Magnetische Phasen im Hubbardmodel

Abstract

This work deals with the calculation of magnetic properties for strongly correlated electron systems. One of most important model in this field of physics is the Hubbard model. It describes the situation of, for example, transition metal oxides. This particular class of materials is known for magnetic order at low temperatures. The reason for this and at the same time the difficulty for theoretically describing them are strong electron-electron-interactions. For calculating the magnetic properties I used the dynamical mean field theory (DMFT), which relates the lattice model to an impurity model. Even this simplified model is highly non trivial. For solving it I used two different renormalization group approaches. I used the numerical renormalization group (NRG) and the density matrix renormalization group (DMRG) and also compared both methods with each other. These modern non-perturbative techniques are able to solve the impurity model, which then can be used for calculating the magnetic phases of the lattice models. In this work I mainly focused on the frustrated one-orbital Hubbard model and the two-orbital Hubbard model. Frustration represents a situation where the aimed magnetic solution cannot be stabilized for the used lattice. Frustration can, for example, be generated by long range hopping of the electrons. This can lead to a situation where new magnetic phases are stable. For systematically analyzing the effects of frustration, I performed calculations for a Bethe lattice with nearest and next nearest neighbor hopping, for which I varied the strength of the interaction and the next nearest neighbor hopping. Another very important aspect for describing transition metal oxides are orbital degrees of freedom. For analyzing the influence of these degrees of freedom I used the two-orbital Hubbard model, which shows new effects.