Disorder, Localization & Slow Quantum Dynamics

Disorder changes quantum dynamics by creating a distribution of local environments rather than a single representative relaxation time. The central question is not simply whether a sample conducts. It is how transport, entanglement, spectral statistics, and memory respond on different length and time scales. Agarwal's work in this area connects rare-region models to numerical observables and asks what survives when a localized system is driven. These results remain useful even when the interpretation of an ultimate many-body localized phase is unsettled.

Figure to read

Disorder-dependent dynamical exponents in the original XXZ-chain study. Figure 1 in the source paper. Agarwal, Kartiek; Gopalakrishnan, Sarang; Knap, Michael; Mueller, Markus; Demler, Eugene. Anomalous diffusion and Griffiths effects near the many-body localization transition, Phys. Rev. Lett. 114, 160401 (2015). Figure linked rather than reproduced; the archived manuscript does not provide an explicit open reproduction license.

Compare the transport and relaxation exponents with the spectral diagnostic. Treat the plotted boundary as the interpretation of the finite-size 2015 study, not a settled modern phase boundary.

The reference picture and its exceptions

In a generic interacting system, conservation laws constrain relaxation while local observables approach values consistent with statistical mechanics. The eigenstate thermalization hypothesis supplies a framework for that expectation; it does not require every observable to relax at the same rate. A conserved density must physically move, whereas a nonconserved local operator can decay without transporting a conserved quantity. This distinction matters when interpreting an apparent power law or a long relaxation plateau. 1

Randomness introduces an additional hierarchy. A long interval that is unusually insulating can impede transport through an otherwise conducting one-dimensional chain. Its probability may decrease exponentially with length while its crossing time increases exponentially. Combining those exponentials produces broad, approximately power-law distributions of waiting times. The resulting slow dynamics are therefore tied to statistically rare structures rather than a typical site's local properties. Agarwal and collaborators developed this perspective for anomalous diffusion near the localization transition. 2

What the contributions establish

The 2015 study links time-domain spin relaxation, frequency-dependent conductivity, and broad resistance distributions through a common scaling picture. The subsequent analysis of Griffiths effects broadens the question to different dimensions, conservation laws, and observables. It explains why rare regions may dominate one measurement while ordinary hydrodynamic long-time tails dominate another. This is a stronger and more useful statement than identifying all slow dynamics with one mechanism. 3

A related branch studies random Heisenberg networks and low-frequency magnetic noise. There the theoretical task includes renormalizing the measured operator as well as the Hamiltonian: which clusters a sensor sees affects its spectrum. That work connects collective spin dynamics to a possible microscopic source of broad noise, while leaving material-specific identification as a separate experimental question. See random spin networks for the role of geometry and probe weighting. 4

Periodic driving tests another boundary of the localization picture. In the strongly driven Anderson problem, extensive motion within one drive cycle need not imply unbounded transport over many cycles. Floquet eigenstates and long-time interference supply information that a short trajectory alone cannot. This is a single-particle problem and should not be used as a direct demonstration of interacting many-body localization. 5

The present research landscape

Contemporary work places particular emphasis on thermal avalanches and finite-size drift. A thermal inclusion inside a localized background raises the reverse rare-region question: can the inclusion absorb neighboring degrees of freedom and grow? Avalanche theories identify a mechanism by which a locally thermal region may destabilize localization. Numerical and experimental studies must then decide whether accessible systems are observing an asymptotic phase or a very long crossover. 6 7

This debate changes how older figures should be read. A disorder-dependent exponent extracted from a finite chain is evidence about the simulated regime. It is not, by itself, a universal critical exponent or proof that the same scaling continues indefinitely. Disorder averages may also obscure the difference between typical and rare samples. A useful comparison should report system size, evolution time, boundary conditions, conservation laws, and the distribution of outcomes rather than only a fitted mean.

Connections and open questions

The four pages in this branch separate thermalization diagnostics, Griffiths bottlenecks, random spin noise, and driven disorder. Together they suggest an experimental strategy: compare several observables that respond differently to the proposed slow mechanism. The unresolved challenge is to connect tractable models to long-time behavior without overstating finite-size evidence. That same challenge reappears in quantum sensing, where an apparently simple noise exponent may encode several distinct microscopic processes.

References

1. D'Alessio, Luca; Kafri, Yariv; Polkovnikov, Anatoli; Rigol, Marcos. From Quantum Chaos and Eigenstate Thermalization to Statistical Mechanics and Thermodynamics. Adv. Phys. 65, 239 (2016). Open manuscript.
2. Agarwal, Kartiek; Gopalakrishnan, Sarang; Knap, Michael; Mueller, Markus; Demler, Eugene. Anomalous diffusion and Griffiths effects near the many-body localization transition. Phys. Rev. Lett. 114, 160401 (2015). Open manuscript.
3. Gopalakrishnan, Sarang; Agarwal, Kartiek; Demler, Eugene; Huse, David A.; Knap, Michael. Griffiths effects and slow dynamics in nearly many-body localized systems. Phys. Rev. B 93, 134206 (2016). Open manuscript.
4. Agarwal, Kartiek; Demler, Eugene; Martin, Ivar. $1/f^\alpha$ noise and generalized diffusion in random Heisenberg spin systems. Phys. Rev. B 92, 184203 (2015). Open manuscript.
5. Agarwal, Kartiek; Ganeshan, Sriram; Bhatt, R. N.. Localization and transport in a strongly driven Anderson insulator. Phys. Rev. B 96, 014201 (2017). Open manuscript.
6. Thiery, Thimothée; Huveneers, François; Müller, Markus; De Roeck, Wojciech. Many-body delocalization as a quantum avalanche. Phys. Rev. Lett. 121, 140601 (2018). Open manuscript.
7. Sierant, Piotr; Lewenstein, Maciej; Scardicchio, Antonello; Vidmar, Lev; Zakrzewski, Jakub. Many-body localization in the age of classical computing
   *. Reports on Progress in Physics 88, 026502 (2025). Open manuscript.

Independent research synthesis. Literature checked 5 September 2026; preprints are identified in the references.