiments provide information about the thermal stability of the VIAl−O−−VIAl center. These data obtained from the VIAl−O−−VIAl center in jeremejevite provide new insights into analogous defects that have been documented in several other minerals.

Silicate glasses are the most common glass types and have impact on almost every aspect in our lives: from window, containers, to glass fibers for telecommunications. Unlike their crystalline counterparts, glass materials lack long-range order in their atomic arrangement but their structures do possess short and medium range characteristics that play critical roles in their physical and chemical properties. Despite active development of characterization techniques that have contributed to the understanding of glass structures, there remain key challenges in obtaining essential structural features of glasses. Atomistic computer simulations have become an increasingly important method in elucidating the atomic structures and in interpretation and/or prediction of composition-structure-property relationships of complex materials. In this dissertation, classical molecular dynamic (MD) simulations were used to investigate the atomic structures, dynamic and other properties of two important glass systems—aluminosilicate glasses and borosilicate glasses, which are the basis of most industrial and technologically important glasses. Firstly, a comprehensive study of peralkaline Na2O-Al2O3-SiO2 glass with varying Al2O3/SiO2, Na2O/Al2O3, Na2O/SiO2 ratios has been performed to obtain better understanding of the composition–structure–property relationships in this glass system. More than 99% of Al were 4-coordinated in these glasses, validating that Na+ tend to charge balance [AlO4]- network forming units first and then, excess Na+ was used to create non-bridging oxygen (NBO) on Si. As the drop of Na/Al ratio, the percentage of NBO decreases, indicating an increase of the glass network connectivity. In addition, polyhedral connection probability results show that Al tend to be randomly distributed in the glass structure, suggesting a violation of Lowenstein's rule. These structural properties were further used to explain macroscopic properties of glass, such as change of glass transition temperature (Tg) and hardness (Hv) with glass composition. Secondly, molecular dynamics simulations were used to understand the structural, thermal mechanical and diffusion behaviors of spodumene (LiAlSi2O6) crystalline phases and glasses. It was found that β-LiAlSi2O6 has a structure much closer to the glass phase. The α-LiAlSi2O6 phase, however, has a more closed-packed structure and higher density. The diffusion behaviors were also found to be closely related to the atomic structures. Thirdly, the surface atomic structures of a series of sodium borosilicate glasses were studied using recently developed compositional dependent partial charge potentials. This provides insight into: a) the structural difference between glass surface and bulk glass; b) the evolution of bulk and surface structures as the change of glass composition. Lastly, pressure and temperature effects on the structure and properties of borosilicate glass were investigated in detail. A serial data derived from different compression temperatures and pressures enable us to explore the link between the microstructure and macroscopic physical properties. The results show that compression temperature and pressure play important roles in glass densification process and may result various glass densification mechanism. This dissertation demonstrates that atomistic simulations coupled with effective potentials and careful validations have become an effective method in research and design of complex glass materials.

La recherche de matériaux dentaires efficaces est une préoccupation constante de toute l'histoire de la dentisterie. Avec l'émergence de meilleurs produits, le développement de matériaux toujours plus innovants s'est imposé. De plus, la combinaison de différentes sciences comme la chimie, la biologie, la physique et l'ingénierie a permis de mieux connaître les exigences liées à la restauration dentaire. Le projet BIODENSOL est un projet de recherche et de mobilité doctorale entre l'Université de Lyon et LUCIDEON Ltd (Stoke-on- Trent, Royaume-Uni), financé par la Commission européenne. Le projet a été conçu pour favoriser les relations entre la recherche académique et les applications commerciales, pour améliorer les innovations médicales et associer des chimistes de l'état solide avec des praticiens hospitaliers en contact direct avec les patients ayant des problèmes dentaires. Ce projet proposait trois thèses de doctorat permettant de répondre aux conséquences des caries et d'érosion de l'émail par les aliments acides, l'idée de base étant de favoriser la reminéralisation pour éviter ces problèmes. Les recherches ont évolué de manière indépendante selon trois voies distinctes examinant trois matériaux différents qui pourraient amener des solutions potentielles. Le sujet principal de cette thèse concerne l'étude de poudres de verre borosilicaté présentant une séparation de phases et pouvant être incorporées dans des ciments verre ionomères. Les verres borosilicatés sont des matériaux prometteurs qui ont été largement étudiés pour des applications biomédicales, comme par exemple les échafaudages dans les tissus mous où la réparation osseuse. Par analogie avec les verres silicatés développés par Hench en 1969, qui sont reconnus pour leur bioactivité et leurs propriétés antimicrobiennes, les verres borosilicatés pourraient intéresser la dentisterie. Le système de verre étudié ici est un verre basé sur l'association de 5 constituants, SiO2-K2O-B2O3-CaO-Al2O3, qui a une forte tendance à la séparation des phases. L'objectif de ce travail est d'élaborer par fusion/trempe puis caractériser une série de nouvelles formulations de verres borosilicatés, puis de comprendre le mécanisme et la cinétique de dissolution en relation avec leur microstructure et leur composition. La variation des proportions de chaque constituant est déterminée par l'approche des plans d'expérience. L'utilisation d'un traitement thermique pour favoriser la séparation des phases en vue d'influencer le taux de libération cationique a été spécialement étudiée. Le processus de dissolution des borosilicates dans une solution aqueuse neutre (fluide corporel simulé) ou dans une solution acide (simulant des scénarios où des bactéries ou des aliments acides sont présents) fournit des informations sur le type et la concentration des espèces libérées par le verre. Différentes compositions ont été étudiées dans lesquelles SiO2 et K2O sont fixés tandis que les autres éléments varient afin d'évaluer l'effet sur la séparation de phases. Les cinétiques de dissolution des ions B-, Si-, Ca-, K- et Al- peuvent alors être reliées à la chimie du verre et à la microstructure. La séparation de phases amorphes (APS) provoque la séparation d'une phase unique initialement homogène en deux phases ou plus de compositions différentes. Le degré d'interconnectivité des deux phases vitreuses dépend de la nature du mécanisme de séparation de phases. Les verres élaborés sont optiquement transparents, puis deviennent plus ou moins opalescents suite à un traitement thermique. Le degré d'opacité est clairement dépendant de la proportion des éléments entrant dans la composition du verre. Le schéma de la libération d'ions implique que l'une des phases est plus réactive et sensible à l'attaque ; elle sera dissoute plus rapidement du verre... [etc] / The research for efficient dental materials has been a constant throughout the history of dentistry. As better materials emerged, the development of ever more innovative materials has been pushed forward. Moreover the combination of different sciences such as chemistry, biology, physics and engineering has provided better knowledge to the demanding requirements of the dental restorations. The BIODENSOL project is a European commission funded mobility research project managed by the University of Lyon and LUCIDEON Ltd, Stoke-on-Trent, UK. The project is designed to help building relations between academic and commercial research to enhance medical innovations and associates solid state chemists and hospital practicioners in direct contact with the patients with dental problems. This project proposes three PhD studies to address the problems of caries and enamel erosion by acidic foods (leading to sensitive teeth) encouraging re-mineralization to help avoid these problems. The researches independently evolved in three different routes investigating three different materials that could provide potential solutions. The main objective of this thesis concerns the study of Borosilicate phase separated glass powders for glass ionomer cements. As silicate bioglasses developed by Hench in 1969, regarding their bioactivity and antimicrobial properties, borosilicates glasses are promising materials and have been widely studied for biomedical applications for scaffolds in soft tissues and for bone repair and could be of interest in dentistry. The glass system studied here is a borosilicate glass (SiO2-K2O-B2O3-CaO-Al2O3) with a strong tendency to phase separate. The objective of this work is to characterize a series of the novel borosilicate formulations and to understand the mechanism and kinetic of dissolution related to their microstructure and composition. The use of thermal treatment to promote phase separation as a means of influencing the rate of ion leaching was especially studied. The dissolution process of borosilicates in neutral body solution (simulating body fluid) or in acid solution (simulating scenarios where bacteria or acid foods are present) provides information regarding the type and concentration of species released by the glass. Different compositions have been investigated in which SiO2 and K2O are fixed while the other elements are varied in order to assess the effect on the phase separation. As the novel borosilicates produced with the meltquenching technique is immersed in an aqueous environment, B-, Si-, Ca-, K- and Al- species are released to different degrees as a function of the time. Ion leaching trends can be related to the glass chemistry and microstructure. It has been shown that the amorphous phase separation (APS) causes an initially homogeneous single phase to separate into two or more phases of different compositions. The degree of interconnectivity of the two glass phases depend on the nature of the phase separation mechanism. This process can occur by a nucleation and growth process which gives isolated spherical particles or by spinodal decomposition where an interconnected structure is obtained. It is significant that before a heat-treatment the glasses are optically clear, but turn opalescent to different degrees following a heat treatment depending on the wt% of the elements in the composition. The pattern of the ion release implies that one of the phases is more reactive and susceptible to acid attack and will be leached out from the glass earlier. The other phase will remain in the cement improving the mechanical properties of the dental restorative material. Moreover, the mixing of the glass powder with a commercial poly(acrylic acid) has shown interesting mechanical and bioactive properties. This work showed how the ions leaching are influenced by the glass composition and the heat-treatment... [etc]

La diffusion chimique est un phénomène clé dans l’élaboration des verres d’intérêt nucléaire. A haute température, dans le liquide, elle permet l’homogénéisation des flux de matière (précurseurs vitreux et déchets) et conduit à la formation d’un verre homogène après refroidissement. A contrario, dans le liquide surfondu, elle peut être à l’origine de processus de séparation de phase ou de cristallisation qui doivent être contrôlés pour le bon fonctionnement du procédé. Dans cette thèse, l’influence de la diffusion chimique sur les processus de cristallisation et d’homogénéisation du liquide est étudiée pour un verre simplifié de type borosilicate de sodium entre sa température de transition vitreuse et sa température d’élaboration. Pour ce type de système, qualifié de multicomposants, la description des phénomènes diffusifs nécessite le calcul d’une matrice de diffusion prenant en compte la diffusion couplée des espèces. Ces couplages sont retranscrits au travers de mécanismes de diffusion ou « échanges diffusifs » qui sont invariants avec la température. Les énergies d’activation associées à ces échanges sont proches de l’énergie d’activation de l’écoulement visqueux ce qui montre que le flux visqueux et la diffusion chimique sont pilotés par un seul et même mécanisme en lien avec la fréquence de rupture des liaisons Si-O et B-O. Nous mettons également en évidence que dans le liquide surfondu, les échanges diffusifs primaire (SiO2-Na2O) et secondaire (SiO2-B2O3) jouent un rôle prépondérant sur la cinétique de cristallisation et la direction de croissance des phases cristallines (cristobalite et tridymite) formées dans nos systèmes. Ces résultats permettent de justifier l’évolution des gradients de compositions à proximité et loin des cristaux. Dans la dernière partie du manuscrit, une complexification des verres a été initiée en ajoutant du lanthane pour simuler un des lanthanides majoritairement présents dans la composition du verre nucléaire de référence R7T7. Les données obtenues ont révélé un couplage diffusif entre le lanthane et le silicium qui entre en compétition avec les autres couplages mis en évidence dans le ternaire SiO2-Na2O-B2O3. Ce couplage, combiné aux autres résultats de la thèse, permet d’expliquer la formation d’une phase de type borosilicate de lanthane (LaBSiO5). / Chemical diffusion is a key-phenomenon during nuclear glass synthesis. At high temperature, diffusion leads to homogenization of the melt contributing to the transformation of heterogeneous waste and frit precursors to a homogeneous glass after cooling. In contrast, in the supercooled liquid, diffusion is a critical factor affecting phase separation and/or crystallization processes that must be avoided when producing a high quality final product.In this manuscript, the impact of chemical diffusion on crystallization and liquid homogenization is studied for a simplified sodium borosilicate glass between its glass transition temperature and its synthesis temperature. For this kind of system, qualified as multicomponent, the description of diffusive phenomena requires the calculation of a diffusion matrix that takes into account diffusive couplings between species. These couplings can be written in the form of diffusive mechanisms or “diffusive exchanges” that are invariant with temperature. The activation energies associated with these exchanges are close to the activation energy of shear viscosity which suggests that viscous flow and chemical diffusion are driven by a single mechanism related to the frequency of Si-O and B-O bond breaking. It is also highlighted that in the supercooled liquid, the principal diffusive exchange (SiO2-Na2O) and the secondary diffusive exchange (SiO2-B2O3) play a significant role on the kinetics and direction of growth of crystalline phases which are formed in our system. These results are used to rationalise the evolution of compositional gradients in the vicinity and far from crystals. In the last part of this work, a complexification of the glasses was initiated by adding lanthanum to simulate one of the main lanthanides of the R7T7 nuclear glass composition. The data collected reveal diffusive couplings between lanthanum and silicon. These couplings, combined with the other results explain the formation of a lanthanum borosilicate phase (LaBSiO5).

Le but de ce travail de thèse est de contrôler via un traitement de polarisation les propriétés physico-chimiques de surface des verres. Pour cela, l'implémentation d'une charge d'espace au sein de différents verres silicates et borosilicates a été caractérisée. Ces études comparatives ont montré que la déplétion des cations de l’anode vers la cathode induit l’apparition d’un champ électrique et des modifications structurales localisées. De plus, nous avons démontré que la formation d’un plasma entre l’anode et le verre pouvait être à l’origine de phénomènes d'échanges ioniques conduisant notamment à la formation d’entités azotées NO2/N2O4 piégées dans le verre. Par la suite, une étude des propriétés physico-chimiques des surfaces polarisées a été conduite. Les résultats préliminaires obtenus concernent l'influence de la charge d'espace sur (i) la mouillabilité, (ii) la chimie de surface et (iii) les réponses optiques de molécules adsorbées. / The purpose of this work is to control physicochemical properties of glasses surfaces using a thermal polarization treatment. For this, the implementation of a space charge layer has been characterized for various silicates and borosilicates glasses. These comparative studies have shown that cations depletion from anode toward cathode induces both an electric field and structural local changes. To go further, we demonstrated that there is a plasma formation between anode and glass surfaces which could be the cause of ionic exchange phenomena leading to the formation of nitrogen NO2/N2O4 entities entrapped inside glass network. Subsequently, physicochemical properties studies of polarized surfaces were performed. Preliminary results concerning the space charge influence on (i) surface wettability, (ii) surface chemistry and (iii) optical responses of adsorbed molecules on poled glass surface have been obtained.