Submitted by MARCIA ROVADOSCHI (marciar@unifra.br) on 2018-08-16T18:21:26Z No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Dissertacao_MarianaZancanTonel.pdf: 3760171 bytes, checksum: 30f56b8eeff4822bbb2bcb24f32cf7be (MD5) / Made available in DSpace on 2018-08-16T18:21:26Z (GMT). No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Dissertacao_MarianaZancanTonel.pdf: 3760171 bytes, checksum: 30f56b8eeff4822bbb2bcb24f32cf7be (MD5) Previous issue date: 2013-08-30 / Graphene has attracted great interest from the scientific community since it was demonstrated its stability in 2004. Because of their structural and electronic characteristics, offers a wide variety of applications, such as in the electronics industry and medicine. Graphene is a single layer of graphite in which carbon atoms are arranged in hexagonal form. Meanwhile, doxorubicin (DOX) is a drug widely used for the treatment of various cancers. However, this molecule has a high toxicity mainly in the cardiovascular system. Recent experimental studies have demonstrated that the combination of doxorubicin and graphene can kill malignant cells selectively. This process is also associated with ambient temperatures due to the ability of carbon nanostructures such as graphene and carbon nanotube that can convert infrared radiation into vibrational energy generating heat and causing apoptosis of the malign cells. In this work, we present the structural and electronic properties of graphene, pure and vacancy-type defects, interacting with the DOX molecule by density functional theory (DFT) making use of computer code SIESTA. The results showed that DOX interacting with the pure graphene has a biding energy in the range from 0.88 to 1.09 eV (LDA and GGA, respectively) and the graphene vacancy type defect binding energy is about 1.09 to 0.84 eV (LDA and GGA, respectively). Likewise, it is observed that no major changes in the original electronic properties of graphene. Also, in this work was analyzed by ab initio molecular dynamics (MD-AI) position of the molecule DOX against the graphene temperature variation and it is noted that the largest spacing occurs at 300K for both as pure graphene for graphene with vacancy-type defect interacting with DOX. / O grafeno vem atraindo grande interesse da comunidade científica desde que foi demonstrada a sua estabilidade em 2004. Devido as suas características estruturais e eletrônicas, oferece uma grande variedade de aplicações, tais como na indústria eletrônica e na medicina. O grafeno é uma única camada de grafite na qual os átomos de carbono estão arranjados de forma hexagonal. Paralelamente, a doxorrubicina (DOX) é um fármaco amplamente utilizado para o tratamento de diversos tipos de câncer. No entanto, essa molécula possui alta toxicidade principalmente no sistema cardiovascular. Estudos experimentais recentes têm demonstrado que a combinação de grafeno e DOX pode causar a morte seletiva de células malignas. Este processo também está associado com temperatura ambiente, pois nanoestruturas de carbono como grafeno e nanotubo de carbono podem converter a radiação infravermelha em energia vibracional gerando calor causando a apoptose de células malignas. Neste trabalho apresentaremos as propriedades estruturais e eletrônicas do grafeno, puro e com defeito do tipo vacância, interagindo com a molécula de DOX por meio da teoria do funcional de densidade (DFT) fazendo uso do código computacional SIESTA. Os resultados mostram que a DOX interagindo com o grafeno puro tem uma energia de ligação na ordem de 0,88 e 1,09eV (LDA e GGA, respectivamente) e no grafeno com defeito do tipo vacância a energia de ligação é da ordem de 1,09 e 0,84eV (LDA e GGA, respectivamente). Da mesma forma, observa-se que não ocorrem alterações significativas nas propriedades eletrônicas originais do grafeno. Também, neste trabalho, foi analisada através da dinâmica ab initio (DM-AI) a posição da molécula de DOX em relação ao grafeno com a alteração de temperatura e observou-se que o maior afastamento ocorreu à 300K tanto para o grafeno puro como para o grafeno com defeito tipo vacância interagindo com a DOX.

Submitted by MARCIA ROVADOSCHI (marciar@unifra.br) on 2018-08-16T20:05:43Z No. of bitstreams: 2 Dissertação_LauraFernandaOsmariVendrame.pdf: 3061643 bytes, checksum: d06cb3bc35f9d2bedf0c39b9c8e35276 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2018-08-16T20:05:43Z (GMT). No. of bitstreams: 2 Dissertação_LauraFernandaOsmariVendrame.pdf: 3061643 bytes, checksum: d06cb3bc35f9d2bedf0c39b9c8e35276 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2015-03-31 / The use of C60 fullerenes, functionalized or in pristine form, as chemical and biological sensors has become a large field of study and application High stability and low chemical reactivity of these nanostructures provide some difficulties in studying the chemical and physical properties. These difficulties can be overcome through the functionalization process. The purpose of this work is to evaluate through abs initio calculations the structural and electronic properties of the temozolamide molecule interacting with functionalized fullerenes. This drug is very important for the treatment of brain tumors. This study is based on the density functional theory, using the SIESTA computational code. First of all, it was evaluated the isolated functionalized fullerenes behavior, as well as the temozolamide molecule. Subsequently, it was studied the interaction of functionalized fullerenes with temozolamide in order to understand the energetic and structural properties of these structures to support the use of these nanomaterials in future biomedical applications. From the results it is possible to observe the depending on the studied configuration, the binding energy values are not the same. It was found a weak interaction for all functionalized C60 interacting with temozolamide through physical adsorption, with values between 0.43 eV and 1.02 eV for all stable configurations. An interaction through physical adsorption was also observed for themozolamide with pristine fullerenes, with energy value of 0.23 eV. Despite some high energy values, the distance in the interaction between the nearest atoms of the nanostructure and temozolomide does not represent chemical bonds. The interaction of the temozolamide drug with fullerenes is important for the future development of central nervous system nanodrugs improving the performance and slowing the elimination of the drug by the body. / O uso de fulerenos C60, funcionalizados ou na forma pura, como carreadores de fármacos tem se tornado um grande campo de estudo e aplicação. A alta estabilidade e baixa reatividade química destas nanoestruturas trazem algumas dificuldades de manipulação no estudo de suas propriedades químicas e físicas. Estas dificuldades podem ser superadas através do processo de funcionalização. O objetivo deste trabalho é avaliar, por meio de cálculos ab initio as propriedades estruturais e eletrônicas da molécula de temozolamida interagindo com fulerenos puros e funcionalizados. A temozolamida é um medicamento muito importante no tratamento de tumores cerebrais. Este estudo é baseado na teoria do funcional da densidade, utilizando o código computacional SIESTA. Primeiramente, estudamos os fulerenos funcionalizados isolados, bem como a molécula de temozolamida. Posteriormente, estudamos a interação do fulereno puro e funcionalizado, com a temozolamida a fim de compreender suas propriedades energéticas e estruturais. A partir dos resultados é possível observar que, dependendo da configuração estudada, os valores de energia de ligação não são os mesmos. Obtivemos uma interação fraca para todos os C60 funcionalizados interagindo com a temozolamida via adsorção física, com valores entre 0,43 eV e 1,02 eV para as configurações mais estáveis. Uma interação via adsorção física tambem foi observada para a interação da temozolamida com fulerenos puros, com energia de 0,23 eV Apesar de alguns valores altos de energia, a distância na interação entre os átomos mais próximos entre a nanoestrutura e a temozolamida e a baixa transferência de carga, não representam características de ligações covalentes. A interação da temozolamida com fulerenos, através da funcionalização, neste estudo, é importante para contribuir com o futuro desenvolvimento de nanomedicamentos do sistema nervoso central melhorando o desempenho e reatardando a eliminação do medicamento pelo organismo.

Conselho Nacional de Desenvolvimento CientÃfico e TecnolÃgico / Nesta dissertaÃÃo, a adsorÃÃo de trihalometanos (THMs) em nanotubos de carbono de parede simples à estudada por meio de cÃlculos de primeiros princÃpios baseados na Teoria do Funcional da Densidade (DFT), implementada no cÃdigo computacional SIESTA que faz uso de pseudopotenciais para descrever os elÃtrons de caroÃo e bases localizadas para expandir as funÃÃes de onda dos elÃtrons de valÃncia. A interaÃÃo do CHCl3 (o mais importante e abundante dos THMs) com nanotubos puros, com vacÃncias e carboxilados à simulada a fim de se entender o processo de interaÃÃo em diferentes geometrias e encontrar as melhores rotas para a possÃvel utilizaÃÃo de nanotubos como filtros. Em todos os sistemas estudados foram utilizados o tubo metÃlico (5,5) e o semicondutor (8,0). Um estudo complementar com THMs bromados (CHCl2Br, CHClBr2 e CHBr3) tambÃm foi realizado. Com estes Ãltimos cÃlculos, busca-se observar a influÃncia do nÃmero de Ãtomos de bromo do THM sobre a interaÃÃo com os nanotubos puros. AtravÃs da anÃlise de propriedades estruturais, eletrÃnicas e energÃticas, à verificado que os THMs sÃo adsorvidos fisicamente na superfÃcie dos nanotubos (energias de ligaÃÃo variando de -0,06 eV a -0,21 eV) em todos os sistemas, tornando-os candidatos para se extrair THMs de soluÃÃes aquosas. Os cÃlculos prevÃem ainda a estabilidade de complexos formados por grupos quÃmicos (como COCCl3 e COOCCl3 oriundos de possÃveis reaÃÃes quÃmicas do CHCl3 com grupos carboxila) agregados na superfÃcie dos tubos. A energia de ligaÃÃo de tais grupos com os nanotubos variou de -1,31 eV atà -2,26 eV. / In this work we study the interaction of thihalomethanes (THM) with single wall carbon nanotubes using first principles calculations based on the Density Functional Theory (DFT), implemented in the SIESTA code which uses pseudopotentials to describe core electrons and localized basis to expand the valence wavefunctions. The adsorption of CHCl3 (the most important and abundant THM) on pristine, vacant and carboxylated carbon nanotubes is simulated aiming to understand the interaction process in different geometries and pursuing the best route for the potential application of carbon nanotubes as filters. The metallic (5,5) and semiconducting (8,0) nanotubes are used in all calculations. A complementary study involving bromated THMs (CHCl2Br, CHClBr2 and CHBr3) is also performed in order to investigate the influence of the bromine atoms on the interaction between THMs and pristine nanotubes. Based on the analysis of structural, electronic and energetic properties it is verified that THMs are adsorbed on the carbon nanotube surface through a physisorption process (binding energies in the range -0,06 eV to -0,21 eV) in all cases. This fact makes carbon nanotubes as promissing candidates to extract THMs from aqueous solutions. The calculations also predict the stability of chemical groups (COCCl3 and COOCCl3) bounded on the nanotube surface. These chemical groups can be derived from possible chemical reactions, through coupling agents, between CHCl3 and the carboxyl groups anchored to the nanotube wall. The binding energies of these groups and the nanotubes are found to be in the range -1,31 eV to -2,26 eV.

Orientador: Edison Zacarias da Silva / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Física Gleb Wataghin / Made available in DSpace on 2018-08-21T01:33:25Z (GMT). No. of bitstreams: 1 Faccin_GiovaniManzeppi_D.pdf: 8582909 bytes, checksum: f53f27084aaa92a73ab76984290715ff (MD5) Previous issue date: 2012 / Resumo: Esta tese apresenta um estudo baseado em simulação computacional de propriedades físicas de nanopartículas de ferro, cobalto, ítrio e compostos híbridos destes elementos visando mapear as estruturas mais estáveis destas nanopartículas e seu comportamento magnético. Para este fim, empregamos cálculos de primeiros princípios baseados no formalismo da teoria do funcional da densidade aliados a uma estratégia heurística visando mapear possíveis isômeros de cada nanopartícula e, posteriormente, caracterizar suas propriedades magnéticas. Os principais resultados encontrados são o mapa estrutural dos isômeros mais prováveis e seu o mapa de magnetização, os quais podem servir de guia para a síntese direcionada de novos materiais com propriedades físicas e químicas de interesse / Abstract: This thesis presents a computational study regarding physical properties of iron, cobalt and yttrium nanoparticles as well as hybrid composites of these elements aiming to map their most stable structures and their magnetic behavior. To this end, we have used first principles calculations based on the density functional theory formalism coupled with a heuristic strategy aiming to map possible isomers of each nanoparticle and also characterize their magnetic properties. The main results found are a map of the structures of the most likely isomers and also their magnetization map, which can be used as a guide for the directed synthesis of new materials that have physical and chemical properties of interest / Doutorado / Física / Doutor em Ciências

Orientador: Iris Concepcion Lineares de Torriani / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Fisica "Gleb Wataghin" / Made available in DSpace on 2018-08-04T15:02:02Z (GMT). No. of bitstreams: 1 Oliveira_CristianoLuisPintode_D.pdf: 31759545 bytes, checksum: a92d80aac97e4e41bdf268c343cbd43e (MD5) Previous issue date: 2005 / Resumo: Apresentamos nesta tese estudos relacionados à obtenção de parâmetros estruturais e modelagem ab initio para macromoléculas biológicas e sistemas densos de partículas elipsoidais a partir de dados de espalhamento de raios X a baixos ângulos. Aspectos teóricos e experimentais são discutidos, complementados por simulações computacionais e pela apresentação de resultados experimentais que exemplificam a aplicação dos métodos de cálculo revelando assim os alcances e limites da técnica. Progressos recentes em biologia molecular estrutural revelam que a correlação entre a forma e a função de macromoléculas biológicas é crucial para a compreensão dos processos biológicos. Além disso, diversos estudos associam várias doenças a defeitos estruturais em proteínas, aumentando ainda mais o interesse na determinação das estruturas protéicas. A técnica de espalhamento de raios X a baixos ângulos permite a obtenção de parâmetros estruturais de proteínas, apresentando diversas vantagens sobre a cristalografia de proteínas e a ressonância magnética nuclear, mesmo não atingindo alta resolução. Trataremos, portanto, dois assuntos principais: estudos de sistemas diluídos e estudos de sistemas densos. Para o estudo de sistemas diluídos o enfoque principal será na modelagem estrutural de proteínas em solução, onde são apresentados e discutidos diversos métodos de simulação tanto da estrutura das macromoléculas quanto de seus parâmetros hidrodinâmicos. Estes métodos foram aplicados ao estudo de diversos problemas, tais como: (i) as estruturas das hemoglobinas extracelulares de Biomphalaria glabrata e Glossoscolex paulistus, proteínas gigantes com elevada cooperatividade funcional entre suas subunidades; (ii) as diferenças estruturais entre as formas livres e ligadas a lipídeos da proteína básica de mielina, uma das proteínas principais do axônio nervoso e cujo mal funcionamento leva a doenças neuronais; (iii) a interação entre proteínas e polímeros, tema relevante para processos de cristalização e construção de sistemas bio-poliméricos; e (iv) a estrutura da proteína precursora amilóide, proteína transmembrana que, apesar de diversas funções conhecidas, é uma das precursoras principais do Mal de Alzheimer. Já no estudo de sistemas densos é apresentada uma nova metodologia baseada em simulações de Monte Carlo para elipsóides de revolução. Como elipsóides podem servir de sistemas modelos para diversos sistemas reais, simulações de propriedades de soluções concentradas destes sistemas são de grande interesse. Com base nestas simulações, funções de espalhamento e funções de correlação puderam ser calculadas para um grande intervalo de anisotropias e concentrações, permitindo a obtenção de funções numéricas que podem ser utilizadas no estudo de sistemas reais uma vez que não existem expressões analíticas para este tipo de sistema. Como aplicação, este método foi utilizado no estudo de nanopartículas de ferro em matriz de silício, fornecendo parâmetros estruturais do sistema. / Abstract: In this dissertation, theoretical and experimental aspects of the calculation of structural parameters and ab-initio modeling for biological macromolecules and condensed particle systems are being discussed. Two main applications will be presented, complemented by computational simulations and many experimental results. For dilute systems the focus will be on the structural modeling of several proteins in solutions, including the calculation of hydrodynamic parameters. Recent developments in structural molecular biology reveal that knowledge about the correlation existing between the structure and function of biological macromolecules is crucial for the understanding of biological processes. Many studies have proved that structural defects detected in certain proteins are responsible for serious diseases. Small angle X-ray scattering has proved to be very useful in the determination of the low resolution structure of macromolecules in solution. This technique offers clear advantages with respect to other techniques, like singlecrystal X-ray crystallography and nuclear magnetic resonance. The low resolution structural studies of proteins included: (i) the 3D solution structure of the extracellular hemoglobin from the species Biomphalaria glabrata and Glossoscolex paulistus, giant proteins with high functional cooperativity between their subunits; (ii) the determination of the structural differences between the lipid free and lipid bound forms of myelin basic protein (MBP), one of the most important proteins of the axon sheath whose bad functioning leads to neural diseases; (iii) the study of protein-polymer interactions, relevant subject for crystallization processes and development of bio-polymeric systems; (iv) the 3D solution structure of the amyloid precursor protein, APP, transmembrane protein which is related to the Alzheimer disease. For the study of dense systems, a Monte Carlo based method for the simulation of a system of hard ellipsoids of revolution was developed. Ellipsoidal particles can be used as model systems for many real problems and simulations of the properties of concentrated solutions are of great interest. As a result of these simulations, scattering functions and correlation functions could be derived for a wide interval of anisotropies, permitting the calculation of numerical functions that can be used in real systems. As an example, this method was used for the study of iron nanoparticles in a silica matrix, leading to the determination of mean radius and size distribution of the particles. / Doutorado / Física da Matéria Condensada / Doutor em Ciências

L'objectif de la thèse est d'identifier et de quantifier les paramètres régissant l'épaississement des macles dans trois métaux hexagonaux (Zr, Ti, Mg). Le mécanisme que nous étudions est le glissement de dislocations d'interfaces, i.e. les disconnections, le long des joints de macles parfaits. Nous nous intéressons alors à la germination des disconnections avant de nous concentrer sur leur migration.Une première étude en potentiel empirique permet de valider un couplage original avec la théorieélastique afin d'extraire des simulations atomiques l'énergie de cœur des disconnections. Cette méthode permet de partitionner l'énergie entre une contribution de cœur, intrinsèque à la disconnection et une partie élastique qui dépend de l'environnement de la disconnection. Cette partition faite, nous modélisons alors l'énergie de formation de dipôles de disconnections isolés, pour chacun des différents joints de macles. Cette modélisation nous permet alors de sélectionner les dipôles de disconnections les plus pertinents que nous étudions par la méthode abinitio.Nous modélisons ensuite la migration des disconnections le long des joints de macles parfaits.Pour cela, nous montrons via l'utilisation de la méthode Nudged Elastic Band, que l'énergie demigration est d'un ordre de grandeur inférieur à l'énergie de formation. Ainsi, l'énergie de formation des dipôles de disconnections apparait comme prédominant dans la croissance des macles / The aim of this thesis is to identify and quantify the parameters of importance when dealing with twin thickening in three hcp metals (Zr, Ti, Mg). The mechanism we study is the glide of twinning dislocations, i. e. the disconnection, along the perfect twin boundaries. We first focus on the nucleation of disconnection before addressing their migration. A study using an EAM potential allows us to validate an original coupling between our atomistic simulations and linear elasticity in order to extract the core energy of disconnection.We then show how this coupling permits to divide the formation energy in two terms: the core contribution, intrinsec to the disconnections, and an elastic one, which depends on the disconnection's environment. Thanks to this partition, we model the formation of isolated disconnection that may appear along the different twin planes. We select the dipoles of lowest formation energies in order to perform ab initio calculations and compare the behavior observed the three different metals. We then model the migration of disconnections along the perfect twin planes. To do so, we use the Nudged Elastic Band method, and find out that the migration energy of disconnections is one order of magnitude lower than their formation energies. We therefor conclude that the disconnection nucleation is the rate limiting factor to explain twin thickening thanks to the creation and motion of disconnection

La combinaison de la spintronique et de I' électronique organique est censée conduire à une nouvelle gamme d'applications en la domaine de spintronique organique. Ce travail se concentre sur la physique des molécules organiques adsorbées sur des surfaces métalliques, en mettant I' accent sur les changements de leurs structures électroniques en raison de I' effet important du substrat métallique. Nous avons étudié le complexe Ni2 adsorbé sur le Cu qui montre un effet Kondo à basses températures. Nous avons ensuite étudié des chaînes d'oligoacènes (anneaux de benzène fusionnè) qui présentent une propriété prédite: une oscillation de la bande interdite d' énergies en fonction de la longueur de la molécule. Cette propriété peut survivre même lorsque la molécule est adsorbée sur une surface, comme I' Au ou le Si02. Pour terminer, nous avons étudié des petites molécules, telles que l'éthane ou I'éthylène, adsorbées sur un subtrat de Cu et de Co. La différence d' énergie entre I' orbitale vide la plus basse (LUMO) et I' orbitale moléculaire occupée la plus élevée (HOMO) est étudiée dans les approximations de la théorie de la fonctionnelle de la densité (DFT) et la méthode GW. / The combination of spintronics and organic electronics, is believed to lead to a new generation of spin based devices, which would likely open a new broad range of applications in the field of organic spintronics. ln this context, this work focuses on organic molecules adsorbed on metallic surfaces and their electronic structure changes due to the important screening of the metallic environment. We have studied different systems, starting with a Ni2 complex adsorbed on Cu (001), where Kondo effect sets in. This study is followed by oligoacene chains (fused benzene rings) where an extraordinary property is seen: an oscillation of the energy band gap with respect to the molecular length. This property is proved to survive under the screening of surfaces such as Au(111) and Si02. We finally focused on small molecules, like ethane and ethylene, adsorbed on Cu and Co. The difference of energy between the lowest unoccupied molecular orbital (LUMO) and the highest occupied molecular orbital (HOMO) is studied using density functional theory (DFT)and GW methodologies.

Nous nous proposons d'explorer les propriétés structurales et vibrationnelles de la glace d'ammoniac en présence de fluor et d'eau sur une large gamme de pression et de température. Ces solides moléculaires et ioniques sont des idéaux afin d'étudier les quatre liaisons hydrogène les plus importantes O-H...O, N-H...N, O-H...N et N-H...F et le mécanisme de transfert de proton le long de ces liaisons. Bien que ces mélanges soient composés de molécules simples, ils sont omniprésents dans l'univers et en particulier à l'intérieur des planètes géantes glacées (Neptune, Uranus) et de leurs satellites (Titan, Ganymède), ce qui a motivé leur exploration sous haute condition de pression et de température. Nous utilisons plusieurs méthodes ab initio afin d'étudier le diagramme de phase de deux cristaux, le fluorure d'ammonium (NH4F) et l'ammoniac monohydraté (NH3.H2O). Après avoir prédit plusieurs structures cristallographiques à 0 K, nous avons pu nous consacrer aux calculs de spectres Raman et infrarouge ainsi qu'à la prise en compte des effets de désordre et dynamique en utilisant la méthode de la dynamique moléculaire ab initio. Nos résultats ont ensuite été comparés à ceux issus des expériences sous enclumes de diamant. Nous avons découvert une transition de type ordre-désordre au sein de NH4F et un diagramme de phase très similaire au diagramme de phase de l'eau. Enfin, nous avons démontré que des conditions thermodynamiques relativement douces étaient suffisantes afin de transformer le mélange NH3.H2O en un cristal exotique où coexiste des molécules d'ammoniac et d'eau avec des ions hydroxyde et ammonium. / In this thesis, we explore the structural and vibrational properties of fluorinated and hydrated ammonia ices over a wide range of pressure and temperature. These crystals are considered as models to study the four main hydrogen bonds O-H…O, N-H…N, O-H…N, N-H…F and proton hopping mechanisms along these bonds. Although these mixtures are composed of simple molecules they are considered to be the major components of the interior of the giant icy planets (Neptune, Uranus) and their satellites (Titan, Ganymede), which has motivated their exploration under high pressure-temperature conditions. Here we investigate the phase diagram of two crystals, ammonium fluoride (NH4F) and ammonia monohydrate (NH3.H2O) by using several ab initio methods. Firstly, we have performed random structural search calculations in order to predict the lowest enthalpy structures at 0 K. On the other hand, we have calculated Raman and infrared spectra of the most promising structures and ab initio molecular dynamics simulations were performed for the understanding of disorder and dynamic effects. Then, our results have been compared to experimental evidence. We have discovered an order-disorder transition in NH4F with a phase diagram surprisingly similar to that of water. Finally, we have demonstrated that relatively mild pressure conditions were sufficient to transform NH3.H2O into an exotic phase where the standard molecular forms of water and ammonia coexist with their ionic counterparts, hydroxide and ammonium ions.

This master's thesis studies the electronic properties changes of graphene caused by substrate SiO2, adsorbed molecules of water and atoms of gallium. There are tested different geometrical configurations of these systems and consequently calculated band structures to derive the changes of the electronic properties: the doping effect and band gap opening of graphene layer.

RÉSUMÉ : Le but de cette thèse est d’étudier en ab initio l’influence des solutés hydrogène ou oxygène sur les défauts étendus dans le titane alpha. Les résultats sont divisés en trois parties. Dans la première partie le site interstitiel octaédrique du Ti-alpha est trouvé être le site énergétiquement le plus favorable pour un H ou un O. Les calculs avec différentes concentrations en H ou O montrent que la présence d’H augmente le volume tandis que O a un effet inverse. La présence d’H en soluté diminue légèrement les constantes élastiques, la présence d’O a un effet opposé. Deux nouvelles SF sont trouvées dans la deuxième partie : une faute 0,57·(c+a) dans le plan π2 et 0,215·[1-102] dans le plan π1. La deuxième est reliée à la faible énergie de formation de la macle {10-11}. Un mécanisme de dissociation triple du cœur de la dislocation vis c+a est proposé. Mais cette dissociation ne semble pas se produire spontanément à partir d'un cœur de dislocation initialement parfait. Puisque la ségrégation à la faute signifie une diminution de l’énergie de faute, nous déduisons de nos calculs de ségrégation que la présence d'O rend sans doute la formation des SF énergétiquement plus difficile, contrairement au cas d’H. H ségrége fortement au cœur d’une dislocation vis a, avec une énergie variant de 0,06 à 0,30 eV, tandis que O y segrége très difficilement. Positionnés dans les sites les plus internes du cœur d’une dissociation prismatique métastable, en glissement, H et O induisent une dissociation dans le plan π1 ou vers une configuration prisme-π1 mélangée. Les barrières d'énergie de Peierls mesurées avec H et O dans différents sites et avec différentes concentrations montrent que H rend le glissement de la dislocation plus difficile, augmentant ainsi sans doute la cission critique résolue dans le plan prismatique, ce qui est en accord avec les mesures expérimentales. Mais les effets de H ne sont pas assez grands pour induire un glissement dévié vers le plan π1 et la dislocation continuera à glisser dans un même plan prismatique. Avec un O, la barrière d'énergie de Peierls est extrêmement élevée, beaucoup plus que celle pour un glissement dans le plan π1 ou un glissement dans le plan prismatique le plus proche. Du glissement dévié devrait ainsi être induit. Finalement, trois méthodes différentes de déformations de macles sont utilisées. Les stabilités structurale des joints de macles dépendent de leurs caractères structuraux intrinsèques mais aussi des modes de déformation appliqués. La macle la plus observée, {10-12}, et la macle {11-22} ne résistent pas à une déformation de plus de 1% ou 2% selon l’axe c. La présence de H ou O ségrégés améliorent la stabilité des macles {10-12} et {11-22}. Un modèle de dipôle de dislocations de mâclage (TD) est proposé pour permettre de simuler une TD dans une super-cellule de petite taille. Pour {10-12} et sa TD, les énergies de ségrégation de H et O mesurées au niveau du joint permettent de valider ce modèle. H et O peuvent se distribuer de manière plus ou moins homogène au joint et niveau de la TD mais pas dans les sites interstitiels de la couche atomique liée à la TD. / ABSTRACT: The aim of this thesis is to study the influence of hydrogen or oxygen solutes on extended defects in alpha titanium by ab initio calculation. Results are divided into three parts. In a first part the octahedral interstitial site of alpha-Ti is found energetically more favorable for a H or an O atom. The presence of H increases the volume while O has the opposite effect. The presence of H slightly decreases the elastic constants of alpha-Ti while O has an opposite effect. In a second part two new SFs are found: 0.57·(c+a) on π2 and 0.215·[1-102] on π1 plane. The second one is related to the low formation energy of the {10-11} twin boundary. A c+a screw dislocation 3-part dissociation mechanism is proposed. However the c+a screw core tends to spread differently according to the initial core position and a complete 3-part dissociation is not found, which may mean that such a dissociation is not easily obtainable from an initially perfect dislocation core. As segregation to SF means a decrease of the SF energy, the presence of O may make the SF formation energetically more difficult, contrary to H case. H strongly segregates to the a screw dislocation core region with segregation energies varing from 0.06 to 0.3 eV while O hardly segregates to it. Both H and O in core sites change the meta-stable gliding prismatic dissociation to π1 plane or a prism-π1 plane mixed configuration. According to our measurements of Peierls energy barriers with H or O in different sites and concentrations, H makes the gliding more difficult, thus increasing the CRSS in prismatic plane, in agreement with experimental measurements. The effect of H is not big enough to induce a cross-slip of the gliding a-screw dislocation to the π1 plane and that screw will prefer to keep on gliding in its same prismatic plane. The Peierls energy barrier is extremely increased when an O is present in the core position, much higher than the barrier for π1 plane glide or a glide in the nearest prismatic plane. A cross-slip could happen in this case. In the last part, three different deformations are applied to TBs. Their structural stabilities depend not only on their intrinsic characters at the atomistic level but also on the deformation mode applied. {10-12}, {11-22} TB structures fail for deformations as low as 1% or 2% along the c-axis. The {11-21} and the {10-11} TBs are much more resistant. The presence of segregated H and O enhances the {10-12} and {11-22} TB limited stability. A twinning disconnection dipole model is proposed which allows the simulation of a TD in a size limited supercell. Segregation energy calculations with the {10-12} TB and its TD validate the model at the TB level and show that H and O should distribute more or less homogeneously to the TD core and the TB, with only a slight preference to the TD core although not at the interstitial sites of the atomic layer related to the disconnection step itself.