A single atom in a one-dimensional optical superlattice with residual harmonic confinement can localize through three distinct mechanisms: topological edge states protected by the Zak phase, an intermediate-trapping-regime effective four-level localization on the central sites, and trap-dominated quasi-classical pairing of eigenstates. The source study (arXiv:2602.04395, 2026) established these mechanisms for the non-interacting atom via exact diagonalization and a tight-binding approximation, mapped their static parameter dependence at sampled values, and illustrated dynamical readout discriminants only at one superlattice parameter set. It explicitly left open whether the edge-site localization persisting beyond the critical trapping frequency admits a topological classification through a generalized topological index, and listed as future perspectives dynamical protocols for observing the mechanisms, the incorporation of interactions, and the effect of other periodic or confining potentials. The problem asks for a topologically complete account of the localization that persists beyond the critical trapping frequency in the finite, trapped setting, together with the source-discussed extensions: readout protocols distinguishing the localization mechanisms beyond the single illustrative parameter set, and the interacting many-body regime. These three threads are independently resolvable but jointly posed parts of the problem: the generalized topological-index classification of the beyond- edge localization, the parameter-general dynamical readout protocols, and the interacting many-body extension (resolved by determining whether and in which interaction regime the phenomenology persists).
Background
Ultracold atoms in standing-wave light fields form optical lattices: periodic potentials whose depth, geometry, and any superimposed confinement are tunable with high precision. Superimposing two standing waves of commensurate wavelengths produces an optical superlattice — a chain of alternating high and low barriers — which realizes the Su-Schrieffer-Heeger (SSH) model, the canonical one-dimensional tight-binding model with two sites per unit cell and alternating hopping amplitudes (Su, Schrieffer and Heeger 1979). Depending on the ratio of the hoppings, the SSH model is in a topologically trivial or non-trivial phase, diagnosed in the continuum by the Zak phase (Zak 1989): the Berry phase accumulated by a Bloch eigenstate over one traversal of the Brillouin zone, which is quantized to 0 or by inversion or chiral symmetry (the symmetry that pairs every energy eigenvalue with one at , here arising from the two-sublattice structure of the model) and serves as the one-dimensional topological invariant. In the non-trivial phase and under open boundaries, the bulk-boundary correspondence — the principle that a bulk topological invariant forces protected states at the system's boundary — predicts topological edge states (TESs): eigenstates exponentially localized at the two ends of the chain and pinned near zero energy by chiral symmetry. In a finite continuous optical superlattice, sharp hard-wall cutoffs deform these edge states; a linear extension of the potential at the boundaries can approximately restore chiral symmetry and the TESs (Katsaris et al. 2024). Real experiments never realize a perfectly flat geometry: a residual harmonic trap is always present, and in a plain optical lattice it is known to break translational invariance and produce energetically paired, spatially localized states at large , a phenomenon related to Wannier–Stark localization — the localization of Bloch states in a lattice subject to a linear potential gradient (Wannier 1960); wave-packet dynamics in such harmonically enveloped lattices remain an active topic (Ali, Holthaus and Meier 2025). The interplay of this trap with superlattice topology in a finite system was analyzed by Katsaris, Englezos, Weitenberg, Diakonos and Schmelcher (arXiv:2602.04395) for a single atom, using exact diagonalization (numerical diagonalization of the continuum Hamiltonian on a finite-difference grid) and a tight-binding reduction to a lowest-band hopping model with on-site energies and hoppings extracted from the Wannier-like orbitals (localized wave functions spanning the lowest Bloch band). They identify three regimes as the trapping frequency grows at fixed superlattice parameters (high barrier , barrier ratio , number of unit cells , phase fixing the trivial or non-trivial Zak-phase configuration, plus optional boundary extension and trap-center alignment): (i) a low- regime in which the TESs of the topological configuration survive up to a critical trapping frequency ; (ii) an intermediate regime in which, in the topological configuration, the four lowest eigenstates form an effective four-level system localized on the two central sites — a novel localization mechanism arising from the interplay of finite size and the trap rather than topology alone; and (iii) a high- trap-dominated regime of quasi-classical pairing and localization of all lower-band states, as in an ordinary lattice. Their static analysis maps these regimes across sampled superlattice parameters (, , up to 10, with the recoil energy of the superlattice lasers at wavenumber ), but the illustrative time-evolution studies — single-site initial states whose transport dynamics exhibit clearly distinct frequencies for each mechanism — are performed only at , , , with representative trapping frequencies ( and ). Crucially, beyond the edge sites remain occupied by localized states, but this localization is trap-dominated and no longer captured by the Zak-phase invariant; the authors state that classifying it would require a generalized topological index, which they leave open; their stated perspectives further include the interacting many-body extension and dynamical observation protocols, while the dynamical readout discriminants are demonstrated only at the single illustrative parameter set.
Problem Statement
For a single atom — and, as the source paper explicitly poses for the future, for interacting many-body systems — in the finite optical superlattice with harmonic confinement (with the two-frequency superlattice potential parameterized by the high barrier , the barrier ratio , the number of unit cells , and the phase fixing the trivial or non-trivial Zak-phase configuration; an optional boundary extension; and the trap-center displacement relative to the superlattice, with the aligned case), the central open question left by the source work is whether the edge-site localization that persists beyond the critical trapping frequency at which the topological edge states cease to exist is topological or not. Resolving it requires determining whether a generalized topological index exists that is well defined for this finite, non-translationally-invariant, trapped system, reduces to the Zak-phase classification in the untrapped limit, and correctly tracks the fate of the edge states; a demonstration that no such index exists is an equally valid resolution. The source work mapped the three localization mechanisms across sampled values of the superlattice and trapping parameters (, , , , , boundary extension, trap alignment) and demonstrated the universality of the intermediate-confinement four-level system; the further directions it raises are readout protocols — dynamical discriminants such as the characteristic transport frequencies of single-site initial states, or alternative measurable observables — that distinguish the mechanisms across the parameter range sampled in the source work and in the experimentally accessible regime, and the interacting many-body extension, determining, for a solver-chosen interacting model whose control parameter and particle number or filling are stated explicitly and justified, whether and in which interaction regime the phenomenology persists.
References
- A. Katsaris, I. A. Englezos, C. Weitenberg, F. K. Diakonos, and P. Schmelcher, "Mechanisms of localization in a finite harmonically confined optical superlattice", arXiv:2602.04395 (2026), https://arxiv.org/abs/2602.04395
- A. Katsaris, I. A. Englezos, C. Weitenberg, F. K. Diakonos, and P. Schmelcher, "Restoring the topological edge states in a finite optical superlattice", Phys. Rev. A 110, 013316 (2024), DOI: 10.1103/PhysRevA.110.013316, arXiv:2404.06924, https://arxiv.org/abs/2404.06924
- W. P. Su, J. R. Schrieffer, and A. J. Heeger, "Solitons in Polyacetylene", Phys. Rev. Lett. 42, 1698 (1979), DOI: 10.1103/PhysRevLett.42.1698, https://doi.org/10.1103/PhysRevLett.42.1698
- J. Zak, "Berry's phase for energy bands in solids", Phys. Rev. Lett. 62, 2747 (1989), DOI: 10.1103/PhysRevLett.62.2747, https://doi.org/10.1103/PhysRevLett.62.2747
- G. H. Wannier, "Wave Functions and Effective Hamiltonian for Bloch Electrons in an Electric Field", Phys. Rev. 117, 432 (1960), DOI: 10.1103/PhysRev.117.432, https://doi.org/10.1103/PhysRev.117.432
- U. Ali, M. Holthaus, and T. Meier, "Wave packet dynamics in parabolic optical lattices: From Bloch oscillations to long-range dynamical tunneling", Phys. Rev. Research 7, 013141 (2025), DOI: 10.1103/PhysRevResearch.7.013141, https://doi.org/10.1103/PhysRevResearch.7.013141
- F. K. Diakonos and P. Schmelcher, "Continuum limit of bipartite lattices -- The SSH model", arXiv:2509.11900 (2025), https://arxiv.org/abs/2509.11900