Duality breaking, mobility edges, and the connection between topological Aubry-André and quantum Hall insulators in atomic wires with fermions
Bar Alluf, C. A. R. Sá de Melo · 2025
Read Duality breaking, mobility edges, and the connection between topological Aubry-André and quantum Hall insulators in atomic wires with fermions by Bar Alluf, C. A. R. Sá de Melo (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-13190.txt.
Introduction
Bar Alluf and a co-author explore what happens when a famous quantum model loses its perfect balance. The Aubry-André model traps particles in a wavy potential and is self-dual, a symmetry that forbids mobility edges, the energy borders between trapped and free states. Breaking that duality, the authors show, lets mobility regions appear and topological phases emerge.
By tuning how strongly the symmetry breaks, they map the new regions and tie the system to quantum Hall physics through a family of Hamiltonians. A laser phase plays the role of an extra dimension. Three kinds of topological phase appear depending on filling and duality breaking: ordinary, Aubry-André flavored, and a hybrid of the two. Special counting numbers, Chern numbers, label the phases.
Catalog record
| CREATOR | Bar Alluf, C. A. R. Sá de Melo |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 66 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-13190
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Duality breaking, mobility edges, and the connection between topological Aubry-André and quantum Hall insulators in atomic wires with fermions Bar Alluf, C. A. R. Sá de Melo Abstract It is well known that the Aubry-André model lacks mobility edges due to its energy-independent self-duality but may exhibit edge states. When duality is broken, we show that mobility regions arise and non-trivial topological phases emerge. By varying the degree of duality breaking, we identify mobility regions and establish a connection between Aubry-André atomic wires with fermions and quantum Hall systems for a family of Hamiltonians that depends on the relative phase of laser fields, viewed as a synthetic dimension. Depending on the filling factor and the degree of duality breaking, we find three classes of non-trivial phases: conventional topological insulator, conventional topological Aubry-André insulator, and unconventional (hybrid) topological Aubry-André insulator. Finally, we discuss appropriate Chern numbers that illustrate the classification of topological phases of localized fermions in atomic wires. I Introduction The concept of mobility edge pertains to an energy boundary that distinguishes between localized and extended states. The existence of mobility edges (MEs) holds immense significance in comprehending intricate localization phenomena observed in diverse physical systems, such as Anderson insulators [ 1 ] , photonic materials [ 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 ] , kicked rotor systems [ 11 ] , flatband networks [ 12 ] , and ultracold atoms with speckles [ 13 , 14 , 15 , 16 , 17 ] . In one dimension, an infinitesimal amount of random disorder is sufficient to cause localization and MEs are not present [ 18 ] . Mobility edges are also absent in the Aubry-André (AA) model [ 19 ] , describing localization in one-dimensional (1D) quasiperiodic potentials, due to a global duality that directly maps extended states to localized states and vice versa. There are some generalizations of the AA model that produce MEs: the addition of mean-field interactions [ 20 , 21 , 22 , 23 , 24 , 25 ] , modified quasiperiodicities [ 26 ] , beyond nearest-neighbor hopping [ 27 , 28 ] , and noninteracting systems with spin-orbit coupling (SOC) but without Rabi fields [ 29 , 30 , 31 ] . Furthermore, MEs may also be found in 1D speckle-disordered systems, such as {}^{87}\rm{Rb} , where it was shown that SOC and Rabi fields facilitate transport and hinder localization [ 32 ] . Motivated by the experimental observation of MEs in Aubry-André atomic wires including mean-field interactions with bosons [ 22 , 23 ] , we propose a mechanism for breaking the global duality of Aubry-André systems, by introducing SOC and Rabi fields, that produces not only MEs but also topological order in atomic wires with fermions. Our global duality-breaking mechanism can also be applied to study many-body localization of ultracold fermions in optical lattices [ 33 , 34 ] . Using the current platform of atomic wires with fermions [ 33 , 34 ] and adding duality-breaking fields, we investigate the phase diagrams for different filling factors, characterize the emergent mobility regions, and identify topological states. Depending on the sequence of MEs, we find three types of topological phases with midgap edge states. The first phase is a conventional topological insulator with a bulk gap separating two conducting bands. The second one is a conventional topological AA insulator with two insulating bands separated by a bulk gap. The third phase is an unconventional (hybrid) topological AA insulator with a bulk gap separating insulating and conducting bands. The last phase is a direct consequence of the existence of MEs. To characterize these topological phases, we introduce a fictitious dimension (the phase difference between weak and strong lasers) in AA atomic wires. We also compute the topological invariants (Chern numbers) of each phase by mapping our problem into a 2D system describing generalized quantum Hall insulators. The remainder of the paper is organized as follows. In II we describe the Hamiltonian used in our analysis. In III we discuss the localization properties of our Hamiltonian by analyzing the inverse participation ratio. In IV we perform a scaling analysis of the inverse participation ratio to be used for the determination of phase diagrams. In V we investigate various phase diagrams showing localized and extended regions. In VI we describe the connection between quantum Hall systems with spin-orbit and Rabi couplings and obtain Chern numbers to identify topological phases. In VII we compare our work with previous efforts. In VIII we summarize and present our conclusions. II Lattice Hamiltonian We start with the duality-broken AA Hamiltonian \mathcal{H}_{\varphi}=-\sum_{\braket{nm}ss^{\prime}}J_{nm}^{ss^{\prime}}c_{ns}^{\dagger}c_{ms^{\prime}}+\sum_{nss^{\prime}}\Gamma_{nn}^{ss^{\prime}}c_{ns}^{\dagger}c_{ns^{\prime}} (1) where c_{ns}^{\dagger} and c_{ns} describe fermions at lattice site n with spin s and \braket{nm} represents nearest neighbors. The first matrix is \bm{J}_{nm}=J_{nm}[\cos(k_{T}\delta x_{nm})\mathbf{I}+i\sin(k_{T}\delta x_{nm}\bm{\sigma}_{z})], (2) describing nearest-neighbor hopping, where k_{T} is the spindependent momentum transfer associated with SOC and \delta x_{nm}=x_{n}-x_{m} are displacements, where x_{n}=na , with a the optical lattice spacing determined by the wavelength \lambda_{S} of a strong laser [ 22 , 23 ] . Here \mathbf{I} is the identity and \bm{\sigma}_{i} is the Pauli matrix for i\in{x,y,z} . The second matrix is \bm{\Gamma}_{nn}=\Delta\cos(2\pi\beta n-\varphi)\mathbf{I}-h_{x}\bm{\sigma}_{x}, (3) where \Delta is a modulation created by a weak laser beam [ 22 , 23 ] of wavelength \lambda_{W} , \varphi=2(\phi_{S}-\phi_{W}) is twice the phase difference between the strong \phi_{S} and weak \phi_{W} lasers, \beta=\lambda_{S}/\lambda_{W} is the ratio of wavelengths, and h_{x} plays the role of the Rabi field. For definiteness, we set \beta=532/738 , compatible with experiments with {}^{40}\rm{K} [ 35 ] . In addition, using the spin-gauge transformation (SGT) \mathbf{c}_{n}=e^{ik_{T}x_{n}\bm{\sigma}_{z}}\tilde{\mathbf{c}}_{n} with \mathbf{c}_{n}=(c_{n\uparrow},c_{n\downarrow})^{T} , where T means transposition, we transfer \exp(ik_{T}\delta x_{nm}) from \bm{J}_{nm} to \bm{\Gamma}_{nn} , leading to \tilde{\mathcal{H}}_{\varphi}=-\sum_{\braket{nm}ss^{\prime}}\tilde{\bm{J}}_{nm}^{ss^{\prime}}\tilde{c}_{ns}^{\dagger}\tilde{c}_{ms^{\prime}}+\sum_{nss^{\prime}}\tilde{\Gamma}_{nn}^{ss^{\prime}}\tilde{c}_{ns}^{\dagger}\tilde{c}_{ns^{\prime}}. (4) Here \tilde{J}^{ss^{\prime}}_{nm} are the matrix elements of \tilde{\bm{J}}_{nm}=J_{nm}\mathbf{I}, (5) where \tilde{\bm{J}}_{nm} does not contain spin-dependent phases. The local spin-diagonal elements \tilde{\Gamma}^{ss}_{nn}=\Gamma^{ss}_{nn} remain invariant; however, the spin-off-diagonal elements become \tilde{\Gamma}_{nn}^{\uparrow\downarrow}=\Gamma_{nn}^{\uparrow\downarrow}e^{-2ik_{T}x_{n}}=-h_{x}e^{-2ik_{T}x_{n}} and \tilde{\Gamma}_{nn}^{\downarrow\uparrow}=\Gamma_{nn}^{\downarrow\uparrow}e^{2ik_{T}x_{n}}=-h_{x}e^{2ik_{T}x_{n}} . Using x_{n}=na , we write the matrix elements \tilde{\Gamma}_{nn}^{ss^{\prime}} in matrix form as \tilde{\bm{\Gamma}}_{nn}=\Delta\cos(2\pi\beta n-\varphi)\mathbf{I}-h_{+}(n)\bm{\sigma}_{+}-h_{-}(n)\bm{\sigma}_{-}, (6) where \bm{\sigma}_{\pm}=(\bm{\sigma}_{x}\pm i\bm{\sigma}_{y})/2 are spin raising (+) and lowering (-) matrices and h_{\pm}(n)=h_{x}e^{\mp i2k_{T}an} (7) are the helical Rabi fields with spatial variation controlled by the SOC parameter k_{T}a . The matrix in 6 can also be written in terms of the Pauli matrices \sigma_{x} and \sigma_{y} as \tilde{\bm{\Gamma}}_{nn}=\Delta\cos(2\pi\beta n-\varphi)\mathbf{I}-h_{x}(n)\bm{\sigma}_{x}-h_{y}(n)\bm{\sigma}_{y}, (8) where the local Rabi field components are h_{x}(n)=h_{x}\cos(2k_{T}an) (9) in the x direction in spin space and h_{y}(n)=h_{x}\sin(2k_{T}an) (10) in the y direction in spin space. Note that the SGT makes explicit in Eqs. ( 4 )-( 10 ) that the Hamiltonian \tilde{\mathcal{H}}_{\varphi} is \pi periodic in k_{T}a and that for h_{x}=0 the system’s energy is independent of the spin-orbit parameter k_{T}a , due to spin-gauge symmetry. Throughout the text, we define the filling factor as \nu=N_{st}/N , where N_{st} is the number of states and N is the number of sites. Using this definition, \nu ranges from 0 to 2, that is 0\leq\nu\leq 2 . Thus, \nu=1/2 corresponds to quarter filling, \nu=1 labels half filling, and \nu=2 describes full filling. Since we are interested in the localization properties of our Hamiltonian as energy E/J or filling factor \nu change, we investigate next the inverse participation