🔬 mesoscale physics

Genus-protected higher-order topological phases

This paper presents construction schemes for higher-order topological phases that are protected solely by the bulk gap, fundamental symmetries, and the system's global genus, eliminating the need for crystalline symmetries to sustain robust boundary states.

Original authors: Shahroze Shahab, Hui Liu, Daniel Varjas, Ion Cosma Fulga

Published 2026-05-08

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Shahroze Shahab, Hui Liu, Daniel Varjas, Ion Cosma Fulga

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a crystal not as a solid block of stone, but as a complex, multi-layered city built on a grid. In this city, the "streets" (the edges of the crystal) are usually safe and empty, while the "buildings" (the bulk) are full of activity. However, in a special type of city called a Higher-Order Topological Phase (HOTP), the rules change. Here, the "streets" are actually closed for construction, but the corners of the city blocks or the hinges where walls meet become special, bustling hubs where energy can flow freely without getting stuck.

For a long time, scientists believed these special hubs only existed because the city was built with perfect symmetry—like a perfect square grid where every corner looks exactly like every other corner. If you broke that symmetry (say, by making the city rectangular instead of square), the hubs would disappear.

The Big Discovery
This paper introduces a new kind of city where these special hubs exist even if the grid is messy or asymmetrical. The protection doesn't come from the shape of the buildings or the symmetry of the streets; it comes from the shape of the entire city itself.

The authors call these "Genus-Protected" phases. In simple terms, "genus" is just a fancy math word for the number of holes in an object.

A donut has one hole (genus = 1).
A pretzel might have three holes (genus = 3).
A smooth ball has zero holes (genus = 0).

The "Donut" Analogy
Imagine you have a rubber band (a loop of energy) running along the edge of a flat, square piece of paper. If you try to remove the rubber band, you can just slide it off the edge or pinch it until it disappears. It's easy to get rid of.

Now, imagine that same rubber band running along the edge of a donut.

If the rubber band goes around the hole of the donut, you cannot slide it off.
You cannot pinch it away without cutting through the rubber itself (which would mean breaking the fundamental rules of the system).
The only way to get rid of it is to tear a hole in the donut itself (which would mean destroying the "bulk" of the material).

The paper shows that by building crystals with holes (like donuts, cylinders, or toruses), you can trap these special energy states in a way that is impossible to remove unless you destroy the material's core.

How They Built It
The researchers didn't just theorize this; they built digital models of these "holey" crystals using two main tricks:

The "Corbino Disk" (The Donut): They took a standard crystal model and cut a square hole right out of the middle. This created two separate edges: an outer edge and an inner edge. Because the edges are disconnected, the special energy states on the inner edge cannot meet the ones on the outer edge to cancel each other out. They are stuck there, protected by the hole.
The "Volterra Construction" (The Twist): They simulated cutting the crystal lattice and re-gluing it with a twist (like a dislocation or a disclination). This creates a "knot" in the fabric of the crystal. Even if the crystal looks normal everywhere else, this knot forces the energy states to appear at the edges, and the global shape of the crystal prevents them from vanishing.

Why It Matters (According to the Paper)
The paper claims that these new phases are a unique mix of two existing types of topological phases:

Like Intrinsic phases, the energy states are robust and cannot be removed by surface tricks.
Like Extrinsic phases, they don't rely on the crystal having perfect symmetry (like rotation or mirror symmetry).

Instead, they rely entirely on the global topology (the number of holes). As long as the fundamental laws of physics (like time-reversal or particle-hole symmetry) are kept, and the "hole" in the material remains, these special states are permanent.

The Bottom Line
The paper proves that you don't need a perfectly symmetrical crystal to have these special, protected energy states. You just need to build your crystal with holes. By changing the "genus" (the number of holes) of the material, you can create a new class of topological matter where the special states are locked in place by the very shape of the object, making them incredibly stable against any surface-level disturbances.

The authors also suggest that these ideas could be tested in electrical circuits (using wires and capacitors to mimic atoms) and photonic systems (using light), where engineers can easily build "donut-shaped" or "pretzel-shaped" networks to see these effects in action.

Technical Summary: Genus-Protected Higher-Order Topological Phases

Problem Statement
Higher-order topological phases (HOTPs) are characterized by gapless boundary modes on boundaries of codimension $n > 1$ (e.g., corners in 2D, hinges in 3D). Current classification schemes divide HOTPs into two categories: intrinsic phases, which rely on specific lattice symmetries (e.g., rotation, mirror) for protection, and extrinsic phases, which arise from strong topological phases appearing on the boundary of a trivial bulk. A critical limitation in both classes is that if the protecting lattice symmetries are broken, the boundary modes can often be removed by surface perturbations without closing the bulk gap. The authors identify a gap in the existing framework: the existence of HOTPs protected not by lattice symmetries, but by the global topology (genus) of the system's shape, independent of crystalline symmetries.

Methodology
The authors propose construction schemes for "Genus-Protected Topological" (GPT) phases in both 2D and 3D systems. The core methodology involves engineering systems with non-zero genus (number of holes, $g$ ) and introducing specific defects or boundary conditions that prevent the annihilation of gapless modes via purely surface perturbations.

2D Construction:
- Local Boundary Defects: The authors modify the Benalcazar-Bernevig-Hughes (BBH) model on a square lattice with a central square hole (Corbino disk geometry, $g=1$ ). By introducing local edge defects (removing specific sites to alter the boundary termination), they create a scenario where Majorana zero modes (MZMs) are localized on disjoint inner and outer edges.
- Topological Defects: They utilize Volterra constructions to introduce dislocations (removing a line of sites) and disclinations (cutting and re-gluing the lattice with a rotation, e.g., $\pi$ ). These defects are placed such that the bulk remains locally uniform, but the global topology forces the existence of unpaired modes on the boundaries.
- Verification: Numerical simulations using the Kwant package compute energy spectra, real-space probability densities, and scattering-matrix invariants. Specifically, the determinant of the reflection matrix, $\det(r)$ , is used as a topological invariant ( $\nu = \text{sign} \det(r)$ ) to distinguish trivial from non-trivial phases.
3D Construction:
- The authors extend a 2D class-D Hamiltonian to 3D by stacking layers and adding $k_z$ modulation.
- They construct a toroidal geometry ( $g=1$ ) from a cubic lattice by introducing a $\pi/2$ disclination and rounding the top and bottom surfaces.
- This geometry supports helical Majorana modes propagating along non-contractible loops. The authors demonstrate that these modes cannot be annihilated by surface perturbations without traversing the bulk or breaking fundamental symmetries (Time-Reversal Symmetry or Particle-Hole Symmetry).

Key Contributions and Results

Definition of GPT Phases: The paper establishes a new class of HOTPs where gapless boundary modes are protected by the bulk gap, fundamental local symmetries, and the global genus of the system, explicitly independent of lattice symmetries.
Robustness Mechanism: In systems with $g > 0$ , gapless modes on disjoint boundaries (e.g., inner and outer edges of a Corbino disk) cannot be brought together and annihilated by any perturbation restricted to the surface. This contrasts with $g=0$ systems (e.g., a disk or sphere), where modes can be moved along the boundary to annihilate unless lattice symmetries prevent it.
Numerical Demonstration:
- In 2D, the authors show that a BBH model on a Corbino disk with specific edge defects hosts stable MZMs even when $C_4$ rotation symmetry is broken. The reflection matrix invariant $\det(r)$ switches from $+1$ (trivial) to $-1$ (GPT phase) in the presence of dislocations.
- In 3D, a toroidal geometry hosts helical modes on non-contractible loops. The scattering matrix invariant confirms the non-trivial topological nature of these phases.
Topological Classification:
- 2D Systems: For a lattice with $g$ holes, the classification depends on the symmetry class. For $\mathbb{Z}_2$ classes (e.g., Class D), the distinct phases are classified by $\mathbb{Z}_2^g$ . For $\mathbb{Z}$ classes, the classification is $\mathbb{Z}^g$ .
- 3D Systems: Using homology theory, the authors classify 1D modes on the boundary surfaces $\Sigma_i$ (each with genus $g_i$ ). For helical modes (associated with $\mathbb{Z}_2$ 2D classification), the distinct phases are classified by $\mathbb{Z}_2^{2\sum g_i}$ . For chiral modes (associated with $\mathbb{Z}$ ), the classification is $\mathbb{Z}^{2\sum g_i}$ . This is derived from the first homology group $H_1(\Sigma_i; G)$ , where the number of independent non-contractible loops on a surface of genus $g$ is $2g$ .

Significance and Claims
The authors claim that GPT phases represent a distinct category that bridges the intrinsic and extrinsic classifications of HOTPs. Like intrinsic phases, their boundary modes cannot be removed without closing the bulk gap (provided fundamental symmetries are preserved). Like extrinsic phases, they do not require lattice symmetries for protection. The protection arises solely from the global topology of the system shape.

The paper concludes that these phases are experimentally accessible. The authors specifically suggest topolectrical circuits as a natural platform, noting that the genus of the lattice in such circuits is determined by the connectivity of the circuit graph rather than physical geometry, making the engineering of non-trivial genus straightforward. They also mention photonic platforms and metamaterials (acoustic/mechanical lattices) as potential realization venues, citing existing work on higher-order topological phases and lattice defects in these systems. The work does not propose new experimental setups but argues that existing platforms capable of realizing HOTPs and lattice defects are sufficient to observe GPT phases.

More like this