L’augmentation des fréquences de commutation des alimentations induit de nouvelles problématiques pour les composants. Que ces composants soient actifs ou passifs, il est nécessaire de contrôler les pertes afin que les rendements restent acceptables. La thèse se propose d’étudier et de fabriquer une nouvelle structure de transformateur destiné à s’insérer dans une alimentation à découpage. Ces transformateurs sont produits à l’aide du procédé PECS/SPS, qui est une technologie de frittage. Cette technologie permet de fritter des ferrites tout autour des enroulements primaires et secondaires. Il en résulte que les composants sont monolithiques.Une première partie présente l’état de l’art, ainsi que le magnétisme dans la matière. Il s’ensuit un chapitre dédié à la fabrication des matériaux magnétiques utilisés durant la thèse : les ferrites.Une deuxième partie concerne les ferrites fabriqués par le procédé PECS/SPS. Il est question dans un premier temps d’étudier l’anisotropie magnétocristalline de ces matériaux, et il est possible de la diminuer en jouant sur la composition chimique. Dans un second temps, d’autres grandeurs telles que la perméabilité ou la polarisation sont mesurées, toujours en fonction de la composition chimique. Les principaux résultats montrent que le frittage par le procédé PECS/SPS est plus réducteur que le frittage classique, ce qui dégrade certaines propriétés comme la résistivité des ferrites. Un dernier chapitre est dédié à la réalisation des transformateurs monolithiques. Un protocole détaillé de la fabrication est alors présenté.Une troisième partie illustre le fonctionnement des transformateurs réalisés. Les mesures usuelles permettent d’identifier les inductances propres, mutuelles et de fuite. Les valeurs de ces inductances montrent qu’il est plus judicieux d’utiliser les composants fabriqués non pas en tant que transformateur, mais en tant que coupleur. Enfin un convertisseur basé sur une structure de type VRM est réalisé. La fréquence de découpage est de 2 MHz, le rendement est supérieur à 90 %, et la densité de puissance est de 15 kW/litre. / The increase in switching frequency of power supply induces new problems for the components. These components are active or passive, it is necessary to control the losses so that efficiency remains acceptable. The thesis deals with the study and production of a new transformer structure intended to be part of a switching power supply. These transformers are produced using PECS/SPS method, which is a sintering technology. This technology can be sintered ferrite around the primary and secondary windings. It follows that the components are monolithic.A first part presents the state of the art, and magnetism in the matter. It follows a chapter dedicated to the manufacture of magnetic materials used in the thesis : the ferrites.A second part concerns the ferrites produced by the PECS / SPS process. Firstly, the magnetocrystalline anisotropy of these materials is studied, and it is possible to decrease it by varying the chemical composition. In a second stage, other variables such as the permeability or the polarization are measured, always depending on the chemical composition. The main results show that the sintering by PECS / SPS method is more reducing than conventional sintering, which degrades certain properties such as the resistivity of the ferrites. The last chapter is dedicated to the realization of monolithic processors. A detailed manufacturing protocol is presented.A third part shows the operation of the realized transformers. The usual measurements allow identifying self, mutual and leakage inductances. The values of these inductances show that it is better to use components made not as a transformer, but as a coupler. Finally a converter based on a VRM structure is realized. The switching frequency is 2 MHz, the efficiency is greater than 90%, and the power density is 15 kW / liter.

No description available.

Thesis. 1976. Ph.D.--Massachusetts Institute of Technology. Dept. of Earth and Planetary Sciences. / Microfiche copy available in Archives and Science. / Vita. / Bibliography: leaves 315-334. / Ph.D.

<p>Long-term stability and chromium (Cr) contamination are two major concerns for application of chromium-bearing metallic materials as interconnects of solid oxide fuel cells (SOFCs) at intermediate temperature (~800°C). Copper-manganese (Cu-Mn) and cobalt-manganese (Co-Mn) spinel can be promising coating materials for the metallic interconnects as they show high electrical conductivities. The first objective of this research is to develop an economical and convenient method through which the spinel coatings can be applied to the metallic substrates. The investigations on the crystal structure of CuᵪMn₃₋ᵪO₄ spinel, e.g., structure symmetry and cation distributions, have always been controversial, which hinders the total understanding of the detailed structure of the material. In order to resolve the inconsistency, in-situ neutron and X-ray diffraction were employed to determine the structure of the spinel.</p> <p>A novel method was developed to obtain high quality manganese coating without any additives (sulphur or selenium compounds). Cu-Mn and Co-Mn spinel coatings were applied to metallic coupons by electrodeposition and subsequent annealing. The method is convenient and easy to control. The performance testing showed that the area specific resistances (ASRs) of the coated samples (0.003 Ω•cm²) are much lower than that of the uncoated UNS 430 (0.189 Ω•cm²) after oxidation at 750°C for 1500 hours. Moreover, both spinel coatings can effectively suppress the outward diffusion of Cr, which resulted in reduction of Cr contamination significantly. The oxidation studies of Cu-Mn coating revealed the transformation mechanisms of Cu-Mn coating to the spinel. In-situ neutron and X-ray diffraction analysis clarified the crystal symmetry of CuᵪMn₃₋ᵪO₄ spinel and CuMnO₂ at high temperatures. Rietveld refinement revealed the cation distribution of Cu and Mn ions on tetrahedral and octrahedral sites of CuᵪMn₃₋ᵪO₄ spinel, which was compared to values in the literatures. / Thesis / Doctor of Philosophy (PhD)

No description available.

<p>Solid oxide fuel cells are an efficient method of converting hydrocarbon fuels to electrical power. However, due to some problems with poisoning, these have made no headway in the energy market. The evaporation of chromium oxides from metallic current collectors causes rapid degradation of the cells on the cathode side. It has been shown that spinel coatings reduce the effects of chromium oxide evaporation. In this thesis, two spinel systems are explored for potential application. Cobalt-manganese spinel is a stable spinel which have a wide range of composition, while remaining sufficiently conductive. Copper-manganese spinel, which is much more conductive than cobalt-manganese, is slightly less stable, but nonetheless a candidate. All components of the spinels explored can be electrodeposited from aqueous solutions, at room temperature. By controlling the concentrations of metallic ions, and other additives, coatings can be deposited on interconnecting plates with reproducible results. The newly coated interconnects can be oxidized in-situ. For characterization, the samples for this thesis were oxidized at 800°C. Two substrate materials were used, ferritic stainless steel and a chromium-iron alloy. Stainless steel substrates showed good coating adhesion, but high concentrations of iron were found in the spinel structure. Chromium alloy substrates were better protected by spinel coatings. However, nitride formation at the substrate interface caused localized delamination of the coating. It was shown that plating operations can be scaled up to 10 cm by 10 cm plates, with little modification of the processes used.</p> / Master of Applied Science (MASc)

<p>The solid state dewetting characteristics of thin gold films sputtered onto (111) MgAl₂O₄ substrates were investigated. Prior research done on this system reported discovering the self-assembly of intricately shaped nanostructures, consisting of a faceted sphere lying above a truncated triangular pyramid, formed by applying a specific two stage heating profile. The current work was done to provide deeper insight on the odd self-assembly observed in this system. The results indicate that the intricate structures are not purely gold self-assemblies, but in fact consist of three distinct materials: a single crystal or polycrystalline gold faceted sphere, separated by an interfacial boundary layer, from above a crystalline MgAl₂O₄ necking structure that spontaneously developed from the initially flat substrate. The boundary separating these two assemblies is confined within a thin, sharp region of a third material consisting of Au and O elements. The composition and crystalline nature of the individual nanostructures were studied using high angle annular dark-field imaging, energy dispersive X-ray spectroscopy, and electron energy loss spectroscopy employed by a high resolution transmission electron microscope. Two-dimensional X-ray diffraction texture analysis revealed that the gold nanoparticles are crystalline, with majority of the maximum intensity signal corresponding to the epitaxial alignment of the gold nanoparticles with the substrate. The MgAl₂O₄ necking structures were found to be sensitive to both the annealing profile, and the thickness of the gold film, which influenced the distinguishable presence, size, and footprint of the MgAl₂O₄ nanostructures. Atomic force microscopy and scanning electron microscopy results were consistent with the gold overlayer playing an essential role in the self-assembly of MgAl₂O₄ nanostructures. While the fundamental mechanisms that govern this phenomenon are not entirely clear, the presented results do provide insight into the role of interfaces in heteroepitaxial systems, especially the self-assembly of crystalline nanostructures from a previously stable substrate.</p> / Master of Applied Science (MASc)

Vanadium extraction methods were developed for iron-vanadium-phosphorus (Fe-V-P) melts derived from processing of V-bearing titanomagnetites and steel slags. Studies on phase relationships of V slags were carried out to provide important understanding of the extraction processes. Phase relationship in vanadiferous slag was investigated experimentally at 1573K, 1673K and 1773K, for the compositional range of 0-6mass% Al2O3, 1-5mass% CaO, 10-17mass% SiO2, with MnO and V2O3 fixed at 5.5mass% and 30mass%, balanced with FeO. The slags were found to be solid-liquid mixtures, of liquid, spinel and in some cases free silica. Alumina was identified as the preferred additive to prohibit precipitation of free silica. A method for V extraction to vanadiferous slag was developed based on Fe-V(2mass%)-P(0.1mass%) melts at 1677K using a semi-industrial scale BOF. Oxidation was carried out with an oxygen enriched air jet and iron ore pellets. The complete dissolution of pellets was achieved by deliberately creating good stirring conditions utilising high momentary decarburisation rates. The P distribution to the slag was low when good stirring conditions was obtained. Phase relationship in Al2O3-CaO(25-35mass%)-SiO2-VOx slag was investigated experimentally at an oxygen partial pressure of 9.37•10-11atm and 1873K. The maximum solubility of V-oxide in the slag was 9-10mass% V2O3. Two solid phases were found, a solid solution of Al2O3 in V2O3 (karelianite) and hibonite with fractionation of V into the crystal structure. V extraction experiments to Al2O3-CaO-SiO2 based slags were carried out in 150kg scale by blowing CO2 gas into the metal bath consisting mainly of 1-10mass% V and 1mass% P. At these conditions, oxidation of V was favoured over Fe. Up to 10-13mass% V2O3 could be dissolved in the slag before a viscous slag saturated in V-oxide was observed. The phosphate capacity in the slag was low and as a result this slag could at once be subjected to a final reduction step for production of ferrovanadium with 40-50mass% V. / Metoder för att utvinna vanadin till högvärdiga vanadinslagger från metallsmältor innehållande främst järn (Fe), vanadin (V) och fosfor (P) utvecklades. Metallsmältorna framställs genom att processa primära V råvaror, såsom titanomagnetit, och sekundära råvaror av i huvudsak vanadinrik stålslagg. Fasstudier av högvärdiga vanadinslagger genomfördes som grund för utvecklingsarbetet. Experimentella fasstudier av vanadinspinellslagg med 30vikt% V2O3 och 5.5vikt% MnO genomfördes vid en temperatur av 1573K, 1673K och 1773K. Övriga komponenter i slaggen varierades inom ett intervall av 0-6vikt% Al2O3, 1-5vikt% CaO och 10-17vikt% SiO2, viktad med järnoxid. Samtliga slagger var sammansatt av både flytande- och fastfas. Den fasta fasen utgjordes främst av en vanadin- och järnrik spinellfas och i vissa fall även av fri SiO2. Genom försök i en stålkonverter i semi-industriell skala utvecklades och validerades en metod för vanadinutvinning från råjärnsmältor innehållande 2vikt% V och 0.1vikt% P, vid en temperatur av 1677K. Oxidationen utfördes med syreanrikad luft via en vattenkyld topplans och genom tillsats av hematit pellets. Omsättningen av pellets säkerhetsställdes genom god omrörning som erhölls under korta perioder med höga gasvolymer som en effekt av hög avkolningstakt. Råjärnet efter behandlingen innehöll cirka 3vikt% C och 0.1vikt% V. Producerad vanadinspinellslagg bestod av upp till 30vikt% V2O3. Fosforfördelningen till slaggen var låg under processbetingelser med god omrörning. Experimentella fasstudier av Al2O3-CaO(25-35vikt%)-SiO2-VOx slagg genomfördes vid en temperatur av 1873K och ett syrepartialtryck av 9.37·10-10atm. Den maximala lösligheten av vanadinoxid i slaggen var 9-10vikt% V2O3. Två fasta faser identifierades, V2O3 (Karelianit) med fast löslighet av Al2O3 och Hibonit med vanadinoxid inlöst i kristallstrukturen. Experimentella försök för att utvinna vanadin från en stålsmälta bestående av 1-10vikt% V och 1vikt% P till en slagg med en initial sammansättning av 7-40vikt% Al2O3, 25-35vikt% CaO och 27-64vikt% SiO2 utfördes i en skala av 150kg. Oxidation av vanadin åstadkoms genom att blåsa in CO2 gas i stålsmältan via en spolsten. Under dessa processförhållanden var oxidationen av vanadin gynnsam framför järn och fosfor. Lösligheten av vanadinoxid i slaggen var upp till 10-13vikt% V2O3. Slagg mättad med vanadinoxid var viskös som en konsekvens av utfällning av V2O3 med inlöst Al2O3. Slaggens gynnsamma vanadin och järn- samt vanadin och fosfor förhållande möjliggör att genom slutreduktion producera ferrovanadin med en vanadinhalt av 40-50vikt% och låg fosforhalt. / <p>QC 20170912</p>

Nanostructured fayalite (α-Fe2SiO4) with a large volume fraction of interfaces is synthesized for the first time via single-step mechanosynthesis, starting from a 2α-Fe2O3 + 2Fe + 3SiO2 mixture. The nonequilibrium state of the as-prepared silicate is characterized by the presence of deformed polyhedra in the interface/surface regions of nanoparticles. / Dieser Beitrag ist mit Zustimmung des Rechteinhabers aufgrund einer (DFG-geförderten) Allianz- bzw. Nationallizenz frei zugänglich.

Iron oxide nanoparticles (NPs) are promising materials for use in many applications, including new cancer treatments and in cleaning water, because they exhibit size-dependent magnetic and absorptive properties. NP properties are caused by structural attributes of the NPs, like surface disorder and cation vacancies. However, NP synthetic methods also impact structure, therefore properties, of NPs. Furthermore, the synthetic method is often changed in order to change the core diameter of NPs. Determining if properties are caused by the dimensions of the NP is impossible if there are also structural features present in the NP caused by the synthetic method, like grain boundaries or polycrystalline shells. In Chapter II of this dissertation, we show a new continuous growth synthesis of spinel iron oxide where the diameter of NPs is changed by the amount of precursor added to the reaction, meaning the only structural feature changing between the NPs is size. Continuous growth, therefore, can be used to probe the impact that size has on NP structure and properties. We report that saturation magnetization of NPs produced from continuous growth is size-dependent and higher in magnitude than NPs of the same core diameter made by other syntheses. In chapter III of this dissertation we determine nanoscale structure by Pair Distribution Function (PDF) analysis of Total X-ray Scattering data of NPs isolated from the reaction with core diameters between 3-10 nm. In Chapter IV of this dissertation we monitored the growth of NPs in situ with Total X-ray Scattering to gain insight on the structures of NPs while forming. In situ measurements of Total X-ray Scattering data gave insights into how precursor oxidation state influences the structures formed during formation of NPs, with more oxidized precursor giving a more oxidized product and a reduced precursor yielding a more reduced product even though the NPs formed by either method are indistinguishable by ex situ analysis. This dissertation includes previously published and unpublished co-authored material. / 2021-04-30