This thesis details the synthesis and coordination chemistry of twenty-five nitrogencontaining heterocyclic ligands, nineteen of which were previously unreported compounds. These ligands were designed for use as synthons for the formation of molecular cages, so contain multiple coordination sites capable of bridging multiple metal atoms. The majority of molecular cages in the literature are formed by rigid bridging ligands, whereas the ligands studied in this research incorporate a higher level of flexibility, thereby lessening the degree of control over the self-assembly process and increasing the number of possible structures that can be formed upon reaction of these ligands with meal salts. Three of the new ligands synthesised were two-armed bridging ligands, which were reacted with a wide variety of metal salts to investigate what self-assembly products were formed. The complexes characterised include a M₃L₃ cyclic trimer, a range of coordination polymers of varying dimensionality, a range of dimeric products and a series of M₄L₆ cage-like molecular squares. However, the majority of ligands studied were three-armed, potentially tripodal compounds, which were envisaged as potential components of M₃L₂ or M₆L₄ molecular cages. The products of self-assembly of these ligands with various metals salts were shown to include a variety of discrete tri- and tetranuclear complexes, a range of coordination polymers of varying dimensionality and interpenetration, and a complex M₆L₄ assembly that appears to be a collapsed coordination cage. Unfortunately some of the ligands synthesised were shown to decompose in the presence of various metal salts, a phenomenon already identified in the literature. Analogues of these decomposition products were synthesised deliberately to identify the potential of a known tridentate ligand as a metallosupramolecular synthon. ¹H NMR spectroscopy, mass spectrometry, elemental analysis, thermogravimetric analysis and X-ray crystallography were used to study the compounds synthesised. The crystal structures of five ligands and fifty-one complexes are discussed.

This thesis details the synthesis and coordination chemistry of twenty-five nitrogencontaining heterocyclic ligands, nineteen of which were previously unreported compounds. These ligands were designed for use as synthons for the formation of molecular cages, so contain multiple coordination sites capable of bridging multiple metal atoms. The majority of molecular cages in the literature are formed by rigid bridging ligands, whereas the ligands studied in this research incorporate a higher level of flexibility, thereby lessening the degree of control over the self-assembly process and increasing the number of possible structures that can be formed upon reaction of these ligands with meal salts. Three of the new ligands synthesised were two-armed bridging ligands, which were reacted with a wide variety of metal salts to investigate what self-assembly products were formed. The complexes characterised include a M₃L₃ cyclic trimer, a range of coordination polymers of varying dimensionality, a range of dimeric products and a series of M₄L₆ cage-like molecular squares. However, the majority of ligands studied were three-armed, potentially tripodal compounds, which were envisaged as potential components of M₃L₂ or M₆L₄ molecular cages. The products of self-assembly of these ligands with various metals salts were shown to include a variety of discrete tri- and tetranuclear complexes, a range of coordination polymers of varying dimensionality and interpenetration, and a complex M₆L₄ assembly that appears to be a collapsed coordination cage. Unfortunately some of the ligands synthesised were shown to decompose in the presence of various metal salts, a phenomenon already identified in the literature. Analogues of these decomposition products were synthesised deliberately to identify the potential of a known tridentate ligand as a metallosupramolecular synthon. ¹H NMR spectroscopy, mass spectrometry, elemental analysis, thermogravimetric analysis and X-ray crystallography were used to study the compounds synthesised. The crystal structures of five ligands and fifty-one complexes are discussed.

L'Imagerie par Résonance Magnétique (IRM) est une technique d'imagerie médicale largement répandue dans les milieux hospitaliers pour le diagnostic de pathologies. Elle repose classiquement sur la détection du proton (IRM 1H) et permet de visualiser des tissus en profondeur avec une très bonne résolution temporelle et spatiale. Cependant, cette méthode souffre encore de sa faible sensibilité. Une des solutions consiste en l'introduction et la détection de xénon hyperpolarisé. En effet, le xénon est un gaz non toxique, très sensible à son environnement chimique et adapté pour l’IRM. Cependant, il n'est spécifique d'aucun récepteur biologique et nécessite des molécules particulièrement adaptées pour son encapsulation. La détection de cibles spécifiques se fait par des biosondes constituées de molécules cages fonctionnalisées par une antenne de reconnaissance d'un récepteur spécifique. Le xénon vient s'encapsuler dans cette molécule hôte et permet la localisation de la cible biologique. Parmi les molécules cages répertoriées dans la littérature, les cryptophanes présentent la plus forte affinité connue pour le xénon et sont donc les plus prometteuses. Les cryptophanes sont des molécules cages constituées de deux unités de type cyclotribenzylène reliées entre elles par trois chaînes pontantes. Ils ont été synthétisés pour la première fois par l'équipe d'A. Collet au Collège de France au début des années 1980. L'objectif de cette thèse a été de synthétiser et de fonctionnaliser de nouveaux cryptophanes. / Non-invasive proton magnetic resonance imaging (1H MRI) is a powerful clinical tool for the detection of numerous diseases. Although MRI contrast agents are often used to improve diagnostic specificity, this technique has limited applications in molecular imaging because of its inherently low sensitivity when compared to nuclear medicine or fluorescence imaging. Laser-polarized 129Xe NMR spectroscopy is a promising tool to circumvent sensitivity limitations. Indeed, optical pumping increases the nuclear spin polarization of xenon by several orders of magnitude (104 to 105), thus small amounts of gas dissolved in biological tissues (blood, lungs…) can be rapidly detected with an excellent signal-to-noise ratio. In addition, the high polarizability of the xenon electron cloud, which induces a very high sensitivity to its environment, makes this nucleus very attractive for molecular imaging. Detection of biomolecules can be achieved by biosensors, which encapsulate xenon atoms in molecular cages that have been functionalized to bind the desired biological target. Cage molecules such as cryptophanes have high affinity for xenon and thus appear as ideal candidates for its encapsulation. During this PhD thesis we worked on the synthesis and the functionalization of new cryptophanes.