Topological materials
Can topological corner states survive inside the bulk energy band?
Open access · cc by · source: Europe PMC
Light injected at a corner of a topological photonic lattice stays trapped there even though the corner states share their energy with bulk states, until the bandgap is made too small.
Study at a glance
- Design
- Other — Laser-written 2D waveguide lattices (C4-symmetric 2D SSH model) probed with heralded single photons injected at corners, at several propagation lengths
- N
- No participant N; physical samples are lattices of 8 x 8 waveguides with propagation lengths from 10 to 30 mm, in topological, trivial and near-transition parameter sets
- Population
- Femtosecond-laser-written photonic waveguide lattices in glass
- Outcome
- Photon intensity distribution at the output and a localization index measuring how much light stays at the injected corner
Structured fields used in claim comparison tables when every cited study has a complete layer.
Key findings
In the topological lattice the light stayed on the injected corner and the localization index remained close to one at every propagation length, whereas in the trivial lattice light spread through the whole array and the index fell towards zero. Injecting a prepared superposition excited just one zero-energy corner eigenstate, which kept its shape as it propagated. When the coupling ratio was raised to 0.68, shrinking the bandgap, light leaked from the corner to the other corners and edges, showing the bound states break down.
Methodology
The team wrote two-dimensional arrays of optical waveguides into glass so that alternating weak and strong couplings mimic a 2D Su-Schrieffer-Heeger model with fourfold rotation symmetry. They sent single photons into one corner, or a four-way equal superposition into all four corners, and imaged where the light ended up after different propagation lengths. They compared a topological lattice, a trivial lattice with the couplings swapped, and a lattice close to the phase transition.
Limitations
The lattices are small (8 by 8 sites), so finite-size effects near the transition are significant and the results are demonstrations rather than scaling studies. Only a few coupling ratios and propagation lengths were tested, so the exact point where protection fails is not mapped. The single-photon source behaves like classical light in a linear lattice, so the 'quantum superposition' framing does not show any effect that needs entanglement. Much of the theoretical argument (orthogonality to bulk states, the topological index) is in supplementary material not included here.
How this study connects
Role on claims
Each row is a claim on a concept or method page where this paper supports, challenges, or qualifies the statement. Roles are hand-checked — not a model guess.
Corner states are protected only while the bulk gap stays open.
In a laser-written 2D SSH waveguide lattice, light injected at a corner of the topological lattice stayed localised at every propagation length tested, while in the trivial lattice it spread; pushing the coupling ratio to 0.68 shrank the gap and let light leak to other corners and edges.
Evidence for the claim as stated.
Measured versus predicted: the electronic and mechanical results (Bi2Se3 films, Mn(Bi,Sb)2Te4 Hall bars, truss lattices, ferrite rods, waveguides) come from fabricated samples, while the Weyl metamaterial, MnBi2Te4 optics and MoS2/CrBr3 flat Chern bands are theory or simulation with idealised, disorder-free structures.
Evidence for the claim as stated.
How robust is 'protected'? The amorphous photonic lattice and the SSH waveguide lattice both show protection failing once disorder or coupling closes the gap, which limits claims of unconditional robustness made in more idealised models.
Evidence for the claim as stated.
Open questions
Tensions this paper is part of
From concept pages' “where studies disagree.” Disagreement means the same question; scope means different assays, populations, or outcomes.
Measured versus predicted: the electronic and mechanical results (Bi2Se3 films, Mn(Bi,Sb)2Te4 Hall bars, truss lattices, ferrite rods, waveguides) come from fabricated samples, while the Weyl metamaterial, MnBi2Te4 optics and MoS2/CrBr3 flat Chern bands are theory or simulation with idealised, disorder-free structures.
- Supports · Do topological edge states need a crystal lattice?
- Supports · Can swapping bismuth for antimony flip a magnet's Hall signals?
- Supports · Can a passive, static structure show one-way topological effects?
- Challenges · Can light bend the 'wrong' way at every angle with no reflection?
- Challenges · Can light absorption reveal the hidden geometry of electron waves?
- Challenges · Can magnetic skyrmions make flat topological bands?
How robust is 'protected'? The amorphous photonic lattice and the SSH waveguide lattice both show protection failing once disorder or coupling closes the gap, which limits claims of unconditional robustness made in more idealised models.
Related papers in this topic
Same topic cluster — not a recommendation engine.
- Can light's frequency act as an extra dimension for topology?
- Do topological edge states need a crystal lattice?
- Can topological insulators triple terahertz frequencies efficiently?
- How do hot electrons change a Weyl semimetal's direction-dependence?
- Can a passive, static structure show one-way topological effects?