Anderson Localization in high dimensional lattices

Abstract

In this thesis, we investigate the behavior of Anderson Localization in high dimension. In the first part we study Levy Matrices (LMs), a Random Matrix model with long-range hopping presenting strong analogy with the problem of Anderson Localization on tree-like structure, representative of the limit of infinite dimensionality. We establish the equation determining the localization transition and obtain the phase diagram. We investigate then the unusual behavior of the delocalized phase. Using arguments based on supersymmetric field theory and Dyson Brownian motion we show that the eigenvalue statistics is the same one as of the Gaussian orthogonal ensemble in the whole delocalized phase and is Poisson-like in the localized phase. Our numerical analysis confirms this result, valid in the limit of infinitely large LMs, and provides information on the behavior of other observables like the wave-functions statistics. At the same time, numerical results also reveal that the characteristic scale governing finite size effects diverges much faster than a power law approaching the transition and is already very large far from it. This leads to a very wide crossover region in which the system looks as if it were in a mixed phase. In the second part we study numerically the behavior of the Anderson Model in dimension from 3 to 6 through exact diagonalization, Transfer Matrix method and an approximate Strong Disorder Renormalization Group technique. The results we find suggest that the upper critical dimension of Anderson Localization is infinite. Finally, we discuss the possible implications of this scenario on the anomalous behavior of the delocalized phase of models representative of the limit of infinite dimension, like Levy Matrices and the Anderson model on tree-like structures.