Modelagem computacional de sistemas de elétrons fortemente correlacionados

• Ano de defesa: 2016
• Autor(a) principal: Souza, Thiago Xavier Rocha de
• Orientador(a): Macêdo, Cláudio Andrade
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Sergipe
• Programa de Pós-Graduação: Pós-Graduação em Física
• País: Brasil
• Palavras-chave em Português: Física; Físico-química; Modelo de Hubbard; Algoritmos; Ferromagnetismo; Método do Lanczos; Fase quântica
• Palavras-chave em Inglês: Explicit Restarted Lanczos; Modified Explicit Restarted Lanczos
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::FISICA
• Link de acesso: https://ri.ufs.br/handle/riufs/5252

Critical phenomena study was for many years dominated by analysis of transitions generated by thermal fluctuations. This thermal fluctuations cease at T-0, however, quantum fluctuations does not end at zero temperature. These quantum fluctuations may, under certain conditions, trigger phase transitions. In this work the Hubbard model is used to study quantum state and quantum phase transitions in strongly correlated electron systems, considering the terms of intersite hopping and Coulomb repulsion intrasite. It was developed an algorithm based on Lanczos method to solving the Hubbard model applied in different types of lattices. Analysis of algorithms efficiency were made an was observed that the standard approaches to evaluate the properties of the ground state in the Hubbard model by Lanczos method presents convergence problems when there is a significant difference between hopping parameters and Coulomb interaction. This difference is very important since the energy convergence does not necessarily reflect in a convergence of the ground state. In this work are discussed several algorithms as standard Lanczos method, the Explicit Restarted Lanczos algorithm and the Modified Explicit Restarted Lanczos algorithm. A protocol based on these algorithms using the operator S2 as s stopping criterion was developed, since through this the operator it is possible to assess the error getting from the ground state itself. The algorithm based on the ERL provides better accuracy and it is 5 times faster compared with conventional ones. The MERL-based algorithm keeps the error at the last significant digit, and its processing time is about 2.5 times longer than the ERL-based algorithm, although it is still faster than the standard Lanczos method. These analyzes pave the way for a reliable and practical evaluation of the ground-state properties not only of the Hubbard model, but also for other manybodies quantum systems. The systems analyzed were clusters of polymeric lattice AB2 tipe, one-dimensional lattice considering nears and next nears neighbors hoppings and cluster of fcc lattice. All systems showed quantum state transitions. Through the study of the spin-spin correlations of the AB2 lattices clusters it was possible to analyze in detail the behavior of these spin-spin correlation functions between sublattices of a finite system. The analysis of one-dimensional lattice with next near neighbor made it possible to use an extrapolation method, which has determined that the quantum phase transition critical point, Uc/t = 4.7, from which the system changes from a paramagnetic behavior to a ferromagnetic behavior. In the fcc lattice clusters were examined the ground state energy as a function of the particle density showed a minimum value for all the structural sizes studied. The minimum energy decreases with increasing the interaction parameter U. It was observed that the ground state energy has a minimum at n = 0.6 for U/t = W, where W denotes the non-interacting bandwidth and the face-centered cubic structure is ferromagnetic. These results, when compared to the nickel properties, shown great similarity analysis in literature, made at finite temperature and support the results of Hirsh, which proposes that the interatomic interaction exchange is dominant to driving the system to a ferromagnetic phase.