Dephasing in strongly disordered interacting quantum wires

TL;DR

This paper investigates how dephasing manifests in time-dependent variances of local observables in disordered quantum wires, revealing disorder-dependent power-law decay and linking it to localization length and multifractality.

Contribution

It introduces a novel analysis of dephasing via variances of local observables in the disordered Hubbard model, connecting decay exponents to localization and multifractal properties.

Findings

Variance decay follows a power law with disorder-dependent exponent

Decay exponent relates to localization length and multifractality

Comparison of exact and Hartree-Fock results elucidates damping mechanisms

Abstract

Many-body localization is a fascinating theoretical concept describing the intricate interplay of quantum interference, i.e. localization, with many-body interaction induced dephasing. Numerous computational tests and also several experiments have been put forward to support the basic concept. Typically, averages of time-dependent global observables have been considered, such as the charge imbalance. We here investigate within the disordered spin-less Hubbard ($t-V$) model how dephasing manifests in time dependent variances of observables. We find that after quenching a N\'eel state the local charge density exhibits strong temporal fluctuations with a damping that is sensitive to disorder $W$: variances decay in a power law manner, $t^{-\zeta}$, with an exponent $\zeta(W)$ strongly varying with $W$. A heuristic argument suggests the form, $\zeta\approx\alpha(W)\xi_\text{sp}$, where $\xi_\text{sp}(W)$ denotes the noninteracting localization length and $\alpha(W)$ characterizes the multifractal structure of the dynamically active volume fraction of the many-body Hilbert space. In order to elucidate correlations underlying the damping mechanism, exact computations are compared with results from the time-dependent Hartree-Fock approximation. Implications for experimentally relevant observables, such as the imbalance, will be discussed.