Localization for the random displacement model

From MaRDI portal



Abstract: We prove spectral and dynamical localization for the multi-dimensional random displacement model near the bottom of its spectrum by showing that the approach through multiscale analysis is applicable. In particular, we show that a previously known Lifshitz tail bound can be extended to our setting and prove a new Wegner estimate. A key tool is given by a quantitative form of a property of a related single-site Neumann problem which can be described as "bubbles tend to the corners".


The random displacement model is a random Schrödinger operator in the continuum with a random perturbation of the periodic potential. In this article, the authors prove spectral localization (a.s. pure point spectrum and exponentially decaying eigenfunctions) and dynamical localization in an interval above the ground state energy. A key ingredient is previous work by two of the authors characterizing the configurations of random displacements that minimize the ground state energy. This is turn is derived from a careful analysis of the ground state energy of Schrödinger operators in a box with von Neumann boundary conditions, which shows that the potential that minimizes the ground state energy is centered in a corner. The structure of the localization proof is standard: (i) smallness of the integrated density of states and (ii) Wegner estimate, used as inputs to the multiscale method. While (i) \& (ii) are proved in details in the paper and both use the a priori knowledge on the ground state energy, the running of the multiscale method from them is known and only briefly discussed for pedagogical reasons.



Cites work


Cited in
(36)








This page was built for publication: Localization for the random displacement model

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q410098)