Cette thèse se propose d'explorer les mérites d'une famille d'approches de simulation quantique ab initio, les théories de perturbation à N-corps, pour l'exploration des propriétés électroniques et optiques de systèmes organiques. Nous avons étudié en particulier l'approximation dite de GW et l'équation de Bethe-Salpeter, très largement utilisées dès les années soixante pour les semiconducteurs de volume, mais dont l'utilisation pour les systèmes organiques moléculaires est très limitée. L'étude de quelques cas d'intérêt pour le photovoltaïque organique, et en particulier de petites molécules pour lesquelles sont disponibles des données expérimentales ou des résultats issus d'approches de chimie quantique, nous ont permis de valider ces approches issues de la physique du solide.Ce doctorat s'inscrit dans le cadre du développement d'un outil de simulation quantique spécifique (le projet FIESTA) dont l'objectif est de combiner les formalismes GW et Bethe-Salpeter avec les techniques de la chimie quantique, c'est-à-dire en particulier l'utilisation de bases localisées analytiques (bases gaussiennes) et des approches de type «résolution de l'identité» pour le traitement des intégrales Coulombiennes. Ce code est aujourd'hui massivement parallélisé, permettant, au delà des études de validation présentées dans ce travail de thèse, l'étude de systèmes complexes comprenant plusieurs centaines d'atomes. En cours de développement, l'incorporation d'approches hybrides combinant mécanique quantique et écrantage à longue portée par des approches modèles de milieu polarisable m'a permis d'une part de me familiariser avec le code et le développement méthodologique, et permet d'autre part d'envisager l'étude de systèmes réalistes en couplage avec leur environnement.Le manuscrit s‘ouvre sur une introduction au photovoltaïque organique afin de mettre en lumière les questionnements spécifiques qui requièrent le développement de nouveaux outils théoriques à la fois fiables en terme de précision et suffisamment efficaces pour traiter des systèmes de grande taille. Le premier chapitre est d'ordre méthodologique et rappelle les fondements des techniques ab initio de type champ-moyen (Hartree, Hartree-Fock et théorie de la fonctionnelle de la densité). En partant des principes de la photoémission, les théories de perturbation à N-corps et la notion de quasi-particule sont ensuite introduites, conduisant aux équations de Hedin et aux approximations GW et COHSEX. De même, à partir de la compréhension d'une expérience d'optique, le traitement des interactions électron-trou est présenté, menant à l'équation de Bethe-Salpeter. Le chapitre 2 introduit brièvement les spécificités techniques liées à l'implémentation des formalismes GW et Bethe-Salpeter. Les propriétés analytiques des bases gaussiennes et les principes mathématiques derrière les techniques de type «résolution de l'identité» et «déformation de contour», sont brièvement décrites. Le troisième chapitre présente les résultats scientifiques obtenus durant cette thèse. Le cas paradigmatique d'un polypeptide model nous permettra de discuter des spécificités de l'approche GW appliquée à des systèmes moléculaires afin d'obtenir des énergies de quasiparticule de bonne qualité. De même, l'utilisation de l'équation de Bethe-Salpeter pour l'obtention du spectre optique de ce système sera présentée, ainsi que le cas d'une famille de colorants d'importance pour les cellules de Graetzel (les coumarines). Finalement, nous explorons dans le cas du fullerène C60 et du graphène le calcul des termes de couplage électron-phonon dans le cadre de l'approche GW, c'est-à-dire au delà des approches standards de type théorie de la fonctionnelle de la densité. Notre étude vise à vérifier si une approximation statique et à écrantage constant au premier ordre permet de garder la qualité des résultats GW pour un coût numérique réduit. Après la conclusion, les appendices donnent le détail de certaines dérivations. / The present thesis aims at exploring the properties and merits of the ab initio Green's function many-body perturbation theory (MBPT) GW and Bethe-Salpeter formalisms, in order to provide a well-grounded and accurate description of the electronic and optical properties of condensed matter systems. While these approaches have been developed for extended inorganic semiconductors and extensively tested on this class of systems since the 60 s, the present work wants to assess their quality for gas phase organic molecules, where systematic studies still remain scarce. By means of small isolated study case molecules, we want to progress in the development of a theoretical framework, allowing an accurate description of complex organic systems of interest for organic photovoltaic devices. This represents the main motivation of this scientific project and we profit here from the wealth of experimental or high-level quantum chemistry reference data, which is available for these small, but paradigmatic study cases.This doctoral thesis came along with the development of a specific tool, the FIESTA package, which is a Gaussian basis implementation of the GW and Bethe-Salpeter formalisms applying resolution of the identity techniques with auxiliary bases and a contour deformation approach to dynamical correlations. Initially conceived as a serial GW code, with limited basis sets and functionalities, the code is now massively parallel and includes the Bethe-Salpeter formalism. The capacity to perform calculations on several hundreds of atoms to moderate costs clearly paves the way to enlarge our studies from simple model molecules to more realistic organic systems. An ongoing project related to the development of discrete polarizable models accounting for the molecular environment allowed me further to become more familiar with the actual implementation and code structure.The manuscript at hand is organized as follows. In an introductory chapter, we briefly present the basic mechanisms characterizing organic solar cells, accentuating the properties which seek for an accurate theoretical description in order to provide some insight into the factors determining solar cell efficiencies. The first chapter of the main part is methodological, including a discussion of the principle features and approximations behind standard mean-field techniques (Hartree, Hartree-Fock, density functional theory). Starting from a description of photoemission experiments, the MBPT and quasiparticle ideas are introduced, leading to the so-called Hedin's equations, the GW method and the COHSEX approach. In order to properly describe optical experiments, electron-hole interactions are included on top of the description of inter-electronic correlations. In this context, the Bethe-Salpeter formalism is introduced, along with an excursus on time-dependent density functional theory. Chapter 2 briefly presents the technical specifications of the GW and Bethe-Salpeter implementation in the FIESTA package. The properties of Gaussian basis sets, the ideas behind the resolution of the identity techniques and finally the contour deformation approach to dynamical correlations are discussed. The third chapter deals with the results obtained during this doctoral thesis. On the electronic structure level, a recent study on a paradigmatic dipeptide molecule will be presented. Further, also its optical properties will be explored, together with an in-depth discussion of charge-transfer excitations in a family of coumarin molecules. Finally, by means of the Buckminster fullerene C60 and the two-dimensional semi-metal graphene, we will analyze the reliability of two many-body formalisms, the so-called static COHSEX and constant-screening approximation, for an efficient calculation of electron-phonon interactions in organic systems at the MBPT level. After a short conclusion, the Appendix containing details and derivations of the formalisms presented before closes this work.

Au cours de cette thèse on étude deux des plus populaires systèmes de nano-échelle, nano fil et points quantiques (quantum dots), dans le cadre d'une approximation basé sur des premiers principes. Afin d'atteindre cet objectif, nous avons utilisé et développé des théories plus sophistiquées qui nous ont permis d'avoir un meilleur aperçu de la façon dont les systèmes se comportent. Un aspect commun qui limite ces deux types de systèmes (nano fil et points quantiques) est la souplesse de contrôler les propriétés électroniques et optiques. Cette accordabilité des propriétés électroniques et optiques les dote d'un grand intérêt technologique, et elle est la raison de sa popularité / In this thesis we studied two of the most popular nano-scale systems, nano-wires and quantum dots, via a first-principles approach. In order to achieve this objective we have used and developed state-of-the-art theories that allowed us to have a greater insight into the the way this systems behave. One common aspect that bounds this two class of systems (nano-wires and quantum dots) is the flexibility to control their electronic and optical properties. This tunability of their electronic and optical properties, endows them of great technological interest, and is the reason behind their popularity

Non-radial pulsations have been identified in a number of accreting white dwarfs in cataclysmic variables. These stars offer insight into the excitation of pulsation modes in atmospheres with mixed compositions of hydrogen, helium, and metals, and the response of these modes to changes in the white dwarf temperature. Among all pulsating cataclysmic variable white dwarfs, GW Librae stands out by having a well-established observational record of three independent pulsation modes that disappeared when the white dwarf temperature rose dramatically following its 2007 accretion outburst. Our analysis of Hubble Space Telescope (HST) ultraviolet spectroscopy taken in 2002, 2010, and 2011, showed that pulsations produce variations in the white dwarf effective temperature as predicted by theory. Additionally in 2013 May, we obtained new HST/Cosmic Origin Spectrograph ultraviolet observations that displayed unexpected behaviour: besides showing variability at a parts per thousand integral 275 s, which is close to the post-outburst pulsations detected with HST in 2010 and 2011, the white dwarf exhibits high-amplitude variability on an a parts per thousand integral 4.4 h time-scale. We demonstrate that this variability is produced by an increase of the temperature of a region on white dwarf covering up to a parts per thousand integral 30 per cent of the visible white dwarf surface. We argue against a short-lived accretion episode as the explanation of such heating, and discuss this event in the context of non-radial pulsations on a rapidly rotating star.

L'échec des modèles pour prévoir l'alignement de bandes dans un empilement MOS (métal-oxyde-semiconducteur) rend utile et intéressant une étude à l'échelle atomique. Nous exposons deux théories ab initio: d'abord la théorie de la fonctionnelle de la densité (DFT), dont dépendent les codes Abinit et Siesta utilisés pour étudier la nature quantique des matériaux ; puis la théorie de l'approximation GW qui corrige les niveaux d'énergies électroniques mal évalués avec l'approximation utilisée en DFT. Ces deux théories sont appliquées à chaque matériau composant l'empilement. Puis nous étudions la façon dont intervient le dipôle dans l'alignement de bandes. La méthode de Van de Walle et Martin est adaptée pour évaluer de façon précise ces alignements grâce à l'approximation GW. Enfin cette méthode est appliquée aux différentes interfaces oxyde/semiconducteur et métal/oxyde et montre que les alignements de bandes estimés par simulation tendent vers un accord avec les mesures expérimentales.

My thesis fits into the domain of theoretical spectroscopy. This term describes a set of theoretical approaches that go hand-in-hand with several experimental techniques such as optical absorption and reflectivity, inelastic X-ray scattering (IXS), electron energy-loss spectroscopy (EELS) and photoelectron (or photoemission) spectroscopy. This set of ab-initio theories is used to simulate, study, predict and understand what is and will be seen in experiment. These spectroscopies are all connected to the dielectric function ε(ω ) of an electronic system which is, in fact, a fundamental quantity in many modern electronic structure theories. In particular I focused my research on photoemission spectroscopy, where the dielectric function enters as the screening of the hole due to the system. During my thesis I have worked on the development of new theoretical approaches, the aim of my project being to go beyond state-of-the-art methods used in electronic structure calculations. These methods stem mainly from two larger theoretical frameworks: Time-Dependent Density-functional Theory (TDDFT) and Green's function theory -- also known as Many-Body Perturbation Theory (MBPT). I carried on the theoretical development in parallel with numerical simulations on real materials and with experimental measurements, performed to verify the reliability of theory.

Thesis is focused on the necessary adjustments to the accounting information required for or the purpose of valuating a company using method EVA. The first part is focused on the development of accounting and accounting systems, in second part is defined by basic terms of valuating a company and the third section describes the accounting information (the necessary adjustments) divided by valuation phase. The most important part of the third part is the transformation of accounting data for economic (financial leasing and intangible assets).

The thesis is focused on the description of LTE system in terms of signaling on interfaces of LTE and EPC subsystems, such as UE initial network connection. The next section describes the types of diagnostic methods for mobile networks using OSS, drive testing and flow analysis. The thesis also aims at description of key performance indicators (independent service QoS parameters and the KPI for the radio part of the network). Part of the network measurement includes a description of the driver settings and views for analysis. It is also described the implementation of measuring and evaluating the results.

Experterna i media har till uppgift att kommentera och analysera nyhetshändelser. Genom rollen att göra skeenden begripliga för en större publik besitter de makten att påverka samhällsopinionen. Med utgångspunkt i mediernas fascination för att rapportera om brott och straff undersöktes i den här studien svensk press användning av kriminologer som experter. Studiens empiriska material bestod av 710 artiklar publicerade i elva storstadstidningar mellan den 1 april 2012 och den 31 mars 2017. Uppsatsens teoretiska ramverk byggde på medielogik som förklarar den process i vilken medierna presenterar och distribuerar information. I den här studien bidrog den med förståelse till svensk press val av kriminologer och hur dessa kan kopplas till mediernas behov. Dessutom användes representationsteori i syfte att analysera kriminologernas citat. För att studera detta tillämpades en kvantitativ innehållsanalys där ett kodschema konstruerades och variabler utformades för att besvara studiens frågeställningar. Senare användes en kvalitativ innehållsanalys för att analysera tematik i kriminologernas citat.    Resultatet visade att fördelningen av kriminologer är obalanserad. Framför allt framträdde Jerzy Sarnecki i en dominerande del av artiklarna, men också Leif GW Persson och Mikael Rying förekom i en förhållandevis stor utsträckning. I resultatet konstaterades dock att Leif GW Perssons namn och ansikte oftast användes för att locka till läsning, något som nyanserar bilden av Jerzy Sarnecki som mediernas favoritkriminolog. Vidare påvisades det att studiens topp tre kriminologer sällan tilldelas en närmare presentation än att de är just kriminologer, detta jämfört med övriga kriminologer som ofta legitimeras med forskningsunderlag och ytterligare yrkesroll. Dessutom uppmärksammades att kriminologerna, beroende på hur de uttalade sig, gavs varierat utrymme i olika frågor. / Recent years the criminologists has possessed a larger quantity of space in the Swedish media; many times to contribute with knowledge and theories for the public regarding crime events. At the same time news desks experience a greater task with an increasing pressure to be first with publishing different stories, which could lead to a comfortable approach to new expert sources. The main purpose of the thesis was therefore to examine how criminologists are represented and presented in some of the major Swedish newspapers from a perspective of Stuart Hall’s ideas of representation. With a quantitative and qualitative content analysis a total of 710 news articles in eleven different papers were examined, from the beginning of April 2012 until the end of March 2017. In an early stage it was clear that just three criminologists occurred in over 70 percent of the articles. Later different themes between them could be identified which displayed how their attributes were used in different ways, giving them access to express themselves in varied topics. Furthermore the analysis showed that the experts were represented in a rather hard boiled manner. For example topics as fatal violence and criticism of the police were two categories in which they often appeared.

In this Thesis a set of functional solid oxides for industrial applications have been addressed by first principles and thermodynamical modelling. More specificially, measurable quantities such as Gibbs free energy, geometry and electronic structure have been calculated and compared when possible with experimental data. These are crystalline and amorphous aluminum oxide (Al2O3), Zirconia (ZrO2), magnesium oxide (MgO), indiumoxide (In2O3) and Kaolinite clay (Al2Si2O5(OH)4). The reader is provided a computation tool box, which contains a set of methods to calculate properties of oxides that are measurable in an experiment. There are three goals which we would like to reach when trying to calculate experimental quantities. The first is verification. Without verification of the theory we are utilizing, we cannot reach the second goal -prediction. Ultimately, this may be (and to some extent already is) the future of first principles methods, since their basis lies within the fundamental quantum mechanics and since they require no experimental input apart from what is known from the periodic table. Examples of the techniques which may provide verification are X-Ray Diffraction (XRD), X-ray Absorption and Emission Spectroscopy (XAS and XES), Electron Energy Loss Spectroscopy and Photo-Emission Spectroscopy (PES). These techniques involve a number of complex phenomena which puts high demands on the chosen computational method/s. Together, theory and experiment may enhance the understanding of materials properties compared to the standalone methods. This is the final goal which we are trying to reach -understanding. When used correctly, first principles theory may play the role of a highly resolved analysis method, which provides details of structural and electronic properties on an atomiclevel. One example is the use of first principles to resolve spectra of multicomponentsamples. Another is the analysis of low concentrations of defects. Thorough analysis of the nanoscale properties of products might not be possible in industry due to time and cost limitations. This leads to limited control of for example low concentrations of defects, which may still impact the final performance of the product. On example within cutting tool industry is the impact of defect contents on the melting point and stability of protective coatings. Such defects could be hardening elements such as Si, Mn, S, Ca which diffuse from a steel workpiece into the protective coating during high temperature machining. Other problems are the solving of Fe from the workpiece into the coating and reactions between iron oxide, formed as the workpiece surface is oxidized, and the protective coating. The second part of the computational toolbox which is provided to the reader is the simulation of solid oxide synthesis. Here, a formation energy formalism, most often applied to materials intended in electronics devices is applied. The simulation of Chemical Vapour Deposition (CVD) and Physical Vapor Deposition (PVD) requires good knowledge of the experimental conditions, which can then be applied in the theoretical simulations. Effects of temperature, chemical and electron potential, modelled concentration and choice of theoretical method on the heat of formation of different solid oxides with and without dopants are addressed in this work. A considerable part of this Thesis is based upon first principles calculations, more specifically, Density Functional Theory (DFT) After Kohn and Pople received the Nobel Prize in chemistry in 1998, the use of DFT for computational modelling has increased strikingly (see Fig. 1). The use of other first principles methods such as hybrid functionals and the GW approach (see abbreviations for short explanations and chapter 4.5 and 5.3.) have also become increasingly popular, due to the improved computational resources. These methods are also employed in this Thesis. / QC 20110201

The Middle and Upper Atmosphere Model (MUAM) is used to implement a modern whole atmosphere gravity wave (GW) parameterization (Yigit et al., 2008). To adjust the resulting model climatology to observations, we compare two different spectra of GW phase speeds and the related momentum fluxes included in this parameterization.The first spectrum, termed Spectrum1, which is more narrow, covers phase speeds up to ±80 m s−1 and the second one, Spectrum2, goes up to ±95 m s−1. We analyzed the zonal GW drag, the acceleration of the zonal mean flow owing to breaking GWs, and the resulting background circulation. We also examined zonal wind amplitudes of the diurnal and semidiurnal migrating solar tide. As a result, we find that both GW phase speed spectra have their advantages and disadvantages. For example, Spectrum1 reproduces the zonal wind reversal in the mesosphere correctly, but it does not lead to a reversed zonal GW drag in the lower thermosphere. While the amplitudes of the diurnal tide tend to be more realistic for Spectrum1, those of the semidiurnal tide are more representational for Spectrum2. Overall, the new GW parameterization is a substantial improvement to the MUAM model. / Das Modell für die mittlere und obere Atmosphäre (MUAM) wird für die Implementierung einer neuen Schwerewellenparametrisierung, die für die gesamte Atmosphäre geeignet ist, verwendet (Yigit et al., 2008). Um die Klimatologie des Modells basierend auf der neuen Parametrisierung an Beobachtungen anzupassen, werden zwei verschiedene Spektren für die Phasengeschwindigkeiten von Schwerewellen und deren Impulsflüsse verglichen. Das erste Spektrum, als Spectrum1 bezeichnet, das einen schmaleren Bereich von Phasengeschwindigkeiten abdeckt, geht bis zu ±80 m s−1 und das zweite, Spectrum2, geht bis zu ±95 m s−1. Wir analysieren die Beschleunigung des mittleren Zonalwindes durch brechende Schwerewellen und die daraus resultierende Hintergrundzirkulation. Wir untersuchen ebenfalls die Zonalwindamplituden der ganztägigen und halbtätigen migrierenden solaren Gezeiten. Das Ergebnis dieses Experiments zeigt, dass beide Phasengeschwindigkeitsspektren ihre Vor- sowie Nachteile haben. Beispielsweise reproduziert Spectrum1 die Umkehr des Zonalwindes in der Mesosphäre korrekt, jedoch führt es nicht zu einer umgekehrten zonalen Beschleunigung durch Schwerewellen in der Thermosphäre. Während die Amplituden der ganztägigen Gezeit bezüglich Spectrum1 wesentlich realistischer zu sein scheinen, sind jene der halbtägigen Gezeit bezüglich Spectrum2 repräsentativer. Insgesamt ist die neue Parameterisierung eine erhebliche Verbesserung des MUAM Modells.