There has been interest in the spin transport properties of the Aubry-Andr'e-Harper model at high temperatures under weak integrability breaking, in particular for small interactions or small fields. We present results that show that the model is diffusive, meaning that the claimed superdiffusion is a finite-size effect.
Recent studies of quantum many-body dynamics uncover many intriguing phenomena of thermalization and localization in the presence of a quasiperiodic potential. In this paper, the authors study nonequilibrium steady states of the one-dimensional interacting Aubry-Andre-Harper model by taking advantage of tensor network-based algorithms. In a weak potential regime, they find a superdiffusive to subdiffusive transition in spin transport within the thermal phase. The correlation structure of the steady states derived from the quantum mutual information shows an unusual oscillation pattern before reaching the localization transition. Our results suggest a rich phase diagram for the model.
In a recent study of the transport properties in the Aubry-Andr\'e-Harper model [Phys. Rev. B 100, 195143 (2019)], the authors point out the failure of the Hellmann-Feynman theorem (HFT) in the case of degenerate states. They try to support such a statement from earlier papers that prove exactly the opposite. It appears that the application of the HFT to degenerate states has not been well understood. For this reason, in this Comment we illustrate the correct application of the HFT to the model just mentioned.
We study a recently discussed XX spin chain with nonlocal dephasing [Y.-P. Wang et al., arXiv:2310.03069] in a steady-state boundary-driven setting, confirming superdiffusive magnetization transport in the thermodynamic limit. The emergence of superdiffusion is rather interesting as the Lindblad operators causing it are a coherent sum of two terms, each of which would separately cause diffusion. One therefore has a quantum phenomenon where a coherent sum of two diffusive terms results in superdiffusion. We also study perturbations of the superdiffusive model, finding that breaking the exact form of dissipators, as well as adding interactions to the XX chain, results in superdiffusion changing into diffusion.
There are two well-established ways of characterizing transport properties. The first is via the spread of correlations in an isolated system. The second is through the system-size dependence of the current, when a potential bias is applied by connecting the two ends of the system to reservoirs. These are two different experiments, but usually the result of one experiment can be inferred from that of the other. This work presents a surprising counterexample. The authors show that, at the critical point of the Aubry-Andr\'e-Harper model, transport in the presence of reservoirs is drastically different from transport in the isolated system. The scaling of current with system size shows clear subdiffusive behavior, while the spread of correlations does not show any subdiffusive behavior. It actually hints at superdiffusion. This implies that the understanding of the relation between transport in open and in isolated systems needs to be reexamined.