Ce travail s’inscrit au sein du projet régional de recherche BEMA (Bois Eco Matériaux Aquitaine) qui allie partenaires universitaires et industriels afin de valoriser, à travers la filière panneaux de particules, des ressources abondantes en Aquitaine : le maïs et le Pin Maritime. L’objectif de cette thèse est de greffer de manière covalente des chaînes de polymère à la surface de fibres de bois afin de modifier leur état de surface et de favoriser leur comptabilisation avec un liant naturel et/ou synthétique. Pour atteindre cet objectif, nous avons choisi d’utiliser une technique de polymérisation particulière, la polymérisation radicalaire contrôlée. Le contexte scientifique international du début de la thèse montre la faisabilité du projet de par la réalisation de plusieurs études sur de la cellulose modèle de type papier filtre. Le défi de cette thèse consiste à transposer ce mécanisme sur des fibres de bois brutes afin de leur conférer de nouvelles propriétés (résistance à l’eau, caractère fongicide, meilleure comptabilisation…). Afin d’atteindre l’objectif énoncé ci-dessus, nous avons développé plusieurs aspects de la chimie des polymères en mettant en oeuvre différentes techniques de synthèse (RAFT, NMP) mais aussi de caractérisation adaptées à la présence de fibres de bois brutes (Mesure d’angle de contact, TGA, XPS). Enfin, nous avons valorisé la présence de polymères fonctionnels en insérant à la surface des fibres de bois des nanoparticules d’oxyde de silice ou de métaux de manière contrôlée. / This work is part of the regional research project BEMA (Bois Eco Matériaux Aquitaine) which combines academic and industrial partners to develop, through the wood pannel industry, abundant resources in Aquitaine: corn and Maritime Pine. The objective of this thesis is to covalently graft polymer chains on the surface of wood fibers in order to modify their surface properties and to facilitate their compatibilization with natural and/or synthetic binders. To achieve this goal, we chose to use a special polymerization technique, the controlled radical polymerization. The international scientific context of the beginning of the thesis shows the feasibility of the project through several studies carried out on model cellulose, such as filter paper. The challenge of this thesis is to implement this mechanism on raw wood fibers to give them new properties (water resistance, fungicidal character, best compatibilization ...). To achieve the objective stated above, we investigated several polymer synthetic pathways (RAFT, NMP) but also characterization techniques (Contact angle measurements, TGA, XPS) suitable with the presence of raw wood fibers. What’s more, we have valued the presence of functional polymers by inserting at the wood fibers surface oxide nanoparticles of silica or metal in a controlled manner.

Advances in medical treatment are leading away from generalized care towards intelligent systems or devices which can sense and respond to their environment. With these devices, the burden of monitoring and dosing for treatment can be removed from the doctor (or the patient) and be placed on the device itself. Implicit closed-loop control systems will allow the device to respond to its environment and release therapeutic agent in response to a specific stimulus. Environmentally responsive hydrogels show great promise in being incorporated in such an intelligent device, such as pH-responsive hydrogels which can swell and deswell in response to changes in the pH of the media. Thus, pH changes can be exploited for controlled and intelligent drug delivery when used in combination with these pH-responsive hydrogels. In this work, heterogeneous, thermal-redox initiated free-radical polymerizations were developed to synthesize novel pH-responsive hydrogels, microparticles, and nanogels. The specific disease of interest was type I diabetes, which requires daily doses of insulin both at a basal amount and either a postprandial or preprandial bolus in order to maintain blood glucose levels within safe limits. To allow pH-responsive hydrogels to be sensitive to glucose, glucose oxidase was incorporated which oxidizes glucose to gluconic acid. A novel inverse-emulsion polymerization method was developed for the synthesis of poly[2-(diethylaminoethyl methacrylate)-grafted-polyethylene glycol monoethyl ether monomethacrylate] (P(DEAEM-g-PEGMMA)) nanogels (100-400 nm) for intelligent insulin delivery. The new polymerization method allowed the incorporation of hydrophilic components, such as glucose oxidase and catalase, as well as PEG surface tethers of lengths 400 Da up to 2000 Da. Surface tethers successfully decreased the surface charge of the nanogels. Insulin loading and release was determined for microparticles which were able to imbibe substantial amounts of insulin from solution when swollen, entrap the insulin when collapsed, and then release the insulin in response to either a pH or glucose stimulus. / text

Radicals are core reactive species that occur in almost every subfield of chemistry. In particular, solution phase radicals find their way into biochemistry (e.g. vitamin B12), and in polymer chemistry (e.g. radical polymerizations) just to name a few. Yet, given the proliferation of radical chemistry, there are still fundamental aspects of it that are poorly understood. This dissertation probed factors that influence the solvent cage effect. The solvent cage effect is where two radicals are held in close proximity to one another and prevented from easily escaping (to form free radicals) by a cage of solvent molecules. A convenient metric of the solvent cage effect is the radical recombination efficiency (FcP). Typically, FcP correlates with the bulk viscosity of the solution, however, this parameter only produces qualitative assessments. This dissertation outlines a method to quantitatively predict FcP using the microviscosity. This microviscosity dependence holds for non polar, aromatic, polar, and hydrogen-bonding solvents, along with solutions that contain polymers. Microviscosity is a great metric because it addresses an underlying reason for the solvent cage effect, the strength of the cage. Not only does the strength of the solvent cage around the radical pair affect FcP, but so does the identity of the radicals themselves. That is, the strength of the solvent cage is one piece to forming a total predictive model. FcP for the Cp'2Mo2(CO)6 dimer also varies with the wavelength of irradiation. Identifying the mechanism by which this wavelength dependence occurs may also provide another factor to include in an overall model of the solvent cage effect. Also, an attempt at synthesizing an asymmetric molybdenum dimer was performed. This asymmetric dimer would allow the study of solvent caged radical pairs that are different from each other. Predicting the photochemical cage pair recombination efficiency (FcP) is the major topic of this dissertation. However, there is also the collisional cage recombination efficiency (Fcʹ). This is where free radicals come together in what is called a collisional solvent cage pair. A method and values of Fcʹ are detailed later in this dissertation. This dissertation contains previously published and unpublished co-authored material.