Phase transitions are fundamental phenomena in (quantum) many-body systems. They are associated with changes in the macroscopic physical properties of the system in response to the alteration in the conditions controlled by one or more parameters, like temperature or coupling constants. Quantum phase transitions are particularly intriguing as they reveal new insights into the fundamental nature of matter and the laws of physics. The study of phase transitions in such systems is crucial in aiding our understanding of how materials behave in extreme conditions, which are difficult to replicate in laboratory, and also the behavior of exotic states of matter with unique and potentially useful properties like superconductors and superfluids. Moreover, this understanding has other practical applications and can lead to the development of new materials with specific properties or more efficient technologies, such as quantum computers. Hence, detecting the transition point from one phase of matter to another and constructing the corresponding phase diagram is of great importance for examining many-body systems and predicting their response to external perturbations. Traditionally, phase transitions have been identified either through analytical methods like mean field theory or numerical simulations. The pinpointing of the critical value normally involves the measure of specific quantities such as local observables, correlation functions, energy gaps, etc. reflecting the changes in the physics through the transition. However, the latter approach requires prior knowledge of the system to calculate the order parameter of the transition, which is uniquely associated to its universality class. Recently, another method has gained more and more attention in the physics community. By using raw and very general representative data of the system, one can resort to machine learning techniques to distinguish among patterns within the data belonging to different phases. The relevance of these techniques is rooted in the ability of a properly trained machine to efficiently process complex data for the sake of pursuing classification tasks, pattern recognition, generating brand new data and even developing decision processes. The aim of this thesis is to explore phase transitions from this new and promising data-centric perspective. On the one hand, our work is focused on the developement of new machine learning architectures using state-of-the-art and interpretable models. On the other hand, we are interested in the study of the various possible data which can be fed to the artificial intelligence model for the mapping of a quantum many-body system phase diagram. Our analysis is supported by numerical examples obtained via matrix-product-states (MPS) simulations for several one-dimensional zero-temperature systems on a lattice such as the XXZ model, the Extended Bose-Hubbard model (EBH) and the two-species Bose Hubbard model (BH2S). In Part I, we provide a general introduction to the background concepts for the understanding of the physics and the numerical methods used for the simulations and the analysis with deep learning. In Part II, we first present the models of the quantum many-body systems that we study. Then, we discuss the machine learning protocol to identify phase transitions, namely anomaly detection technique, that involves the training of a model on a dataset of normal behavior and use it to recognize deviations from this behavior on test data. The latter can be applied for our purpose by training in a known phase so that, at test-time, all the other phases of the system are marked as anomalies. Our method is based on Generative Adversarial Networks (GANs) and improves the networks adopted by the previous works in the literature for the anomaly detection scheme taking advantage of the adversarial training procedure. Specifically, we train the GAN on a dataset composed of bipartite entanglement spectra (ES) obtained from Tensor Network simulations for the three aforementioned quantum systems. We focus our study on the detection of the elusive Berezinskii-Kosterlitz-Thouless (BKT) transition that have been object of intense theoretical and experimental studies since its first prediction for the classical two-dimensional XY model. The absence of an explicit symmetry breaking and its gappless-to-gapped nature which characterize such a transition make the latter very subtle to be detected, hence providing a challenging testing ground for the machine-driven method. We train the GAN architecture on the ES data in the gapless side of BKT transition and we show that the GAN is able to automatically distinguish between data from the same phase and beyond the BKT. The protocol that we develop is not supposed to become a substitute to the traditional methods for the phase transitions detection but allows to obtain a qualitative map of a phase diagram with almost no prior knowledge about the nature and the arrangement of the phases -- in this sense we refer to it as agnostic -- in an automatic fashion. Furthermore, it is very general and it can be applied in principle to all kind of representative data of the system coming both from experiments and numerics, as long as they have different patterns (even hidden to the eye) in different phases. Since the kind of data is crucially linked with the success of the detection, together with the ES we investigate another candidate: the probability density function (PDF) of a globally U(1) conserved charge in an extensive sub-portion of the system. The full PDF is one of the possible reductions of the ES which is known to exhibit relations and degeneracies reflecting very peculiar aspects of the physics and the symmetries of the system. Its patterns are often used to tell different kinds of phases apart and embed information about non-local quantum correlations. However, the PDF is measurable, e.g. in quantum gas microscopes experiments, and it is quite general so that it can be considered not only in the cases of the study but also in other systems with different symmetries and dimensionalities. Both the ES and the PDF can be extracted from the simulation of the ground state by dividing the one-dimensional chain into two complementary subportions. For the EBH we calculate the PDF of the bosonic occupation number in a wide range of values of the couplings and we are able to reproduce the very rich phase diagram containing several phases (superfluid, Mott insulator, charge density wave, phase separation of supersolid and superfluid and the topological Haldane insulator) just with an educated gaussian fit of the PDF. Even without resorting to machine learning, this analysis is instrumental to show the importance of the experimentally accessible PDF for the task. Moreover, we highlight some of its properties according to the gapless and gapped nature of the ground state which require a further investigation and extension beyond zero-temperature regimes and one-dimensional systems. The last chapter of the results contains the description of another architecture, namely the Concrete Autoencoder (CAE) which can be used for detecting phase transitions with the anomaly detection scheme while being able to automatically learn what the most relevant components of the input data are. We show that the CAE can recognize the important eigenvalues out of the entire ES for the EBH model in order to characterize the gapless phase. Therefore the latter architecture can be used to provide not only a more compact version of the input data (dimensionality reduction) -- which can improve the training -- but also some meaningful insights in the spirit of machine learning interpretability. In conclusion, in this thesis we describe two advances in the solution to the problem of phase recognition in quantum many-body systems. On one side, we improve the literature standard anomaly detection protocol for an automatic and agnostic identification of the phases by employing the GAN network. Moreover, we implement and test an explainable model which can make the interpretation of the results easier. On the other side we put the focus on the PDF as a new candidate quantity for the scope of discerning phases of matter. We show that it contains a lot of information about the many-body state being very general and experimentally accessible.

No description available.

No description available.

No description available.

The work presented in this thesis was broadly focussed on the investigation of the magnetic behaviour of different superconducting materials in the form of bulk (singe crystals and pellets) and thin films (nanomagnetic devices like superconducting spin valves etc). Neutrons and muons were extensively used to probe the structural and magnetic behaviour of these systems at the nanoscale along with bulk characterisation techniques like high-sensitive magnetic property measurements, scanning probe microscopy and magneto-transport measurements etc. The nanoscale interplay of Superconductivity and Ferromagnetism was studied in the thin film structures using a combination of Polarised Neutron Reflectivity (PNR) and Low Energy Muon Spin Rotation (LE-µSR) techniques while bulk Muon Spin Rotation (µSR) technique was used for microscopic magnetic investigation in the bulk materials. In the Fe/Pb heterostructure, evidence of the Proximity Effect was observed in the form of an enhancement of the superconducting penetration depth (λs) with an increase in the ferromagnetic layer thickness (dF) in both the bilayered and the trilayered structures. The existence of an Inverted Magnetic Region was also detected at the Ferromagnet-Superconductor (F/S) interface in the normal state possibly originating from the induced spin polarisation within the Pb layer in the presence of the neighbouring Fe layer(s). The spatial size (height and width) of the Inverted Magnetic Region did not change much while cooling the sample below the superconducting transition temperature(Tc)and it also stayed unaffected by an increase in the Fe layer thickness and by a change of the applied magnetic field. In the superconducting spin valve structure containing Permalloy (Py) as ferromagnetic layer and Nb as the superconducting layer, LE-µSR measurements revealed the evidence of the decay of magnetic flux density (as a function of thickness) within the Nb layer symmetrically from the Py/Nb interfaces towards the centre of the Nb layer in the normal state. The thickness dependent magnetisation decay occurred over two characteristic length scales in the normal state that stayed of similar values in the superconducting state also. In the superconducting state, an additional contribution towards the magnetisation was found in the vicinity of the Py/Nb interfaces possibly originating from the spin polarisation of the singlet Cooper pairs in these areas. The nanoscale magnetic investigation on a highly engineered F/S/F structure (where each of the F blocks made of multiple Co/Pd layers with magnetic moments aligned perpendicular to the plane of these layers and neighbouring magnetic blocks separated by Ru layers giving rise to antiferromagnetic alignment) using LE-µSR showed an antisymmetric thickness dependent magnetic flux density profile with two characteristic length scales. In the superconducting state, the magnetic flux density profile got modified within the superconducting Nb₆₇Ti₃₃ layer near the F/S interfaces in a way similar to that of observed in the case of Py/Nb system, most likely because of the spin polarisation of the superconducting electron pairs. The vortex magnetic phase diagram of Bi₂Sr₂Ca₂Cu₃O10-δ was studied using the Muon Spin Rotation (µSR) technique to explore the effects of vortex lattice melting and rearrangements for vortex transitions and crossover as a function of magnetic field and temperatures. At low magnetic fields, the flux vortices undergo a first order melting transition from a vortex lattice to a vortex liquid state with increasing temperature while another transition also occurred with increasing field at fixed temperature to a vortex glass phase at the lowest temperatures. Evidence of a frozen liquid phase was found in the intermediate field region at low temperature in the form of a lagoon in the superconducting vortex state which is in agreement with earlier observations made in BiSCCO-2212. The magnetic behaviour of the unconventional superconductor Sr₂RuO₄ was investigated using µSR to find the evidence of normal state magnetism and the nature of the vortex state. In the normal state, a weak hysteretic magnetic signal was detected over a wide temperature and field range believed to be supporting the evidence of a chiral order parameter. The nature of the vortex lattice structure was obtained in different parts of the magnetic phase diagram and the evidence of magnetic field driven transition in the lattice structure was detected from a Triangular→Square structure while the vortex lattice stayed Triangular over the entire temperature region below Tc at low fields with a disappearance of pinning at higher temperatures.

Ce mémoire traite des propriétés du La2CuO4 dopé en trous, le premier supraconducteur à haute température critique ayant été découvert. Les différentes phases électroniques du cristal y seront présentées, ainsi que le diagramme de phases en dopage de ce matériau. Les trois structures dans lesquelles on peut retrouver ce cristal seront décrites en détail, et leurs liens présumés avec les phases électroniques seront présentés. Il s’en suivra une étude utilisant la théorie de la fonctionnelle de la densité combinée au modèle de Hubbard (DFT+U) des différentes phases structurales, en plus des phases antiferromagnétiques et paramagnétiques. L’effet de la corrélation électronique sur la structure cristalline sera également étudié par l’intermédiaire du paramètre de Hubbard. Le but sera de vérifier si la DFT+U reproduit bien le diagramme de phases expérimentales, et sous quelles conditions. Une étude des effets de l’inclinaison des octaèdres d’oxygène sur la structure électronique sera également présentée. / Presented here is a study on the hole doped La2CuO4, the first discovered high-­‐Tc superconductor of the cuprate family. The different electronic phases of this crystal are briefly reviewed. The three crystal structures present in this material are described, and the link between those phases and the electronic structure are discussed. The relationship of those structural phases with the magnetic phases is investigated with the help of calculations based on the density functional theory where an additional Hubbard term has been added (DFT+U). With the help of the Hubbard parameter, the effect of the electronic correlation’s strength on the structural parameters of the crystal is also studied. The idea here is to verify how well the DFT+U is able to reproduce the experimental phase diagram of this material. The effect of the tilting of the oxygen octahedras on the electronic structure is also addressed.

L’électrofilage est une technique permettant de fabriquer des fibres polymériques dont le diamètre varie entre quelques nanomètres et quelques microns. Ces fibres ont donc un rapport surface/volume très élevé. Les fibres électrofilées pourraient trouver des applications dans le relargage de médicaments et le génie tissulaire, comme membranes et capteurs chimiques, ou dans les nanocomposites et dispositifs électroniques. L’électrofilage était initialement utilisé pour préparer des toiles de fibres désordonnées, mais il est maintenant possible d’aligner les fibres par l’usage de collecteurs spéciaux. Cependant, il est important de contrôler non seulement l’alignement macroscopique des fibres mais aussi leur orientation au niveau moléculaire puisque l’orientation influence les propriétés mécaniques, optiques et électriques des polymères. Les complexes moléculaires apparaissent comme une cible de choix pour produire des nanofibres fortement orientées. Dans les complexes d’inclusion d’urée, les chaînes polymères sont empilées dans des canaux unidimensionnels construits à partir d’un réseau tridimensionnel de molécules d’urée liées par des ponts hydrogène. Ainsi, les chaînes polymère sonts très allongées à l’échelle moléculaire. Des nanofibres du complexe PEO-urée ont été préparées pour la première fois par électrofilage de suspensions et de solutions. Tel qu’attendu, une orientation moléculaire inhabituellement élevée a été observée dans ces fibres. De tels complexes orientés pourraient être utilisés à la fois dans des études fondamentales et dans la préparation de matériaux hiérarchiquement structurés. La méthode d’électrofilage peut parfois aussi être utilisée pour préparer des matériaux polymériques métastables qui ne peuvent pas être préparés par des méthodes conventionnelles. Ici, l’électrofilage a été utilisé pour préparer des fibres des complexes stables (α) et "métastables" (β) entre le PEO et l’urée. La caractérisation du complexe β, qui était mal connu, révèle un rapport PEO:urée de 12:8 appartenant au système orthorhombique avec a = 1.907 nm, b = 0.862 nm et c = 0.773 nm. Les chaînes de PEO sont orientées selon l’axe de la fibre. Leur conformation est significativement affectée par les ponts hydrogène. Une structure en couches a été suggérée pour la forme β, plutôt que la structure conventionnelle en canaux adoptée par la forme α. Nos résultats indiquent que le complexe β est thermodynamiquement stable avant sa fonte et peut se transformer en forme α et en PEO liquide par un processus de fonte et recristallisation à 89 ºC. Ceci va dans le sens contraire aux observations faites avec le complexe β obtenu par trempe du complexe α fondu. En effet, le complexe β ainsi obtenu est métastable et contient des cristaux d’urée. Il peut subir une transition de phases cinétique solide-solide pour produire du complexe α dans une vaste gamme de températures. Cette transition est induite par un changement de conformation du PEO et par la formation de ponts hydrogène intermoléculaires entre l’urée et le PEO. Le diagramme de phases du système PEO-urée a été tracé sur toute la gamme de compositions, ce qui a permis d’interpréter la formation de plusieurs mélanges qui ne sont pas à l’équilibre mais qui sont été observés expérimentalement. La structure et le diagramme de phases du complexe PEO-thiourée, qui est aussi un complexe très mal connu, ont été étudiés en détail. Un rapport molaire PEO :thiourée de 3:2 a été déduit pour le complexe, et une cellule monoclinique avec a = 0.915 nm, b = 1.888 nm, c = 0.825 nm et β = 92.35º a été déterminée. Comme pour le complexe PEO-urée de forme β, une structure en couches a été suggérée pour le complexe PEO-thiourée, dans laquelle les molécules de thiourée seraient disposées en rubans intercalés entre deux couches de PEO. Cette structure en couches pourrait expliquer la température de fusion beaucoup plus faible des complexes PEO-thiourée (110 ºC) et PEO-urée de forme β (89 ºC) en comparaison aux structures en canaux du complexe PEO-urée de forme α (143 ºC). / Electrospinning is a technique that allows production of polymeric fibers with diameters ranging from nanometers to a few microns, and thus with an inherent high surface-to-volume ratio. Electrospun fibers are finding potential applications in drug delivery and tissue engineering, as membranes and chemical sensors, and in nanocomposites and electronic devices. Electrospinning was initially used to prepare disordered, non-woven mats, but it is now possible to produce highly aligned fibers by using different target collectors. However, it is of great interest to not only control the macroscopic alignment of the fibers but also their orientation at the molecular level since it influences the mechanical, optical and electrical properties of polymers. Molecular complexes were targeted as a means of increasing molecular orientation in electrospun fibers. In the host-guest urea inclusion complexes (ICs), polymer chains are packed in one-dimensional channels constructed from an essentially infinite three-dimensional network of hydrogen-bonded urea molecules. The polymer chains are thus highly extended at the molecular scale. PEO-urea complex nanofibers have been prepared for the first time by electrospinning of suspension and solutions. As predicted, an unusually large molecular orientation in the fibers was achieved. Such highly ordered IC fibers could find use both for fundamental studies of the inclusion complexes and for the preparation of hierarchically structured materials. Electrospinning can also sometimes be used to prepare metastable polymeric materials that cannot be prepared by the conventional methods. Here, solution electrospinning was used to prepare fibers of both the stable (α) and "metastable" (β) complexes between PEO and urea. Detailed characterization of the ill-studied β complex reveals that it possesses a 12:8 PEO:urea stoichiometry and belongs to the orthorhombic system with a = 1.907 nm, b = 0.862 nm, and c = 0.773 nm. The PEO chains are oriented along the fiber axis and present a conformation significantly affected by strong hydrogen bonding with urea as compared to the pure polymer and the stable α complex. A layered structure, rather than the conventional channel structure, is suggested. In contrast with previous suggestions based on melt-quenched PEO-urea α complex, our results further indicate that the β complex is thermodynamically stable before melting and can phase-transfer to the α complex and liquid PEO through a thermodynamic melt-recrystallization process at 89 ºC. In contrast, the β complex obtained by melt-quenching the α complex is mixed with urea crystal and is metastable. It can experience a kinetic solid-solid phase transition process to produce α complex within a large temperature range. This transition is induced by a PEO conformation change and by the formation of intermolecular hydrogen bonds between urea and PEO. The phase diagram of the PEO/urea system was drawn over the complete composition range, which allowed interpreting the formation of various out-of-equilibrium mixtures observed experimentally. The structure and phase diagram of the PEO/thiourea complex, another poorly understood system, was also studied in detail. An EO:thiourea molar ratio of 3:2 was deduced for the complex, and a monoclinic unit cell with a = 0.915 nm, b = 1.888 nm, c = 0.825 nm and β = 92.35º was determined. Just as for the PEO-urea β complex, a layered structure was suggested for the PEO-thiourea complex, in which the thiourea molecules would be arranged into a ribbon-like structure intercalated between two PEO layers. This layered structure could explain the much lower melting temperature of the PEO-thiourea (110 ºC) and PEO-urea β complexes (89 ºC) as compared to the well known channel-structured PEO-urea α complex (143 ºC).

Ce mémoire traite des propriétés du La2CuO4 dopé en trous, le premier supraconducteur à haute température critique ayant été découvert. Les différentes phases électroniques du cristal y seront présentées, ainsi que le diagramme de phases en dopage de ce matériau. Les trois structures dans lesquelles on peut retrouver ce cristal seront décrites en détail, et leurs liens présumés avec les phases électroniques seront présentés. Il s’en suivra une étude utilisant la théorie de la fonctionnelle de la densité combinée au modèle de Hubbard (DFT+U) des différentes phases structurales, en plus des phases antiferromagnétiques et paramagnétiques. L’effet de la corrélation électronique sur la structure cristalline sera également étudié par l’intermédiaire du paramètre de Hubbard. Le but sera de vérifier si la DFT+U reproduit bien le diagramme de phases expérimentales, et sous quelles conditions. Une étude des effets de l’inclinaison des octaèdres d’oxygène sur la structure électronique sera également présentée. / Presented here is a study on the hole doped La2CuO4, the first discovered high-­‐Tc superconductor of the cuprate family. The different electronic phases of this crystal are briefly reviewed. The three crystal structures present in this material are described, and the link between those phases and the electronic structure are discussed. The relationship of those structural phases with the magnetic phases is investigated with the help of calculations based on the density functional theory where an additional Hubbard term has been added (DFT+U). With the help of the Hubbard parameter, the effect of the electronic correlation’s strength on the structural parameters of the crystal is also studied. The idea here is to verify how well the DFT+U is able to reproduce the experimental phase diagram of this material. The effect of the tilting of the oxygen octahedras on the electronic structure is also addressed.

Dans cette thèse, nous avons synthétisé de nouveaux matériaux moléculaires hétérométalliques pontés par des groupements cyanures en utilisant la chimie de coordination donneur-accepteur. Les précurseurs moléculaires basés sur un ion métallique 3d (MnII/III, FeII, NiII, CuII) agissent en tant qu’unités acceptrices, alors que les précurseurs 4d (RuIII) et 5d (ReIV) de type trans-cyanido ont été utilisés comme groupements donneurs. Le chapitre I présente une approche théorique avec quelques exemples des matériaux magnétiques de basse dimensionnalité : molécule et chaîne aimant qui illustrent le rôle du ligand cyanido pour concevoir les matériaux magnétiques. En outre, nous avons sélectionné des précurseurs cyanido à base de métaux 4d/5d pour leur anisotropie magnétique plus élevée par rapport aux ions de métaux 3d. Le choix des précurseurs moléculaires, leur synthèse et leur caractérisation ont été largement décrits dans le chapitre II. Au cours de ce travail de recherche, nous avons obtenu de nombreux systèmes moléculaires hétérométalliques mais également des systèmes unidimensionnels. Les chapitres III, IV et V décrivent la synthèse, la structure cristallographique et la caractérisation magnétique de tous les complexes synthétisés. L’empilement cristallin compact des molécules conduit à un état fondamental antiferromagnétique pour la plupart des complexes ce qui perturbe souvent les phénomènes de relaxation magnétique. Nous avons donc établi les diagrammes de phases magnétiques pour la plupart de ces systèmes. De plus, certaines des chaînes ReIV/MnIII et une chaîne ReIV/FeII présentent des comportements magnétiques de type "chaîne aimant" et donc une bistabilité magnétique. / In this thesis, we have synthesised new cyanido bridged heterometallic molecule-based magnetic materials with the tool of simple donor-acceptor coordination chemistry. The 3d metal ion (MnII/III,FeII, NiII, CuII) based molecular precursors acted as acceptor building blocks, whereas the 4d (RuIII)and 5d (ReIV) trans-cyanido type molecular precursors were used as donor moieties.Chapter I contains a theoretical approach with examples of low dimensional magnetic materials:Single Molecule and Single Chain Magnets that illustrate the role of the cyanido ligand to design such magnetic materials. Furthermore, we emphasized the 4d/5d metal based cyanido precursors for their higher magnetic anisotropy over the 3d metal ions. The choice of molecular precursors, their synthesis and characterization were extensively described in chapter II. In this research work, we have obtained many heterometallic molecular complexes and also one dimensional systems. Chapter III, IV and V contain the synthesis, X-ray crystallographic and magnetic characterization of all the newly synthesized complexes. The close crystal packing of the molecules lead to an antiferromagnetic ground state for most of the complexes and this often perturbed the magnetic relaxation phenomena. As expected, we have found metamagnetic phase diagrams for most of these systems. Some of the ReIV/MnIII and a ReIV/FeII one dimensional compounds exhibited Single-Chain magnet like properties and thus magnetic bistability.

Cette étude porte sur la décontamination de matrices textiles en milieu CO2 dense (CO2 liquide ou CO2 supercritique). Elle s'inscrit dans le cadre de la décontamination des textiles utilisés dans l'industrie nucléaire. Le CO2 dense est proposé comme alternative au milieu aqueux utilisé dans le procédé actuel et qui génère une importante quantité d'effluents aqueux contaminés nécessitant un post-traitement. Le contaminant étudié est le cobalt qui peut se présenter sous forme ionique ou particulaire. L'extraction du cobalt en milieu CO2 dense est assurée par un additif : un tensioactif macromoléculaire CO2-phile/CO2-phobe complexant. Plusieurs familles d'additifs ont été synthétisées par polymérisation radicalaire contrôlée : des copolymères à gradient comportant des motifs CO2-philes, siliciés ou fluorés, et des motifs CO2-phobes complexants de types acétoacétoxys, diéthylphosphonates ou acides phosphoniques. Le comportement de ces copolymères dans le CO2 dense a été évalué grâce à la détermination des diagrammes de phases copolymère-CO2 (par la mesure du point de trouble) et grâce à l'étude de leur autoorganisation dans le CO2 dense (par diffusion de neutrons aux petits angles). Les copolymères fluorés se sont avérés être les plus avantageux en termes de solubilité. Néanmoins, les copolymères siliciés présentent une solubilité compatible avec le procédé et ils constituent donc une alternative intéressante pour éviter la présence de fluor gênant pour le conditionnement des déchets nucléaires. L'étude de la complexation du cobalt par les copolymères (par spectrométrie UV-visible et par torche à plasma couplée à un spectromètre d'émission atomique) a permis d'établir des relations entre le type de motif complexant et l'affinité avec le cobalt. La solubilité dans le CO2 dense de ces complexes copolymères-cobalt est comparable à celle des copolymères seuls. De plus, l'étude de l'auto-organisation en milieu CO2 dense a révélé un faible taux d'agrégation des complexes copolymères-cobalt. Enfin, les copolymères synthétisés ont été mis en oeuvre dans les procédés de décontamination particulaire et ionique. Dans le cas du procédé de décontamination ionique, l'emploi du copolymère à gradient poly(acrylate de 1,1,2,2-tétrahydroperfluorodécyle-co-diacide vinylbenzylphosphonique) a permis d'atteindre environ 70% de décontamination grâce à la formation d'une microémulsion d'eau dans le CO2 dense. L'efficacité du procédé dedécontamination a été portée à 97% grâce à l'emploi de pyridine comme tiers additif. / This study is about textile decontamination in dense CO2 (liquid CO2 or supercritical CO2). The study is carried out in the framework of decontamination of textile used in the nuclear industry. The dense CO2 offers an alternative to aqueous medium used in the current process which generates a huge quantity of contaminated aqueous effluent requiring a post-treatment. Cobalt is the targeted contamination and can be found as ionic species or particles. The cobalt extraction in dense CO2 is achieved with an additive : a complexing CO2-philic/CO2-phobic macromolecular surfactant. Several types of additives were synthesized by controlled free radical polymerization : gradient copolymers made with CO2-philic groups (silicone-based or fluorinated moieties) and CO2-phobic complexing groups (acetoacetoxy, diethylphosphonate or phosphonic acid moieties). The copolymer behavior in dense CO2 was determined by phase diagram measurements (cloud point method) and their self-assembly in dense CO2 was investigated by small angle neutron scattering. The fluorinated copolymers were found advantageous in terms of solubility. Nevertheless, the silicone-based copolymers showed solubilities which are compatible with the process, therefore they are a good alternative to avoid fluorinated compounds which are unwanted in the conditioning of nuclear wastes. The study of cobalt complexation by the copolymers (UV-vis spectroscopy and inductively coupled plasma-mass spectroscopy) established relations between the type of complexing group and the affinity with the cobalt. The solubility of copolymer-cobalt complexes in dense CO2 is similar to those of copolymers. Moreover, the self-assembly study of the complex revealed a low aggregation. Finally, the synthesized copolymers were used in particle or ionic decontamination processes. In the case of ionic decontamination process, a rate of 70% of decontamination was reached with the use of gradient copolymer poly(1,1,2,2-tetrahydroperfluorodecyle acrylate-covinylbenzylphosphonic diacid) which allowed the formation of water-in-CO2 microemulsion. The efficiency of the decontamination process was even improved up to 97% with the addition of pyridine in the process.