The structure, magnetic behaviour and chemistry of layered oxychalcogenides of composition A₂MO₂X₂Ch₂ (where A = Sr, Ba; M = Mn, Co, Ni, Cu, Zn; X = Cu, Ag and Ch = S, Se, Te) has been investigated by the synthesis of new compounds of this type and control of the properties of these compounds by oxidative deintercalation of Cu/Ag. I₂ can be used to oxidatively deintercalate Cu from Sr₂MnO₂Cu_1.5S₂, forming Sr₂MnO₂Cu_1.33S₂, an incommensurately modulated compound, with a completely different Cu/vacancy ordering and antiferromagnetic ordering structure to the parent. This reaction is also probed in real-time, using in situ powder X-ray diffraction. Sr₂MnO₂Ag_1.5Se₂ was found to have an A-type magnetic ordering structure, similar to Sr₂MnO₂Cu_1.5Se₂. Sr₂MnO₂Cu_1.8Te₂ on the other hand with a lower Mn oxidation state shows only two-dimensional magnetic correlations, rather than long-range order. Extending the reaction with I₂ to several Co-containing analogues revealed that ~ 25% Ag could be removed from Sr₂CoO₂Ag₂Se₂, sufficient to observe a change in magnetic behaviour, from antiferromagnetic to ferromagnetic. By contrast only ~11% Cu can be deintercalated from Sr₂CoO₂Cu₂S₂ and even less (~5%) from Sr₂CoO₂Cu₂Se₂. Neutron diffraction was used to examine the resultant changes in magnetic ordering. The novel compounds Sr₂CuO₂Cu₂Se₂ and Ba₂CuO_2-xCu₂Se₂ are related by substitution of the alkali-earth metal, but while Sr₂CuO₂Cu₂Se₂ is a stoichiometric compound with metal-like character, Ba₂CuO_2-xCu₂Se₂ is an oxygen-deficient semiconductor, with tuneable oxygen content. Unusual features are observed in the magnetic susceptibility measurements of Sr₂NiO₂Cu₂Se₂ that appear unrelated to this compound's long-range magnetic ordering, as probed by neutron diffraction. Furthermore, unusual peak splitting is observed in low-temperature powder X-ray diffraction patterns of this compound; this may plausibly be due to a photon-induced effect arising from the use of a high-energy beamline; although further measurements are required to examine this. Overall the work shows the flexibility and range of behaviour exhibited by a series of the transition metal oxide chalcogenides.

Ce travail de thèse porte sur l’élaboration et la caractérisation des propriétés physiques et chimiques d’une nouvelle famille de composés thermoélectriques, et plus particulièrement le composé BiCuSeO. Les composés de cette famille, dite 1111, présentent une structure en couche de type ZrCuSiAs. L’une des particularités de cette structure est la nature distincte des couches qui la composent, la couche Bi2O2 étant décrite comme isolante tandis que la couche Cu2Se2 est appelée couche conductrice. L’étude approfondie du composé BiCuSeO montre qu’en dépit d’un facteur de puissance (S²σ) relativement modéré, ce composé est un matériau thermoélectrique prometteur, notamment à haute température. En effet, BiCuSeO présente une conductivité thermique remarquablement faible, qui permet d’atteindre des facteurs de mérite relativement élevés. De plus, BiCuSeO présente de nombreuses voies d’améliorations possibles. L’une d’elle concerne l’étude d’un dopage aliovalent sur le site du bismuth. L’analyse des résultats a montré que l’insertion d’un élément divalent permet d’optimiser la concentration des porteurs de charges, entrainant ainsi une forte augmentation du facteur de mérite du composé. Une autre voie possible d’exploration est l’étude de l’influence de l’ion chalcogène, au travers notamment de la substitution du sélénium par le tellure, avec l’obtention d’une solution solide complète BiCuSe(1-x)Te(x)O. L’étude des propriétés électriques des composés de cette série a permis de mettre en évidence la présence d’une transition métal – semi-conducteur – métal pour les fractions de tellure inférieures à 0.5. Ainsi, bien que l’influence du tellure sur le facteur de puissance soit relativement limitée en raison de cette anomalie, des résultats intéressants ont été obtenus pour les fractions de tellure élevées. Par ailleurs, des problématiques autour d’une méthode de synthèse alternative du matériau ainsi que sa stabilité sous air sont également abordées dans ce travail. / This thesis addresses the issues of the elaboration and the characterization of the chemical and physical properties of a new family of thermoelectric materials, the oxychalcogenides with the general formula BiCuSeO. This compound, called 1111, cristallises in the ZrCuSiAs structure-type. One feature of this structure lies in the fact that the layers are considered as electronically distinct: the Bi2O2 layers are described as the insulating layers whereas the chalcogenide layers Cu2Se2 are presented as the conductive ones. The study of BiCuSeO exhibits that in spite of a relatively moderate power factor (S²σ), this compound is very promising as possible thermoelectric material, especially at high temperature. Indeed, BiCuSeO shows a remarkably low thermal conductivity, which can achieve relatively high figures of merit. In addition, BiCuSeO offers many ways for improvement. One of them concerns the study of aliovalent doping on the bismuth site. The results showed that the insertion of a divalent element optimizes the charge carriers concentration, leading to a sharp increase in the figure of merit of the compound. Another possible way of exploration lies the study of the influence of the chalcogen ion, notably through the substitution of selenium and tellurium, with a complete solid-solution BiCuSe(1-x)Te(x)O. The study of the electrical properties of this solid solution has highlighted the presence of a metal - semiconductor - metal transition for tellurium fractions below 0.5. Thus, although the influence of tellurium on the power factor is relatively limited due to this anomaly, interesting results were obtained for the high tellurium fractions. In addition, issues around an alternative method of synthesis of the material and its stability in air are also discussed in this work.