English
 
Help Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT

Released

Preprint

Operator Spreading and the Absence of Many-Body Localization

MPS-Authors
/persons/resource/persons185050

Weisse,  A.
Max Planck Institute for Mathematics, Max Planck Society;

External Resource
Fulltext (restricted access)
There are currently no full texts shared for your IP range.
Fulltext (public)

2401.08031.pdf
(Preprint), 2MB

Supplementary Material (public)
There is no public supplementary material available
Citation

Weisse, A., Gerstner, R., & Sirker, J. (submitted). Operator Spreading and the Absence of Many-Body Localization.


Cite as: https://hdl.handle.net/21.11116/0000-000F-54B9-2
Abstract
We consider the spreading of a local operator A in Euclidean time in one-dimensional many-body systems with Hamiltonian H by calculating the k-fold commutator [H,[H,[...,[H,A]]]]. We derive general bounds for the operator norm of this commutator in free and interacting fermionic systems with and without disorder. We show, in particular, that in a localized system the norm does grow at most exponentially and that the contributions of operators to the total norm are exponentially suppressed with their length. We support our general results by considering one specific example, the XXZ chain with random magnetic fields. We solve the operator spreading in the XX case without disorder exactly. For the Anderson and Aubry-André models we provide strict upper bounds. We support our results by symbolic calculations of the commutator up to high orders. For the XXX case with random magnetic fields, these symbolic calculations show a growth of the operator norm faster than exponential and consistent with the general bound for a non-localized system. Also, there is no exponential decay of the contribution of operators as function of their length. We conclude that there is no indication for a many-body localization transition. Finally, we also discuss the differences between the interacting and non-interacting cases when trying to perturbatively transform the microscopic to an effective Hamiltonian of local conserved charges by consecutive Schrieffer-Wolff transformations. We find that such an approach is not well-defined in the interacting case because the transformation generates ∼4ℓ terms connecting sites a distance ℓ apart which can overwhelm the exponential decay with ℓ of the amplitude of each individual term.