Die Untersuchung von korrelationsinduzierten Effekten, wie beispielsweise der kollektiven magnetischen Ordnung, verlangt einen Einblick in die elektronische Struktur der Festkörper. Hier stehen mit den Ein-Teilchen-Spektroskopien (Photoemission und inverse Photoemission) sowie den Zwei-Teilchen-Spektroskopien (Auger-Elektronen- und Appearance-Potential-Spektroskopien) nützliche experimentelle Werkzeuge zur Verfügung. Eine adäquate Interpretation der experimentell ermittelten Spektren erfordert die Berücksichtigung (i) der elektronischen Korrelationen, (ii) der orbitalen Entartung sowie (iii) der Übergangsmatrixelemente. Der vorliegenden Arbeit liegt ein Multiband-Hubbard-Modell zugrunde, das die Hopping- und Hybridisierungsprozesse der für die 3d-Übergangsmetalle relevanten Orbitale (4s-, 4p- und 3d-Orbitale) und die lokalen Coulomb-Wechselwirkungen der stark lokalisierten 3d-Elektronen beschreibt. Die Hopping- und Hybridisierungsparameter werden Bandstrukturrechnungen (lokale Dichteapproximation der Dichtefunktionaltheorie) entnommen. Als Ein-Teilchen-Basis werden quasiatomare Orbitale verwendet, deren hohe Symmetrie es erlaubt, die vollständige lokale 3d-Coulomb-Matrix mithilfe von lediglich drei Zahlen (effektive Slater-Parameter) zu parametrisieren. Fasst man die effektiven Slater-Parameter zu einer mittleren direkten Wechselwirkungsstärke U und einer mittleren Austauschwechselwirkungsstärke J zusammen, so sind U und J die beiden einzigen Parameter des Modells, die dann an experimentelle Daten (z.B. Spinmoment bei T=0K) angepasst werden können. Für die Berechnung der Ein-Teilchen-Green-Funktion, die die Photoemissionsspektren bestimmt, wird die Störungstheorie zweiter Ordnung um die Hartree-Fock-Lösung verwendet und damit die Magnetisierung als Funktion der Temperatur für Nickel berechnet. Die Kombination der errechneten Green-Funktion mit den Übergangsmatrixelementen der Photoemission gestattet dann einen quantitiven Vergleich mit experimentellen spin- und winkelaufgelösten Daten für endliche Temperaturen und verschiedene Photonenenergien. Die Zwei-Teilchen-Spektren lassen sich als Funktionale der Ein-Teilchen-Green-Funktion auffassen. Durch die Verwendung sogenannter Leiter-Näherungen werden nicht nur die Wechselwirkungen der an den Übergangsprozessen beteiligten Teilchen mit dem Restsystem (Selbstenergieeinschübe) berücksichtigt, sonderen auch die Wechselwirkungen der direkt am Prozess beteiligten Teilchen untereinander (Endzustandskorrelationen). Die Verwendung des Zwei-Stufen-Modells impliziert jedoch, dass die Valenzbandelektronen das Potential des zuvor erzeugten Core-Lochs abschirmen werden, was sich insbesondere auf die Auger-Spektren auswirkt. Am Beispiel der Appearance-Potential-Spektren wird die vielteilchentheoretisch berechnete Green-Funktion mit entsprechenden Übergangsmatrixelementen kombiniert. Die resultierenden Spektren sind für alle betrachteten Temperaturen in sehr guter Übereinstimmung mit gemessenen Spektren. / The investigation of correlation-induced effects, as for example the collective magnetic order, requires an insight into the electronic structure of solids. In this context the one-particle spectroscopies (photoemission and inverse photoemission) as well as the two-particle spectroscopies (Auger electron and appearance potential spectroscopies) represent useful experimental tools. An adequate interpretation of the experimentally determined spectra requires the consideration of (i) electronic correlations, (ii) orbital degeneration as well as (iii) transition-matrix elements. The present work uses a multi-band Hubbard model, which describes the hopping and hybridization processes of the relevant orbitals in the 3d-transition metals (4s, 4p and 3d orbitals) and the local Coulomb interactions of the strongly localized 3d electrons. The hopping and hybridization parameters are taken from band-structure calculations (local density approximation of the density functional theory). As one-particle basis quasi-atomic orbitals are used, whose high symmetry permits it to parameterize the complete local Coulomb-matrix among 3d-electrons by only three numbers (effective Slater parameters). If one combines the effective Slater parameters into an averaged direct interaction strength U and an averaged exchange interaction strength J, then U and J are the only parameters of the model, which can be fitted to experimental data (e.g. spin moment at T=0K). For the calculation of the one-particle Green function, which determines the photoemission spectra, the second order perturbation theory around the Hartree-Fock solution is used. Within this framework the magnetization as function of the temperature is calculated for nickel. The combination of the calculated Green function and the photoemission transition-matrix elements permits a quantitative comparison with spin- and angle-resolved measurements for finite temperatures and different photon energies. The two-particle spectra can be understood as functionals of the one-particle Green function. Due to the usage of so called ladder approximations one is able to account not only for the correlations between the considered particles and the remainder system (self-energy insertations) but also for the correlations among those particles directly involved in the transition (final state correlations). The usage of the two-step model implies, however, that the valence band electrons tend to screen the additional potential of the core hole created before, which in particular affects the Auger spectra. As an example, the appearance potential spectra of nickel are calculated by combining the corresponding Green function with appropriate transition-matrix elements. The resulting spectra are in a very good agreement with measured spectra for all temperatures.

No description available.

La réalisation de condensats de Bose-Einstein et de gaz de Fermi dégénérés ont déclenché d'énormes progrès dans les méthodes théoriques ainsi que dans la mise en place de nouvelles techniques expérimentales. Parmi celles-ci, de fascinantes possibilités viennent de l'implémentation de réseaux optiques : potentiels périodiques pour atomes neutres créés à travers l'interférence de rayons laser. Un gaz dégénéré dans un réseau optique peut être forcé dans des pièges fortement anisotropes, jusqu'à réduire la dimensionnalité du système physique. Du point de vue fondamental, le comportement des ondes de matière en dimensions réduites éclaircit les propriétés intrinsèques des interactions entre particules. En outre, ces systèmes à dimensionnalité réduite peuvent être manipulés afin de créer des simulateurs quantiques de la matière condensée, comme par exemple des réseaux à deux dimensions, dans un environnement pur et contrôlable. Motivés par les passionnantes perspectives de ce domaine, on a consacré cette Thèse à l'étude théorique de deux systèmes dans lesquels une onde de matière se propage en dimensions réduites. L'interaction dipôle-dipôle, à longue portée et anisotrope, affecte fortement le comportement des gaz quantiques. Les progrès expérimentaux dans ce domaine florissant permettront bientôt de piéger dans des réseaux optiques un gaz dégénéré de dipôles. Dans la première partie de cette thèse, on considère l'apparition d'une seule résonance dipolaire dans l'interaction entre deux particules pour différents systèmes quasi-unidimensionnels. On propose une approche à deux canaux qui décrit cette résonance dans un piège harmonique fortement allongé “en forme de cigare”, qui représente l'approximation d'un site d'un réseau optique quasi-unidimensionnel. A` ce stade, on développe un nouveau modèle étendu de Bose-Hubbard atome-dimère, qui est valable pour des bosons dipolaires dans un réseau optique quasi-unidimensionnel. On étudie donc le diagramme de phase du modèle pour T =0 par la diagonalisation exacte de systèmes de petite taille, en soulignant les effets de la résonance dipolaire sur la physique à plusieurs corps dans le réseau. Dans la seconde partie de la thèse, on propose un modèle pour réaliser des simulateurs quantiques de cristaux bidimensionnels avec des atomes froids, basé sur le piégeage indépendant de deux espèces atomiques. La première constitue une onde de matière bidimensionnelle qui interagit exclusivement avec les atomes de la seconde espèce, piégés aux nœuds d'un réseau optique bidimensionnel. En introduisant une approche théorique générale, on examine les propriétés de transport de l'onde de matière. On propose des exemples d'application pour réseaux soit de Bravais (carré, triangulaire), soit de non-Bravais (graphène, kagomé), en étudiant soit des systèmes périodiques idéaux, soit des systèmes de taille expérimentale et désordonnés. Les caractéristiques d'un réseau atomique artificiel dépendent de l'intensité de l'interaction entre les deux espèces, qu'on montre être largement réglable grâce à des résonances à dimensionnalité mixte de type 0D-2D. / The experimental achievement of Bose-Einstein condensation and Fermi degeneracy with ultracold gases boosted tremendous progresses both in theoretical methods and in the development of new experimental tools. Among them, intriguing possibilities have been opened by the implementation of optical lattices: periodic potentials for neutral atoms created by interfering laser beams. Degenerate gases in optical lattices can be forced in highly anisotropic traps, reducing the effective dimensionality of the system. From a fundamental point of view, the behavior of matter waves in reduced dimensions sheds light on the intimate properties of interparticle interactions. Furthermore, such reduced-dimensional systems can be engineered to quantum-simulate fascinating solid state systems, like bidimensional crystals, in a clean and controllable environment. Motivated by the exciting perspectives of this field, we devote this Thesis to the theoretical study of two systems where matter waves propagate in reduced dimensions.The long-range and anisotropic character of the dipole-dipole interaction critically affects the behavior of dipolar quantum gases. The continuous experimental progresses in this flourishing field might lead very soon to the creation of degenerate dipolar gases in optical potentials. In the first part of this Thesis, we investigate the emergence of a single dipolar-induced resonance in the two-body scattering process in quasi-one dimensional geometries. We develop a two-channel approach to describe such a resonance in a highly elongated cigar-shaped harmonic trap, which approximates the single site of a quasi-one- dimensional optical lattice. At this stage, we develop a novel atom-dimer extended Bose- Hubbard model for dipolar bosons in this quasi-one-dimensional optical lattice. Hence we investigate the T=0 phase diagram of the model by exact diagonalization of a small- sized system, highlighting the effects of the dipolar-induced resonance on the many-body behavior in the lattice.In the second part of the Thesis, we present a general scheme to realize cold-atom quantum simulators of bidimensional atomic crystals, based on the possibility to independently trap two different atomic species. The first one constitutes a two-dimensional matter wave which interacts only with the atoms of the second species, deeply trapped around the nodes of a two-dimensional optical lattice. By introducing a general analytic approach, we investigate the matter-wave transport properties. We propose some illustrative appli- cations to both Bravais (square, triangular) and non-Bravais (graphene, kagomé) lattices, studying both ideal periodic systems and experimental-sized, eventually disordered, ones. The features of the artificial atomic crystal critically depend on the two-body interspecies interaction strength, which is shown to be widely tunable via 0D-2D mixed-dimensional resonances.

Neste trabalho utilizamos sistemas modelos no desenvolvimento, implementação e análise de funcionais orbitais da densidade, focando, em particular, nas correções de autointeração de Perdew-Zunger (PZSIC) e Lundin-Eriksson (LESIC). Aplicamos as correções de auto-interação ao funcional local (LDA) do modelo de Hubbard e de poços quânticos semicondutores, ambos unidimensionais, no caso estático e dependente do tempo, respectivamente. Para o modelo de Hubbard unidimensional, comparamos a LDA, LDA+PZSIC e LDA+LESIC, identificando o desempenho para energias e densidades do estado fundamental, com e sem impurezas locais, além do gap fundamental de energia. Em adição, averiguamos o desempenho diante de cargas fracionárias, estabelecendo conexões com o erro de delocalização da LDA. Mostramos a possibilidade da correta descrição das freqüências das oscilações de Friedel no modelo de Hubbard, além de investigar como a falha da LDA em reproduzir esse aspecto pode estar relacionada com os erros de auto-interação e delocalização. Investigamos ainda as diferentes possibilidades de implementação autoconsistente de qualquer funcional orbital da densidade, analisando a relação entre funcionais aproximados e suas implementações aproximadas. Nos poços quânticos, sob o enfoque dependente do tempo, analisamos a descontinuidade do potencial de troca e correlação ao variarmos o número de partículas, em dois processos distintos: a ionização eletrônica em um poço simples e dissociação de um poço duplo assimétrico. No último caso, avaliamos os efeitos da descontinuidade no número total de partículas em cada poço, revelando os mecanismos que resgatam a neutralidade elétrica durante processos de dissociação, com a correta carga final inteira. / In this work we use model systems to develop, implement and analyse orbital-dependent density functionals, focusing, specifically, on the self-interaction corrections (SICs) of Perdew and Zunger (PZSIC) and of Lundin and Eriksson (LESIC). These self-interaction corrections are applied to the local-density approximation (LDA) for the one-dimensional Hubbard model and for semiconductor quantum wells, in one-dimensional static and time-dependent situations. For the one-dimensional Hubbard model we compare LDA, LDA+PZSIC and LDA+LESIC, and investigate the performance of these approaches for ground-state energies, densities and energy gaps, with and without impurities in the system. We also consider the case of fractional charges, where a connection to the delocalization error of the LDA can be made. We show that in principle a correct description of the frequences of Friedel oscillations in the Hubbard model can be obtained from DFT, and investigate how the failure of the LDA in reproducing this is related to the selfinteraction and delocalization errors. Moreover, we investigate different procedures for the selfconsistent implementation of any orbital-dependent functional, and analyse the question of the interplay between an approximate functional and its approximate implementation. For quantum wells sytems we analyse, in a time-dependent framework, the discontinuity of the exchange-correlation potential under variation of the particle number in two different processes: the ionization of a simple quantum well and the dissociation of an asymmetric double well. In the latter case, we also consider the effect of changes in the particle number in each subwell, thus revealing the mechanism that restores electric neutrality during dissociation, with correct final charge.

Orientador: Eduardo Miranda / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Física Gleb Wataghin / Made available in DSpace on 2018-08-23T18:51:04Z (GMT). No. of bitstreams: 1 SuarezVillagran_MarthaYolima_D.pdf: 7321255 bytes, checksum: d76479a0e0c1143207cb4ee380a8034d (MD5) Previous issue date: 2013 / Resumo: Estudamos nesta tese alguns aspectos da transição metal-isolante de Mott no caso desordenado. O modelo no qual baseamos nosso estudo é o modelo de Hubbard desordenado, que é o modelo mais simples a apresentar a transição metal-isolante de Mott. Analisamos esse modelo através da Teoria Dinâmica de Campo Médio Estatística (StatDMFT). Essa teoria é uma extensão natural da Teoria Dinâmica de Campo Médio (DMFT), que foi usada com relativo sucesso nos últimos anos para analisar a transição de Mott no caso limpo. Como no caso dessa última, a StatDMFT incorpora os efeitos de correlação eletrônica apenas nos seus aspetos locais. A desordem é tratada de maneira a incorporar todos os efeitos de localização de Anderson. Com essa técnica, analisamos a transição de Mott desordenada no caso bi-dimensional, usando o Monte Carlo quântico para resolver os problemas de impureza única de Anderson requeridos pela StatDMFT. Encontramos as linhas espinodais nas quais o metal e o isolante deixam de ser meta-estáveis. Também estudamos os padrões espaciais das flutuações de quantidades locais, como a auto-energia e a função de Green local, e mostramos como há o aparecimento de regiões metálicas dentro do isolante e viceversa. Analisamos efeitos de tamanho finito e mostramos que, em consonância com os teoremas de Imry e Ma, a transição de primeira ordem desaparece no limite termodinâmico. Analisamos as propriedades de transporte desse sistema através de um mapeamento a um sistema de resistores aleatórios clássicos e calculamos a corrente média e sua distribuição através da transição metal-isolante. Finalmente, estudamos o comportamento da parede de domínio que se forma entre o isolante e o metal no caso limpo. Isso foi feito através de um modelo de uma cadeia unidimensional conectada a reservatórios, um metálico e um isolante, cada um em uma de suas extremidades. Nesse caso, utilizamos o método da Teoria de Perturbação Iterada para a solução dos modelos de impureza única. Encontramos o comportamento da parede como função da temperatura e das interações / Abstract: In this thesis, we studied some aspects of the Mott metal-insulator transition in the disordered case. The model on which we based our analysis is the disordered Hubbard model, which is the simplest model capable of capturing the Mott metal-insulator transition. We investigated this model through the Statistical Dynamical Mean-Field Theory (statDMFT). This theory is a natural extension of the Dynamical Mean-Field Theory (DMFT), which has been used with relative success in the last several years with the purpose of describing the Mott transition in the clean case. As is the case for the latter theory, the statDMFT incorporates the electronic correlation effects only incorporate Anderson localization effects.. With this technique, we analyzed the disordered two-dimensional Mott transition, using Quantum Monte Carlo to solve the associated single-impurity problems. We found the spinodal lines at which metal and insulator cease to be meta-stable. We also studied the spatial fluctuations of local quantities, such as the self-energy and the local Green¿s function, and showed the appearance of metallic regions within the insulator and vice-versa. We carried out an analysis of finite-size effects and showed that, in agreement with the theorems of Imry and Ma, the first-order transition is smeared in the thermodynamic limit. We analyzed transport properties by means of a mapping to a random classical resistor network and calculated both the average current and its distribution across the metalinsulator transition. Finally, we studied the behavior of the domain wall which forms between the metal and the insulator in the clean case. This was done by means of a model of a one-dimensional chain connected to two reservoirs, one metallic and the other insulating, each attached to one of the chain¿s ends. In this case, we used the Iterated Perturbation Theory technique in order to solve the associated singleimpurity problems. We then established the behavior of the domain wall width as a function of temperature and interactions / Doutorado / Física / Doutora em Ciências

Orientador: Marcos César de Oliveira / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Física Gleb Wataghin / Made available in DSpace on 2018-08-18T12:10:11Z (GMT). No. of bitstreams: 1 Farias_ReginaldodeJesusCosta_D.pdf: 4228413 bytes, checksum: 9e7ff892eed5441839ee091ddb080501 (MD5) Previous issue date: 2011 / Resumo: Partindo de uma derivação matemática do modelo de Bose-Hubbard, desenvolvemos os cálculo para o emaranhamento bi e multipartido, através da transição Isolante de Mott-superuida, entre os modos de uma rede óptica para as situações mais simples de dois, três e quatro átomos (N) depositados nestas com igual número de sítios M, ocasionando um fator de preenchimento ? = N=M unitário. Apresentamos uma investigação sobre o controle da dinâmica de um condensado de Bose-Einstein aprisionado em um poço duplo através da ação de um potencial externo dependente do tempo. Apresentamos também uma investigação preliminar de codificação e operações quânticas embasadas em colisões controladas entre átomos em múltiplos poços / Abstract: Starting from a mathematical derivation of the Bose-Hubbard Model (BHM) we analyze the developing of bipartite and multipartite entanglement through the Mott insulator-superuid quantum phase transition, among the modes of an optical lattice to the simplest situations of two, three and four atoms (N) deposited in such optical lattice with equal number of sites (M), where a filling factor ? = N=M = 1 per lattice site is considered. We present an investigation about the controlled dynamic of a Bose-Einstein condensate in a double well driven by an external time dependent potential. Beyond we present some preliminar notes on codi cations and quantum operations in cold controlled collisions among atoms in multiples wells / Doutorado / Física da Matéria Condensada / Doutor em Ciências

Neste trabalho utilizamos sistemas modelos no desenvolvimento, implementação e análise de funcionais orbitais da densidade, focando, em particular, nas correções de autointeração de Perdew-Zunger (PZSIC) e Lundin-Eriksson (LESIC). Aplicamos as correções de auto-interação ao funcional local (LDA) do modelo de Hubbard e de poços quânticos semicondutores, ambos unidimensionais, no caso estático e dependente do tempo, respectivamente. Para o modelo de Hubbard unidimensional, comparamos a LDA, LDA+PZSIC e LDA+LESIC, identificando o desempenho para energias e densidades do estado fundamental, com e sem impurezas locais, além do gap fundamental de energia. Em adição, averiguamos o desempenho diante de cargas fracionárias, estabelecendo conexões com o erro de delocalização da LDA. Mostramos a possibilidade da correta descrição das freqüências das oscilações de Friedel no modelo de Hubbard, além de investigar como a falha da LDA em reproduzir esse aspecto pode estar relacionada com os erros de auto-interação e delocalização. Investigamos ainda as diferentes possibilidades de implementação autoconsistente de qualquer funcional orbital da densidade, analisando a relação entre funcionais aproximados e suas implementações aproximadas. Nos poços quânticos, sob o enfoque dependente do tempo, analisamos a descontinuidade do potencial de troca e correlação ao variarmos o número de partículas, em dois processos distintos: a ionização eletrônica em um poço simples e dissociação de um poço duplo assimétrico. No último caso, avaliamos os efeitos da descontinuidade no número total de partículas em cada poço, revelando os mecanismos que resgatam a neutralidade elétrica durante processos de dissociação, com a correta carga final inteira. / In this work we use model systems to develop, implement and analyse orbital-dependent density functionals, focusing, specifically, on the self-interaction corrections (SICs) of Perdew and Zunger (PZSIC) and of Lundin and Eriksson (LESIC). These self-interaction corrections are applied to the local-density approximation (LDA) for the one-dimensional Hubbard model and for semiconductor quantum wells, in one-dimensional static and time-dependent situations. For the one-dimensional Hubbard model we compare LDA, LDA+PZSIC and LDA+LESIC, and investigate the performance of these approaches for ground-state energies, densities and energy gaps, with and without impurities in the system. We also consider the case of fractional charges, where a connection to the delocalization error of the LDA can be made. We show that in principle a correct description of the frequences of Friedel oscillations in the Hubbard model can be obtained from DFT, and investigate how the failure of the LDA in reproducing this is related to the selfinteraction and delocalization errors. Moreover, we investigate different procedures for the selfconsistent implementation of any orbital-dependent functional, and analyse the question of the interplay between an approximate functional and its approximate implementation. For quantum wells sytems we analyse, in a time-dependent framework, the discontinuity of the exchange-correlation potential under variation of the particle number in two different processes: the ionization of a simple quantum well and the dissociation of an asymmetric double well. In the latter case, we also consider the effect of changes in the particle number in each subwell, thus revealing the mechanism that restores electric neutrality during dissociation, with correct final charge.

The localized/delocalized duality of 5f electrons plays an important role in understanding the complex physics of actinides. Band-structure calculations based on the ad hoc assumption that 5f electrons are simultaneously localized and delocalized explained the observed dHvA experiments very well. This ad hoc assumption also gives the correct equilibrium volume for delta-Pu. Experimentally, the duality of 5f electrons is observed by inelastic neutron scattering experiments, or by soft X-ray angle-resolved photoelectron spectroscopy. It is worth recalling that the origin of partial localization in the 3d and 5f systems is quite different. In compounds with 3d electrons, the large crystalline electric field set up by the surrounding environment of transition metal ions plays a major role. On the other hand, in 5f systems, the Hund's rule correlations play the key role whilst the crystalline electric field is less important. In this thesis we have studied the effect of intra-atomic correlations on anisotropies in hopping matrix elements of different 5f orbitals. For that purpose, we used the effective model that includes on-site interactions that are responsible for Hund's rules and effective hopping terms that result from the hybridization of different 5f orbitals with the environment. Two different approximations, namely, rotationally invariant slave-boson mean-field (RISBMF) and infinite time-evolving block decimation (iTEBD), have been used to investigate the ground-state properties of the Hamiltonian. We have demonstrated that Hund's rule correlations enhance strongly anisotropies in hopping matrix elements. For a certain range of 5f bandwidth parameters this effect may result in a complete suppression of hopping processes for some of 5f orbitals, i.e., the system is in a partially localized phase. Within the RISBMF method, we calculated the ground-state properties and the phase diagram of the system. The suppression of hopping processes in some of 5f orbitals due to Hund's rule correlations can be seen through orbital-dependent quasiparticle weights. In a mean-field theory, a quasiparticle weight of zero for an orbital means a complete suppression of hopping processes in this orbital. Thus, quasiparticle weights and occupation numbers were used to classify partially localized phases. In the calculated phase diagram we obtain four partially localized phases that can be separated into two different sets. In the first set electrons in two orbitals are localized. In the second, electrons in one orbital are localized. The difference between the two sets is not simply the number of localized orbitals but the mechanism for the partial localization. For the first set, the Hund's rule mechanism applies: only those 5f electrons that enable the remaining ones to form a Hund's rule state will delocalize. This mechanism requires to have at least two localized orbitals, therefore it is definitely not applicable to those phases with only one localized orbital. For the second set, a situation similar to a single-band Mott-Hubbard transition applies. The direct on-site Coulomb interaction between jz and -jz electrons plays the key role for understanding the partial localization transition. In order to assess the validity of the RISBMF results we have used the iTEBD method to calculate the ground-state properties of a 1D system. Qualitatively, the two approaches agree with each other. However, we found an area where the RISBMF yields an artificial ground-state. Note that the mean-field method is worst for a 1D system. Therefore one shoud not judge from it the quality of the RISBMF method for the more general case.

Supervisor at Deutsches Elektronen-Synchrotron (DESY) in Zeuthen: Dr. Habil. Karl Jansen / Mit Tensor Netzwerken (TN) untersuchen wir auf einem hexagonalen Gitter das Hubbard-Modell mit einem chemischen Potential. Wir zeigen, dass ein TN als Ansatz für die Zustände des Modells benutzt werden kann und präsentieren die berechneten Eigenschaften bei niedrigen Energien. Unser Algorithmus wendet eine imaginäre Zeitentwicklung auf einen fermionischen projected engangled pair state (PEPS) auf einem endlichen Gitter mit offenen Randbedingungen an. Der Ansatz kann auf einen spezifischen fermionischen Paritätssektor beschränkt werden, was es uns ermöglicht, den Grundzustand und den Zustand mit einem Elektron weniger zu simulieren. Mehrere in unserer Arbeit entwickelte Verbesserungen des Algorithmus führen zu einer erheblichen Steigerung der Effizienz und Genauigkeit. Wir messen Erwartungswerte mit Hilfe eines boundary matrix product state. Wir zeigen, dass Observablen in dieser Näherung mit einer weniger starken Trunkierung, als in der Literatur erwartet wird, berechnet werden können. Dies führt zu einer erheblichen Reduzierung der numerischen Kosten des Algorithmus. Für verschiedene Stärken der lokalen Wechselwirkung, sowie für mehrere chemische Potentiale berechnen wir die Energie, die Teilchenzahl und die Magnetisierung mit guter Genauigkeit. Wir zeigen die Abhängigkeit der Teilchenzahl vom chemischen Potential und berechnen die Energielücke. Wir demonstrieren die Skalierbarkeit zu großen Gittern mit bis zu 30 × 15 Gitterpunkten und machen Vorhersagen in einem Teil des Phasenraums, der für Monte-Carlo nicht zugänglich ist. Allerdings finden wir auch Limitierungen des Algorithmus aufgrund von Instabilitäten, die die Berechnungen im Paritätssektor behindern, welcher orthogonal zum Grundzustand ist. Wir diskutieren Ursachen und Indikatoren und schlagen Lösungen vor. Unsere Arbeit bestätigt, dass TN genutzt werden können, um den niederenergetischen Sektors des Modells zu erforschen. Dies eröffnet den Weg zu einem umfassenden Verständnis des Phasendiagramms. / Using tensor network (TN) techniques, we study the Hubbard model on a honeycomb lattice with a chemical potential, which models the electron structure of graphene. In contrast to Monte Carlo methods, TN algorithms do not suffer from the sign problem when a chemical potential is present. We demonstrate that a tensor network state can be used to simulate the model and present the calculated low energy properties of the Hubbard model. Our algorithm applies an imaginary time evolution to a fermionic projected entangled pair state (PEPS) on a finite lattice with open boundary conditions. The ansatz can be restricted to a specific fermionic parity sector which allows us to simulate the ground state and the state with one electron less. Several improvements of the algorithm developed in our work lead to a substantial performance increase of the efficiency and precision. We measure expectation values with a boundary matrix product state and show that observables can be calculated with a lower bond dimension of this approximation than expected from the literature. This decreases the numerical costs of the algorithm significantly. For varying onsite interactions and chemical potentials we calculate the energy, particle number and magnetization with good precision. We show the dependence of the particle number on the chemical potential and compute the single particle gap. We demonstrate the scalability to large lattices of up to 30 × 15 sites and make predictions in a part of the phase space that is not accessible to Monte Carlo methods. However, we also find limitations of the algorithm due to instabilities that spoil the calculations in the parity sector orthogonal to the ground state. We discuss the causes and indicators of such instabilities and propose solutions. Our work validates that TNs can be utilized to study the low energy properties of the Hubbard model on a honeycomb lattice with a chemical potential, thus opening the road to finally understand its phase diagram.

No description available.