Peter D> Hislop,Werner Kirsch,M. Krishna
Abstract:We provide a simple proof of dynamical delocalization, that is, time-increasing lower bounds on quantum transport for discrete, one-particle Schrodinger operators on $\ell^2 (\mathbb{Z}^d)$, provided solutions to the Schrodinger equation satisfy certain growth conditions. The proof is based on basic resolvent identities and the Combes-Thomas estimate on the exponential decay of the Green's function. As a consequence, we prove that generalized eigenfunctions for energies outside the spectrum of $H$ must grow exponentially in some directions. We also prove that if $H$ has any absolutely continuous spectrum, then the Schrodinger operator exhibits dynamical delocalization. We apply the general result to $\Gamma$-trimmed Schrodinger operators, with periodic $\Gamma$, and prove dynamical delocalization for these operators. These results also apply to the $\Gamma$-trimmed Anderson model, providing a random, ergodic model exhibiting both dynamical localization in an energy interval and dynamical delocalization.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is about the phenomenon of dynamical delocalization in quantum transport. Specifically, the authors provide a simple proof method to prove the dynamical delocalization of the discrete one - particle Schrödinger operator on \( \ell^2(\mathbb{Z}^d) \) under certain conditions, that is, the lower bound of quantum transport as time increases. These conditions are related to the growth characteristics of the solutions of the Schrödinger equation. Through basic analytic identities and Combes - Thomas estimates, the authors prove that if the Hamiltonian \( H \) has generalized eigenfunctions outside its spectrum, these eigenfunctions must grow exponentially in certain directions. Moreover, if \( H \) has any absolutely continuous spectrum, the Schrödinger operator exhibits dynamical delocalization.
The paper also applies these general results to \( \Gamma \)-pruned Schrödinger operators and proves the dynamical delocalization of these operators. These results also apply to the \( \Gamma \)-pruned Anderson model, providing a random, ergodic model that exhibits both dynamical localization and dynamical delocalization in energy intervals.
### Main contributions
1. **Simple proof of dynamical delocalization**: The authors provide a method based on basic analytic identities and Combes - Thomas estimates to prove dynamical delocalization.
2. **Growth conditions of generalized eigenfunctions**: The authors prove the exponential growth of generalized eigenfunctions in certain directions, and this growth is related to dynamical delocalization.
3. **Application of \( \Gamma \)-pruned models**: The authors apply these results to \( \Gamma \)-pruned Schrödinger operators and Anderson models and prove the dynamical delocalization of these models.
### Mathematical expressions
- **Definition of dynamical delocalization**: If a moment of the position operator increases unboundedly with time, the Schrödinger operator is said to exhibit dynamical delocalization.
- **Growth conditions of generalized eigenfunctions**: For almost all \( E\in J \), there exist generalized eigenfunctions \( \psi_E \) and a constant \( C_E>0 \) such that \( |\psi_E(x)| < C_E \) holds for almost all \( E\in J \).
- **Sufficient conditions for dynamical delocalization**: If there exists a set \( J\subset \mathbb{R} \) of positive measure, for almost all \( E\in J \), there exist generalized eigenfunctions \( \psi_E \) satisfying certain growth conditions, then the moment \( M_q(T) \) of the position operator satisfies \( M_q(T)\geq C_J T \).
### Application examples
- **\( \Gamma \)-pruned Anderson model**: The authors prove that for some \( \Gamma \)-pruned Anderson models, even if they exhibit dynamical localization in certain energy intervals, they still exhibit dynamical delocalization overall.
These results not only expand the understanding of quantum transport but also provide new tools and methods for studying the dynamical behavior in complex systems.