La transition de phase isostructurale du cérium a été et reste l'objet de nombreuses études pour tester les méthodes permettant de décrire les matériaux fortement corrélés.La Théorie du Champ Moyen Dynamique (DMFT) jointe à la Théorie de la fonctionnelle de la densité à permis de décrire de tels systèmes.Pourtant, le calcul des propriétés de l'état fondamental nécessite une très bonne précision de calcul à la fois de la part de la DFT et de la DMFT.Nous utilisons un résolveur Monte Carlo Quantique en Temps Continu (CT-QMC), rapide et capable de simuler les basses températures, combiné à une implantation ondes planes augmentées par projection de la DMFT pour calculer les énergies internes et libres -- et par conséquent l'entropie -- au cours de la transition de phase du cérium.D'importants calculs, utilisant cette implantation, nous ont permis de reconsidérer les propriétés de l'état fondamental et une grande partie de la thermodynamique de la transition de phase α<->γ du cérium à basses températures.En particulier, le bruit stochastique est suffisamment faible pour interpréter, sans ambiguïté, les courbes énergie en fonction du volume.Sur ces dernières, un double point d'inflexion est clairement visible pour l'énergie interne jusqu'à une température relativement basse.Les courbes d'énergie libre mettent, de plus, en évidence l'importance de l'entropie pour ce système.D'autre part, les spectres de photoemission tout au long de la transition de phase sont analysés.Le schéma DMFT est comparé avec des calculs DFT récents et des données expérimentales récentes.Enfin, nous mettons en avant les approximations utilisées et nous nous interrogeons sur leurs validité. / The isostructural phase transition of cerium has been and remains the aim of many studies in order to test methods developed to describe strongly correlated materials.The Dynamical Mean Field Theory (DMFT) combined with density functional theory (DFT) has been successful to describe such systems.However, the computation of the ground state properties requires a very good accuracy from both DFT and DMFT sides.We use thus a strong coupling Continuous Time Quantum Monte Carlo (CT-QMC) solver, which is fast and able to reach low temperatures, in combination with a projector augmented wave (PAW) DMFT implementation to calculate internls and free energies -- and thus the entropy -- during the phase transition of cerium.Extensive calculations using this implementation allows us to carefully reassess the ground state properties and almost all thermodynamics of the α<->γ phase transition in cerium at low temperatures.In particular, stochastic noise is small enough to avoid any ambiguity on the interpretation of energy versus volume curves.On those curves, a double inflexion point is clearly observable ont the internal energy curves untill a relatively low temperature.Moreover, free energy curves highlight the importance of including the entropy contribution.The DMFT picture is put in perspective with recent DFT calculations and recent experimental investigations.Furthermore, photoemission spectra are analysed while the phase transition.Finaly, we discuss the approximations used and raise curiosity about their consideration.

O Brasil tem a maior produção de cana-de-açúcar do mundo. O cultivo de cana-de-açúcar no Brasil está voltado principalmente para a produção de açúcar ou Etanol e nos últimos anos para a produção de bioeletricidade através da utilização da biomassa do bagaço e da palha. Apesar da importância econômica e do potencial sustentável que a cana-de-açúcar apresenta, o genoma de referência para esta cultura ainda não está disponível na literatura. A principal justificativa para isso está na complexidade do mesmo, em especial pela alopoliploidia e autopoliploidia. De fato esta característica é a principal barreira para o desenvolvimento de novas variedades comerciais. Na literatura há diferentes estratégias que visam contribuir com o conhecimento genômico de cana-de-açúcar sendo mais prevalente dados de transcriptoma e pouca informação sobre o processo de regulação gênica. Além disso, diferente do que é observado em outras culturas comerciais, em cana-de-açúcar não há trabalhos associados com a caracterização in silico da região Promotora, assim como na identificação de sítios de ligação para Fatores de Transcrição (TFBSs). Por esta razão, o nosso trabalho foi direcionado para a caracterização in silico de regiões regulatórias em cana-de-açúcar. Para esta tarefa nós realizamos apenas a rotulação de sequências de DNA não codificante que estavam a upstream de cada gene anotado em cana-de-açúcar. Todos os genes foram selecionados de dados de transcriptoma e a sequência de DNA da região Promotora foi isolada do Genespace de cana-de-açúcar SP80-3280 gerado pelo projeto de sequenciamento do genoma de referência do nosso grupo. A rotulação da região regulatória em cana-de-açúcar foi executada em duas subsequências: Core Promoter e Promotor Proximal. Na região Core Promoter nós realizamos a identificação do sítio de inicio de transcrição (TSS), a estimativa do tamanho da região 5\' UTR e a classificação da região Core Promoter em TATA-box ou TATA-less. Todos os processos foram realizados através da ferramenta TSSPlant. A utilização da ferramenta TSSPlant motivou o desenvolvimento de uma nova ferramenta para predição do sinal de TSS que aqui chamamos de TSSFinder. A ferramenta TSSinder apresentou resultados de predição do sinal de TSS superior aos seus pares, além disso esta ferramenta foi bem sucedida em diferentes organismos como Arabidopsis thaliana, Gallus gallus e Saccharomyces cerevisiae. Na região Promotora Proximal nós realizamos a identificação de TFBSs através de duas metodologias: predição de novo e mapeamento de matrizes de TFBS (PSSM). O processo de predição de novo foi realizada por meio de dois modelos: Maximização da expectativa e Gibbs Sampler e esse processo foi executado apenas para o subgrupo de genes co-expressos ou apenas para o conjunto de sequências homeólogas de cada gene de cana-de-açúcar selecionado. Para o restante das sequências foi realizado apenas o mapeamento das matrizes de TFBSs identificadas durante o processo de predição de novo. Em paralelo todos TFBSs identificados no nosso trabalho foram comparados com o banco de TFBS para plantas. Através desse procedimento foi possível estimar qual classe de Fator de Transcrição está interagindo com o TFBS identificado na região Promotora Proximal dos genes Scdr1, ScSuSy, ScPAL. Com este trabalho, nós cobrimos parte da lacuna observada em estudos in silico paras regiões regulatórias de cana-de-açúcar. Além disso, nós aperfeiçoamos o processo de identificação do sinal de TSS para diferentes organismos; inclusive para plantas Dicotiledôneas e Monocotiledôneas. / Brazil has the highest production of sugarcane in the world. Its cultivation in Brazil is aimed at producing of sugar or ethanol and in recent years, biomass for bioenergy from bagasse and straw. Despite the economic importance and the sustainable potential that sugarcane presents, a reference genome for this crop is not yet available in the literature. One justification for this absence lies in the sugarcane genome complexity, allopolyploidy and autopolyploidy. In fact these characteristics are the main barrier for the development of new commercial varieties. In the literature different strategies aimed at contributing to genomic sugarcane mostly on the transcriptome and little information on the process of gene regulation. Furthermore, unlike other commercial crops, sugarcane has no reported in silico characterization of its promoter regions and identification of Transcription Factor binding sites. For this reason, our work was directed to an in silico characterization of regulatory regions in sugarcane. For this task we performed the labeling of non-coding DNA sequences that were upstream of each gene annotated in sugarcane. All genes were using from transcriptome data and the promoter region DNA sequence was isolated from Genespace of the SP80-3280 reference genome obtained of our group. The labeling of the regulatory region in sugarcane was carried out in two subsections: Core Promoter and Proximal Promoter. In the Core Promoter region we performed the identification of the TSS signal, the estimation of the size of the 5\' UTR region and the classification of the Core Promoter region in TATA-box or TATA-less. All processes were performed using the TSSPlant tool. The use of the TSSPlant tool motivated the development of a new tool to predict the TSS signal that we call TSSFinder. The TSSinder tool presented TSS signal prediction results superior to its peers, moreover this tool was successful in different organisms - Arabidopsis thaliana, Gallus gallus and Saccharomyces cerevisiae. In the Proximal Promoter region we performed the identification of TFBSs through two methodologies: de novo prediction and mapping of TFBS matrices (PSSM). The de novo prediction process was performed using two models: Expectancy Maximization and Gibbs Sampler and this process was performed only for subgroups of coexpressed genes or only for the set of homeologues sequences from each sugarcane gene. For the rest of the sequences only the mapping of the matrices of TFBSs identified during the de novo prediction process was conducted. In parallel all TFBSs identified in our work were compared with the TFBS database for plants. Through this procedure it was estimated which class of Transcription Factor is interacting with the TFBS identified in the Proximal Promoter region of the Scdr1, ScSuSy, ScPAL genes.With this work, we cover part of the gap observed in in silico studies for the regulatory region of sugarcane. In addition, we improved the process of identification the TSS signal for different organisms including dicotyledonous and monocotyledonous plants.

Kohlenstoffnanoröhrchen (engl. carbon nanotube, CNT) sind vielversprechende Kandidaten für den Ersatz von Kupferleitbahnen die bei weiterer Strukturverkleinerung von integrierten Schaltkreisen notwendig wird. In dieser Arbeit wird mit Hilfe von ab-initio Simulationen auf Basis der Dichtefunktionaltheorie die elektronische Struktur von halbleitenden CNTs beispielhaft anhand des (8,4)-CNTs untersucht. Nach Besetzung des CNT mit Metallatomen, hier Kobalt, zeigen sich massive Änderungen der Bandstruktur. Es reichen bereits überraschend kleine Mengen des Metalls aus, um einen starken Effekt zu erreichen. Die Änderungen der elektronischen Struktur sind stark abhängig von der genauen Position der Metallatome relativ zum Kohlenstoffgerüst der CNTs, der Einfluss der mechanischen Verformung des CNTs als Reaktion auf die Anlagerung ist hingegen sehr gering. Die relevanten Bänder der Kobaltatome liegen leicht unterhalt der Fermi-Energie und sorgen bei der Integration in die Bandstruktur des CNTs für die Schließung der Bandlücke und somit für die Transformation eines vorher halbleitenden CNTs in ein leitendes. Diese Transformation konnte auch mit Simulationsrechnungen zum Elektronentransport bestätigt werden. Ferner wurden bei weiteren Rechnungen eine ausgeprägte Spinabhängigkeit der Bandstruktur ermittelt, welche noch weiterer Untersuchung bedarf.

Im Rahmen der vorliegenden Arbeit wird eine Tight-Binding-Formulierung der Korringa-Kohn-Rostoker-Greenschen-Funktionsmethode vorgestellt. Dabei werden mittels eines geeignet gewählten Referenzsystems abgeschirmte Strukturkonstanten konstruiert. Es werden die Vorteile und Grenzen dieser Transformation des Formalismus diskutiert. Es wird gezeigt, daß der numerische Aufwand zur erechnung der Elektronenstruktur von Systemen mit langgestreckter Elementarzelle linear mit der Systemgröße wächst. Damit ist eine Behandlung von Systemen mit 500 und mehr Atomen pro Elementarzelle möglich. Anhand von umfangreichen Testrechnungen kann demonstriert werden, daß das neue Verfahren bezüglich seiner Genauigkeit mit dem traditionellen KKR-Verfahren vergleichbar ist. Es werden Anwendungen zur Berechnung der Elektronenstruktur sowie zur Zwischenlagenaustauschkopplung von Co/Cu(100)-Vielfachschichten vorgestellt. / A newly developed ab initio tight-binding-formulation of the Korringa-Kohn-Rostoker-Green's function method for layered systems is presented. Screened structure constants are calculated by means of a repulsive reference system. Advantages and limits of this transformation of the formalism are discussed in detail. The numerical effort for self consistent electronic structure calculations of systems with a large prolonged supercell scales linearly with the system size. Systems with up to 500 atoms per unit cell can be treated easily. The accuracy of the new method is of the same order as the traditional KKR method. Applications to electronic structure calculations and magnetic interlayer exchange coupling in Co/Cu(100) multilayers are presented.

No description available.

No description available.

La présente thèse porte sur l'utilité de la théorie de la fonctionnelle de la densité dans le design de polymères pour applications photovoltaïques. L'étude porte d'abord sur le rôle des calculs théoriques pour la caractérisation des polymères dans le cadre de collaborations entre la théorie et l'expérience. La stabilité et les niveaux énergétiques de certaines molécules organiques sont étudiés avant et après la sulfuration de leurs groupements carbonyles, un procédé destiné à diminuer le band gap. Les propriétés de dynamique électronique, de séparation des porteurs de charges et de spectres de vibrations Raman sont également explorées dans un polymère à base de polycarbazole. Par la suite, l'utilité des calculs théoriques dans le design de polymères avant leurs synthèses est considérée. La théorie de la fonctionnelle de la densité est étudiée dans le cadre du modèle de Scharber afin de prédire l'efficacité des cellules solaires organiques. Une nouvelle méthode de design de polymères à faible band gaps, basée sur la forme structurale aromatique ou quinoide est également présentée, dont l'efficacité surpasse l'approche actuelle de donneur-accepteur. Ces études sont mises à profit dans l'exploration de l'espace moléculaire et plusieurs candidats de polymères aux propriétés électroniques intéressantes sont présentés. / This thesis focuses on the role of density functional theory in the design of polymers for photovoltaic applications. Theoretical calculations are first studied in the characterization of polymers in the context of collaborations between theory and experiment. The stability and the energy levels of some organic molecules are studied before and after a sulfurization of their carbonyl groups, a process destined to lower the band gaps. The dynamics of the electronic processes and the Raman vibration spectra are also explored in a polycarbazole-based polymer. From then, the usefulness of theoretical calculations in the design of polymers before their syntheses is explored. Density functional theory calculations are studied under the Scharber model in order to predict the efficiency of organic solar cells. Then, a new approach for the design of low band gap polymer based on the aromatic or quinoid structures is established, whose efficiency surpasses the actual donor-acceptor approach. These studies are used in the exploration of the chemical space and several candidate for polymers with interesting electronic properties are presented.

In der vorliegenden Diplomarbeit wurden sekundäre Isotopeneffekte auf NMR-chemische Verschiebungen mit ab-initio quantenchemischen Methoden bestimmt. Um die zu erwartenden Fehler verschiedener Methoden abschätzen zu können, wurden die sekundären Isotopeneffekte von Fluor(2-2H1)ethan auf HF, MP2 und CCSD(T)-Niveau mit tripel- und quadrupel-Zeta Basissätzen berechnet. Anhand der Ergebnisse dieser Rechnungen lässt sich zeigen, dass gauche und antiperiplanare Konformere bereits mithilfe von sekundären Isotopenverschiebungen unterschieden werden können, die auf HF/tz2p-Niveau berechnet wurden. Weiterhin wurde untersucht, ob es eine Abhängingkeit des Isotopeneffektes von den Diederwinkeln zwischen Deuterium- und Fluorsubstituenten gibt. Die sekundären Isotopeneffekte auf 19F-NMR-chemische Verschiebungen von exo- und endo-2-Fluornorbornanen mit Deuterium Substituenten in den endo-3, exo-3, syn-7 oder anti-7 Positionen wurden auf HF/tz2p-Niveau berechnet. Es wird gezeigt, dass die berechneten Isotopeneffekte an 2-Fluornorbornanen eine Identifizierung der verschiedenen Stereoisomere erlauben. / In the present Diploma thesis secondary isotope effects on NMR chemical shieldings were determined using ab-initio quantum chemical methods. In order to estimate errors of various methods, secondary isotope effects on fluoro(2-2H1)ethane were calculated at the HF, MP2 and CCSD(T) level of theory using triple- and quadruple zeta basis sets. On the basis of these calculations it can be shown that gauche and antiperiplanar conformers can already be distinguished by their secondary isotope shifts calculated at the HF/tz2p level of theory. Furthermore it was investigated, whether a dependency of the isotope effects on dihedral angles between the deuterium and the fluorine substituent exists. The secondary isotope effects on 19F chemical shifts of exo- and endo-2-fluoronorbornanes with deuterium subsituents in the endo-3, exo-3, syn-7 or anti-7 positions were calculated at the HF/tz2p level. It is shown that the calculated isotope effects of 2-fluoronorbornanes allow to identify various stereoisomers.

Ce travail porte sur la synthèse et la caractérisation de nouveaux matériaux d’électrodes positives pour batteries au lithium. Nos recherches se sont principalement orientées vers les matériaux de type phosphates de métaux de transition, et notamment vers la famille des tavorites de composition (Li,H)FePO4(OH), qui présente une structure tridimensionnelle comportant plusieurs types de tunnels propices à l’insertion d’ions lithium. La structure du matériau LiFePO4(OH) a ainsi été parfaitement résolue, de même que celle du matériau FePO4.H2O, qui est un nouveau phosphate de fer (III) découvert au cours de ces travaux. Ces deux matériaux, ainsi que ceux obtenus par traitement thermique de la phase FePO4.H2O, ont été caractérisés à l’aide de différentes techniques d’analyse physico-chimiques. Leur comportement électrochimique vis-à-vis de l’intercalation / désintercalation du lithium a été étudié, ainsi que les mécanismes redox et structuraux associés mis en jeu. / This work deals with the synthesis and characterization of new positive electrode materials for lithium batteries. Our researches were mainly focused on phosphates of transition metals, and especially on the tavorite-type materials of composition (H,Li)FePO4(OH). Their structure is characterized by a three-dimensional network with different types of tunnels, which can host inserted lithium ions. In this context, LiFePO4(OH) structure was perfectly solved, as well as that of FePO4.H2O, which is a new iron (III) phosphate discovered during this work. These two materials, together with those obtained by heat-treatment of FePO4.H2O, were characterized using different analytical techniques. Their electrochemical behavior toward intercalation / deintercalation of lithium was also studied, as well as the structural and redox processes involved.

La biomasse ligno-cellulosique est principalement constituée de cellulose, d’hémicellulose et de lignine. Par conversion thermochimique, la lignine se transforme en bio-huile riche en composés phénoliques. Ces composés phénoliques sont généralement récupérés à l’aide de plusieurs étapes d‘extractions liquide-liquide consécutives impliquant des solvants aqueux et organiques. À l’aide d’une approche multi-échelle, nous avons étudié, la faisabilité et l’efficacité d’un liquide ionique, la [Choline][NTf2], pour l’extraction liquide-liquide de ces composés. En effet, l’utilisation de ce solvant permettrait d’améliorer l’efficacité d’extraction, tout en diminuant la toxicité et les coûts liés à l’utilisation de solvants organiques classiques. Des calculs quantiques ont été effectués afin de mieux comprendre les interactions régissant les systèmes clés de ces extractions. Les structures des composés phénoliques dans leur état isolé ou en présence de solvant montrent que les conformations sont stabilisées par la présence de liaisons hydrogène. De plus, la détermination des énergies d’interaction indique que la [Choline][NTf2] est efficace pour l’extraction de composés phénoliques présents en solution aqueuse. Par ailleurs, la détermination des diagrammes de phases des systèmes binaires {eau-composé + phénolique} et des systèmes ternaires {eau + composé phénolique + [Choline][NTf2]} montre que l’extraction des composés est également possible à une échelle macroscopique. Les paramètres NRTL issus de ces expériences ont permis de simuler l’extraction de trois composés phénoliques majoritairement présents dans les bio-huiles, i.e. le phénol, le guaiacol et le syringol, et ce, à moindre coût. Enfin, l’extraction de ces composés à partir d’une bio-huile obtenue par pyrolyse rapide de lignine a également été particulièrement efficace avec le liquide ionique [Choline][NTf2]. De ce fait, cette étude multi-échelle a permis de montrer que la [Choline][NTf2] est un excellent solvant pour la récupération des composés phénoliques. Enfin, l’étude des propriétés anti-oxydantes témoigne de la valeur ajoutée de ces composés, notamment à travers leur pouvoir réducteur et leur propriété anti-radicalaire / The lignocellulosic biomass is mostly composed of cellulose, hemicellulose and lignin. Upon thermal conversion of lignin, a bio-oil rich in phenolic compounds is obtained. These latter are then generally recovered through several liquid-liquid extraction involving aqueous and organic solvents. In this work we investigated, by a multi-scale study, the feasibility and the efficiency of the ionic liquid [Choline][NTf2] for the extraction of these phenolic compounds by liquid-liquid extraction. Indeed, such a solvent could improve the extraction efficiency and at the same time, reduce the toxicity and the cost of the classic organic solvents. Quantum calculations were performed in order to better understand the interaction governing the key systems of these extractions. The structures of the phenolic compounds in their isolated forms and in contact with solvents show that the conformations are stabilized by the presence of hydrogen bonds. More, the determination of the interaction energies indicates that the [Choline][NTf2] ionic liquid is efficient for the extraction of phenolic compounds present in aqueous solution. Furthermore, the study of phase diagrams of binary systems {water-phenolic compound} and ternary systems {water-phenolic compound-[Choline][NTf2]} show that the extraction of these compounds is also possible at a macroscopic scale. Then, the NRTL parameters coming from these experiments allowed the simulation of the extraction of three compounds mostly present in the bio-oils, namely phenol, guaiacol and syringol, as so at low cost. Finally, the extraction of these compounds from bio-oil obtained from lignin fast pyrolysis was also particularly efficient with the [Choline][NTf2] ionic liquid. Therefore, this multi-scale study demonstrated that [Choline][NTf2] is an excellent solvent for the recovery of phenolic compounds. Lastly, the study of the antioxidant properties testify the added-value of these compounds, especially through their reducing power and their radical scavenging capacity