Genus-protected higher-order topological phases
Shahroze Shahab, Hui Liu, Daniel Varjas, Ion Cosma Fulga · 2026
Read Genus-protected higher-order topological phases by Shahroze Shahab, Hui Liu, Daniel Varjas, Ion Cosma Fulga (2026) 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-2605-06383.txt.
Introduction
Shahroze Shahab and three co-authors construct higher order topological phases protected by the shape of the system itself. Such phases host protected boundary modes at corners or hinges, one dimension below the usual edge states. They were understood as protected by lattice symmetries, so breaking those symmetries should let the boundary modes be removed without closing the bulk energy gap, a vulnerability this work designs around entirely.
Their construction schemes protect the phases with only the bulk gap, fundamental local symmetries, and the global genus of the system's shape, meaning its number of holes. No crystalline symmetry is needed at any point. As long as the fundamental local symmetries hold, no purely surface perturbation can remove the boundary states. The result ties topological protection to the topology of the sample's own geometry.
Catalog record
| CREATOR | Shahroze Shahab, Hui Liu, Daniel Varjas, Ion Cosma Fulga |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.06383
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Genus-protected higher-order topological phases Shahroze Shahab, Hui Liu, Daniel Varjas, Ion Cosma Fulga Abstract Higher-order topological phases (HOTPs) feature protected gapless modes on boundaries of higher codimension, such as the corners or hinges of a crystal. They are understood as being protected by lattice symmetries: If the latter are broken, it becomes possible to remove the boundary modes without closing the bulk gap. In this work, we present construction schemes for HOTPs protected solely by the bulk gap, by fundamental symmetries, and by the global topology of the system shape (its genus, or number of holes), independent of any crystalline symmetries. As long as the fundamental local symmetries are preserved, the resulting boundary states cannot be removed by any purely-surface perturbation. I Introduction Strong topological insulators and superconductors (TIs/TSCs) are characterized by a gapped bulk spectrum accompanied by robust gapless boundary modes 21 ; 37 . These boundary states are protected by fundamental local symmetries: time-reversal, particle-hole, and chiral symmetry 2 ; 42 ; 26 . The possible gapped Hamiltonians can be systematically organized according to their symmetry content within the ten-fold way classification 2 ; 42 ; 26 ; 39 . A recent development is the discovery of higher-order topological phases (HOTPs) 6 ; 5 ; 40 , which extend this paradigm by hosting gapless excitations on boundaries of codimension n>1 . In such systems, the full boundary need not be gapless. For instance, a two-dimensional (2D) second-order topological insulator features a gapped bulk and gapped edges, but gapless, zero-dimensional topologically protected modes appear on the boundary, for instance at the corners 5 ; 6 . In three dimensions (3D), second-order phases host gapless 1D states forming closed loops on the 2D boundary 40 , and third-order phases exhibit gapless 0D modes, e.g. at the corners 6 ; 5 . More generally, the boundary of a n^{\text{th}} -order TI/TSC in d dimensions is gapped except for a (d-n) -dimensional manifold that hosts protected gapless modes. A paradigmatic example is the Benalcazar-Bernevig-Hughes (BBH) model 6 , where sublattice symmetry together with rotation symmetry lead to robust, mid-gap corner states. Lattice symmetries play a central role in the classification of HOTPs. According to the classification of Refs. 19 ; 53 ; 54 , the latter fall within two classes: intrinsic and extrinsic. Intrinsic HOTPs are bulk topological phases which explicitly require lattice symmetries, e.g. mirror, rotation, or inversion for their protection. Provided that the system globally respects these spatial symmetries, it is not possible to remove the gapless boundary modes without closing the bulk gap. Extrinsic HOTPs, on the other hand, are neither bulk phases, nor do they require any lattice symmetry. As explained in Ref. 54 , they can be understood as being the result of strong topological phases appearing on the boundary, rather than in the bulk. For example, if one of the facets of a 3D cube-shaped crystal enters a Chern insulating phase, chiral hinge modes will appear on the hinges bordering that facet 54 . Expanding the family of possible HOTPs has been a subject of intense research in the last decade, resulting in generalizations to a wide range of systems. These include aperiodic HOTPs protected by rotation symmetries that do not occur in crystals 56 ; 9 , higher-order topological semimetals 30 ; 14 ; 46 ; 59 , or systems with a so-called mixed-order topology 47 . Various systems have been put forward for their experimental realization, including elemental bismuth 41 , SnTe 40 , {\mathrm{EuIn}}_{2}{\mathrm{As}}_{2} 64 , {\mathrm{MnBi}}_{2n}{\mathrm{Te}}_{3n+1} 69 , as well as a wide range of meta-material platforms 65 ; 35 ; 43 ; 60 ; 63 ; 4 ; 58 ; 48 ; 15 ; 61 ; 70 ; 67 . Despite this rapid progress and the large resulting variety of HOTPs, the intrinsic/extrinsic distinction has remained central to their classification. Here, we explore a family of HOTP models that show a combination of intrinsic and extrinsic features. On the one hand, their gapless boundary modes cannot be removed by any surface gap closing, provided that the fundamental local symmetries are preserved. On the other hand, these boundary modes do not require lattice symmetries for their protection. The combination of these features originates from the nontrivial global topology of the system and its boundary. In this sense, our work parallels those which show how strong topological phases can show different properties when the system shape is nontrivial, e.g. in Möbius strip 23 ; 8 or Klein bottle geometries 49 ; 38 . Here, instead, we show that HOTPs can go beyond the intrinsic/extrinsic classification in systems with nonzero genus (or a nonzero number of holes). Note that we use the word “genus” accurately in the case of 3D HOTPs, where it represents the number of handles of a closed 2D surface 22 . For the 2D systems we consider, however, the genus is strictly speaking zero, so we will use the letter g and the term “genus” to refer to the number of holes present in the system (the number of boundary components minus one), as common in the physics literature. Figure 1: Second-order states on different geometries. (a) On the disk, zero modes can be gapped out by bringing them together along the edges. (b) In a Corbino-disk geometry, the zero modes cannot be removed without closing the bulk gap, simply because they occur on two disjoint edges. (c) On the sphere, the 1D edge mode can be moved along the surface until it shrinks to a point (e.g. at the north pole) and gaps out. (d) In a torus-shaped 3D system, propagating modes on non-contractible loops cannot be removed without traversing the bulk. A simple illustration of our main idea is shown in Fig. 1 . Panel (a) shows a typical 2D second-order HOTP. Both the bulk and the edge are gapped, while two 0D gapless modes (red) appear at different points on the boundary. For this topological phase to be protected by the bulk gap (as opposed to the surface gap), an additional lattice symmetry is required to prevent the modes from being moved along the boundary and pairwise annihilated. Examples include inversion, a twofold rotation symmetry, or a mirror symmetry along the line connecting the two zero modes. In contrast, no lattice symmetry is required in panel (b). The Corbino disk has one hole ( g=1 ), hosting two disjoint edges. As such, if one zero mode forms on each edge, the two cannot be made to overlap by any perturbation that is restricted to the system boundary. The same general principle applies also in higher-dimensional systems, as shown in panels (c, d) for the case of 3D HOTPs. Genus g=0 systems (the sphere) require an additional lattice symmetry, lest the 1D topological mode be shifted along the surface until it contracts to a point and gaps out. The 3D torus geometry (genus g=1 ), however, naturally features noncontractible loops. A chiral edge mode formed along the red loop in Fig. 1 (d) cannot be gapped out without passing through the bulk, regardless of any symmetries. In the following, we will show how genus-protected topological phases (GPT phases) can be constructed, focusing on concrete examples of HOTPs in