{"id":"4990a717-8b8d-4410-8cac-a05fd4fbd2d7","arxiv_id":"2412.15333","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A symmetric scattering geometry with threaded flux yields reflection-matrix invariants that identify intrinsic higher order topological phases, verified on several models and with disorder.","lead":"Researchers developed a new way to measure the topology of higher order topological insulators, materials whose protected edge states are pinned to corners or hinges rather than whole surfaces. They use a ring-shaped sample with a magnetic flux through the hole and show that the way particles reflect off the inner and outer surfaces reveals the material's topological phase, even when disorder is present.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The annulus flux-winding is verified only as a transition detector for examples; the paper does not prove it equals the bulk HOTI index, and Sec. 4.1 admits the 1D invariant is constructed ad hoc.","rationale":"The reader's conditional verdict is well matched to the evidence. The paper gives correct-looking numerical demonstrations for the BBH model, a C4T+P network model, a three-dimensional C4T-symmetric HOTI, and an inversion-symmetric axion insulator, and the code is deposited on Zenodo. Those examples support the claim that the flux-dependent reflection matrix tracks the phase transitions in those specific systems. What they do not support, without further argument, is the abstract's stronger statement that the flux dependence of the reflection matrix 'captures' intrinsic higher-order topology, or the conclusion that 'all second order HOTIs have protected defect modes appearing at flux lines.' The load-bearing gap is the equality between the scattering-derived invariant and the established bulk topological index. The paper itself flags this gap in Sec. 4.1 when it states that the topology of 1D systems with fractional translations and sublattice symmetry had not been studied before and that the invariant is constructed rather than derived. The conclusion narrows the universality claim to 'examples, together with general arguments' and explicitly excludes third-order HOTIs. A decisive check is to compare Q with a real-space bulk invariant on the same parameter sets and disorder realizations. If the two agree across a broader family, the central claim is substantially strengthened; if they disagree, the scattering invariant would be a transition detector rather than a bulk invariant. The secondary gauge-independence check is important because App. D introduces a smooth-gauge construction that could, in principle, change the value of the winding; confirming invariance under independent gauge implementations would protect the invariant's definition. Neither check invalidates the paper's current contribution; they determine whether the universal bulk-edge correspondence claim can be accepted as stated.","tokens_in":21463,"tokens_out":6977,"duration_ms":52639,"concrete_test":"Independently benchmark Q from Eq. (14), and the analogous expressions in Secs. 4.2 and 4.3, against a real-space bulk invariant for a family of second-order HOTIs beyond the published examples: for instance, a C4-symmetric chiral model with anisotropic hoppings or an added C4-symmetric perturbation, scanning a two-dimensional parameter region that crosses the bulk transition line. Compute the nested Wilson loop or Bott index on the same parameters and disorder realizations, and compare with Q. If Q disagrees with the bulk index anywhere away from finite-size-shifted transition points, the central equality claim fails. As a secondary control, recompute the BBH invariant with two independent smooth-gauge implementations of App. D to confirm that Q is gauge-independent.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Load-bearing concern: the central claim is not merely that an annulus geometry yields a flux-dependent reflection matrix with a stable winding, but that this winding is the bulk higher-order topological index. The paper demonstrates the former empirically for four models, and the latter only by observing that Q changes at the phase transition of each example. In Sec. 4.1, after obtaining Heff from r, the authors explicitly write: 'to the best of our knowledge, the topology of one-dimensional systems combining fractional translations and sublattice symmetries has not been studied before. Therefore, we proceed to construct the invariant.' Equation (14) is therefore an unproven candidate invariant. The conclusion itself limits the generality: 'Our examples, together with general arguments, indicate that our approach applies to all second order HOTIs,' and it concedes that third-order HOTIs 'do not seem to fit into our framework.' Nothing in the paper rules out a second-order HOTI whose flux-winding Q stays constant across a bulk phase transition, or a trivial model whose Q changes without a bulk transition; the only evidence against this is numerical agreement at a single transition line for the BBH model and the phase diagrams for the other examples. The same gap affects the concluding inference that 'all second order HOTIs have protected defect modes appearing at flux lines,' which is drawn from geometry-independence of Q rather than from an equality proof. This is an external-validity gap rather than an internal inconsistency: the examples appear correctly computed, but they do not establish the universal bulk-edge correspondence claimed in the abstract.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a scattering-theoretic framework for probing intrinsic higher-order topological insulators and superconductors (HOTIs) with spatial symmetries. The central construction is an annulus or cylindrical scattering geometry: leads are attached to the inner and outer boundaries, a magnetic flux φ is threaded through the center, and the reflection matrix r(φ,k) is computed at zero energy. Symmetry constraints on r are derived from the spatial symmetries of the HOTI, and effective Hamiltonians Heff(φ,k) are built from r. Topological invariants are then assigned to Heff, taking the form of winding numbers or Pfaffian-based Z2 expressions (Eqs. 14, 16, 19, and 22). The approach is applied to four model systems: the Benalcázar–Bernevig–Hughes (BBH) model, a two-dimensional C4T-symmetric network model, a three-dimensional C4T-symmetric HOTI, and a three-dimensional inversion-symmetric axion insulator. In each case the invariant is numerically found to change at the independently known bulk phase transition, and for the BBH model the invariant remains well-defined in the presence of globally C4-symmetric disorder.","tokens_in":21781,"tokens_out":5940,"duration_ms":56265,"significance":"If the central claim holds, this is a valuable contribution: it would provide a real-space, disorder-compatible probe of intrinsic HOTI topology, connect higher-order topology to flux response, and offer a new route toward bulk–edge correspondence in these phases. The paper has clear strengths: the symmetry-constraint derivation in Sec. 3 and App. B is systematic; the numerical verification covers four distinct models with different symmetry groups; the code is deposited on Zenodo (Ref. [54]); and the invariant is shown to work for a network model, where a bulk Hamiltonian is not available. However, the paper does not prove that the flux-winding of the reflection matrix equals the bulk higher-order topological index. The authors are candid about this in Secs. 4.1 and 5, but the abstract and conclusion make stronger claims than the evidence supports. The universality for all second-order HOTIs is asserted on the basis of examples and general arguments, not a proof, and the spectral-flow identification is justified by a single numerical example (Fig. 4).","major_comments":[{"comment":"The central claim that the flux-winding of the reflection matrix equals the bulk higher-order topological index is not proven. In Sec. 4.1 the authors state that the topology of the effective one-dimensional system has 'not been studied before' and that they 'proceed to construct the invariant,' making Eq. (14) a candidate invariant rather than a derived one. The evidence that Q changes only at the bulk transition is numerical agreement at the known transition point (λ=γ for the BBH model) and at the corresponding transitions in the other examples. The conclusion (Sec. 5) itself says the examples 'indicate' the approach applies to all second-order HOTIs. As written, the abstract's 'We demonstrate' and the final sentence that 'all second order HOTIs have protected defect modes appearing at flux lines' exceed what is established. The authors should either supply a proof for at least one nontrivial case (e.g., relating Eq. (14) to the nested Wilson loop or corner charge of the BBH model) or reformulate the universality claim explicitly as a conjecture supported by numerical evidence.","section":"Sec. 4.1, Eq. (14); Sec. 5"},{"comment":"The correspondence between the threaded flux φ and a momentum parameter of Heff is an assumption rather than a derived statement. Section 3 asserts that C_n^n(φ)=C_n^n(0)e^{iφ} and that this 'invites interpreting φ as a momentum of Heff,' but no argument rules out a second-order HOTI whose invariant Q stays constant across a bulk transition, or a trivial model whose Q changes in the absence of a bulk transition. The only evidence for the spectral-flow identification is the numerical example in Fig. 4 for the BBH model in the topological phase. This correspondence is load-bearing: it is the step that connects the scattering setup to the bulk index. A precise statement of the conditions under which the flux acts as a faithful momentum parameter, with at least a heuristic derivation or additional nontrivial examples showing that the correspondence fails when those conditions are violated, is needed.","section":"Sec. 3; Fig. 4"},{"comment":"The concluding inference that all second-order HOTIs have protected defect modes at flux lines overreaches the presented evidence. The argument uses geometry-independence of Q, but geometry-independence alone does not imply that the flux-line defect modes are protected by the bulk topology unless Q has been shown to equal the bulk index. The paper also concedes that third-order HOTIs 'do not seem to fit into our framework,' yet the abstract and conclusion do not adequately convey this limitation. The defect-mode conclusion should be presented as a consequence of the (conjectured) equality between the scattering invariant and the bulk invariant, rather than as a separate established result.","section":"Sec. 5; Defect-mode claim"},{"comment":"Figure 8(d) shows an extended range of µ over which the sign of the invariant changes, which the text attributes to the Fermi surface. This needs clarification: if the bulk is gapless at zero energy in that range, the reflection matrix is not unitary and the invariant defined in Eq. (22) may not be a well-defined topological index. The authors should state explicitly whether Q remains quantized in the gray region, whether the reflection gap ∆ in Fig. 8(c) closes there, and how a 'change of sign' is defined when the invariant may not be quantized. Without this clarification, the axion-insulator example does not unambiguously demonstrate that the scattering invariant detects the phase transition.","section":"Sec. 4.3, Fig. 8; Eq. (22)"},{"comment":"The numerical smooth-gauge construction in App. D does not establish that the invariants are independent of the gauge choices made in the eigenvector and SVD updates. The integrals in Eqs. (14), (19), and (22) involve log det of the reflection matrix, and a non-smooth or branch-choosing gauge could in principle change the parity or the winding. The authors should either prove gauge invariance analytically or demonstrate numerically that different gauge seeds and different step sizes δφ give the same value of Q.","section":"App. D; Eqs. (14), (19), (22)"}],"minor_comments":[{"comment":"In the paragraph after Fig. 1, 'surface gap glosing' should be 'surface gap closing.'","section":"Sec. 1"},{"comment":"The symbol K for complex conjugation is used without definition; please define it or avoid it by writing the conjugation explicitly.","section":"App. B.2"},{"comment":"Equation (22) is difficult to parse because of the nested fractions and exponentials; please present the invariant with an explicit definition of the branches of the logarithms and square roots used in the determinants.","section":"Eq. (22)"},{"comment":"The phrase 'the crossing at ϕ=π is fine-tuned' is unclear; please specify which parameter is tuned, to what value, and why the crossing is not generic.","section":"Fig. 4 caption"},{"comment":"The phrase 'higher order HOTIs' is redundant; consider using 'third-order HOTIs' or 'HOTIs of order three.'","section":"Sec. 5"},{"comment":"The network-models package is cited as 'to be published'; if a stable version or DOI is available, please cite it instead.","section":"Ref. [40]"}],"recommendation":"major_revision","confidential_remarks":"The main gap is between the paper's strong claims (abstract, conclusion) and the evidence, which is numerical and acknowledged in Secs. 4.1 and 5. I believe this is fixable within the manuscript's scope: the authors could add a proof of the invariant for at least the BBH model, or explicitly soften the universality and defect-mode claims to conjectures. The code availability and the breadth of examples are clear strengths, and the self-citation of Refs. [20] and [50] is appropriate given their direct relevance. The axion-insulator Fermi-surface issue in Sec. 4.3 should also be clarified before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new thing here is real: an intrinsic scattering invariant for second-order HOTIs, built from an annulus geometry with flux and a reflection matrix whose symmetry constraints are derived carefully. The mapping of rotation to translation and inversion to glide in the effective Hamiltonian is a genuine conceptual step, and the numerical work is solid across four distinct models, including a disordered BBH model, a network model, a 3D C4T-symmetric insulator, and an axion insulator. The code is on Zenodo, so the results are reproducible. This is a useful toolkit, especially for disordered or amorphous HOTIs where real-space invariants are awkward.\n\nWhat I'd push back on is the headline claim. The abstract says the flux dependence of the reflection matrix 'captures' intrinsic higher-order topology and the conclusion promises an 'alternative approach for proving bulk–edge correspondence.' But the paper does not prove that the flux-winding invariant equals the bulk HOTI index. It verifies that the invariant changes at known phase transitions in each example. That is evidence, not a theorem. The authors are candid about this in places: Sec. 4.1 admits the one-dimensional invariant is 'constructed' because the topology of such systems was not studied before, and the conclusion says the examples 'indicate' generality, with third-order HOTIs left as a conjecture. I would have liked the abstract to say 'we construct scattering invariants that correctly classify these examples' rather than implying a general proof.\n\nThe other soft spot is the annulus-geometry assumption. The invariant relies on having protected modes on both inner and outer boundaries, with flux acting as a faithful momentum parameter. The paper offers general arguments but no rigorous condition for when this is possible. That said, this is an external-validity gap, not an internal error: the symmetry derivations and numerics check out, and the reflection-gap closing at transitions is strong supporting evidence for the examples treated.\n\nMinor point: a side-by-side comparison with real-space invariants (e.g., Bott index) on the same disordered samples would have made the practical value more concrete.\n\nOverall, this is honest, careful work by people who know the subject. It deserves a serious referee, but the authors should either supply a sharper theorem or scale back the universal claims in the abstract. I would send it to peer review with that expectation.","headline":"Solid new scattering construction for intrinsic HOTIs, honestly verified on examples, but the abstract's universality claim outruns a proof and should be tempered.","tokens_in":22258,"tokens_out":1840,"would_cite":true,"duration_ms":13142,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A symmetric annular scattering geometry with a threaded flux turns the flux dependence of the reflection matrix into a universal invariant for intrinsic higher-order topological phases.","keywords":["higher-order topological insulators","scattering theory","reflection matrix","spectral flow","flux line","bulk-edge correspondence","topological invariant","disorder"],"falsifier":"Compute the scattering invariant of an intrinsic second-order HOTI in an annulus and check whether $\\det h(\\phi)$ (or the Pfaffian) winds while the reflection matrix has no zero eigenvalue in the topological phase, or whether it changes when only the surface gap closes; a single such example, or a demonstration that the flux-phase convention $V^n_{C_n} = C^n(0) e^{i\\phi}$ depends on gauge, would falsify the claim that the reflection matrix captures the bulk index.","tokens_in":21263,"feed_emoji":"🌀","tokens_out":5890,"duration_ms":46987,"temperature":0.7,"pith_summary":"The paper develops a scattering theory for intrinsic higher-order topological insulators and superconductors—phases whose boundary states are protected by the spatial symmetries of a finite sample. The authors show that standard scattering invariants fail for these phases because the geometry either breaks the protecting symmetry or only probes surface physics. Their fix is an annulus or cylinder geometry with leads on both the inner and outer boundary and a magnetic flux through the hole. The flux dependence of the reflection matrix then encodes the bulk higher-order index, changing only when the bulk gap closes. This gives a transport-based route to bulk–boundary correspondence that also works with disorder and in network models that lack a bulk Hamiltonian.","feed_headline":"Flux through a hole reveals higher-order topology","feed_subtitle":"A symmetric annulus turns flux-dependent reflection into a bulk invariant that survives disorder.","key_machinery":"The load-bearing object is the reflection matrix $r(\\phi, k_{d-2})$ of an annulus in two dimensions or a cylinder in three dimensions, with leads attached symmetrically to both the inner and outer boundaries and a flux $\\phi$ threaded through the hole. The flux acts as an extra momentum parameter, completing a dimensional reduction that maps the $d$-dimensional bulk onto a $(d-1)$-dimensional effective Hamiltonian $H_{\\mathrm{eff}}(\\phi, k_{d-2})$ constructed from $r$ according to a symmetry-dependent prescription (Hermitian when the bulk has chiral symmetry, off-diagonal otherwise). The crucial structural insight is that spatial symmetries of the bulk become nonsymmorphic symmetries of $H_{\\mathrm{eff}}$: a rotation becomes a translation combined with a phase, and inversion becomes a glide. The invariants are then evaluated on $H_{\\mathrm{eff}}$—for the BBH model as the parity of crossings of $\\det h(\\phi)$ through the real axis, and for the magnetic and inversion-symmetric examples as Pfaffians at high-symmetry points—so that a zero eigenvalue of $r$ is exactly a bulk-gap-closing transition.","core_discovery":"The central claim is that the flux dependence of the reflection matrix of a symmetric annular scattering geometry is a universal probe of intrinsic higher-order topology. Concretely, the paper demonstrates that the winding of $\\det h(\\phi)$, or equivalently Pfaffian-type expressions built from the reflection matrix, equals the bulk higher-order topological index in a series of examples: the two-dimensional BBH model, a two-dimensional $C_4\\mathcal{T}$-symmetric network model, a three-dimensional $C_4\\mathcal{T}$-symmetric higher-order topological insulator, and a three-dimensional inversion-symmetric axion insulator. The invariant is defined through a dimensionally reduced effective Hamiltonian $H_{\\mathrm{eff}}(\\phi, k_{d-2})$ that is gapped whenever the reflection matrix has no zero eigenvalue, so it can only change when the bulk gap closes. Accompanying this, the paper identifies the spectral flow of zero modes as a function of the flux threaded through the annulus as a signature of higher-order topology, present only in the topological phase. The scattering approach thereby provides an alternative proof route for bulk–edge correspondence in intrinsic higher-order topological phases, including disordered ones.","pith_inferences":["(Extension) The conjecture that third-order HOTIs require two holes and two fluxes suggests a natural generalization: a $d$-dimensional $N$-th order phase should be probed by an $N$-fold annular geometry with $N$ independent fluxes, mapping to a $d-1$-dimensional effective Hamiltonian.","(Extension) Because the invariant depends only on the reflection matrix, it could be measured in metamaterial realizations where reflection-phase measurements are routine, turning the flux-winding into an observable.","(Extension) The equivalence between flux response and higher-order topology implies that flux lines act as local topological defects carrying the bulk index, which may yield quantized transport or pumping signatures in driven versions of these systems.","(Extension) The dimensional-reduction map from rotation to translation suggests a unified classification where point-group-protected higher-order invariants are indexed by glide-symmetric one-dimensional invariants, possibly extending to magnetic space groups."],"forward_implications":["The intrinsic higher-order index can be computed from reflection data alone, without diagonalizing the full spectrum.","The invariant is robust to disorder that preserves the global symmetry, so it can certify higher-order topological phases in disordered or amorphous samples.","The approach resolves the ambiguous topological classification of network models, where no bulk Hamiltonian exists.","The flux-response (spectral flow) signature connects higher-order topology to Laughlin-type flux arguments, giving a local, experimentally relevant probe.","The method provides an alternative proof of bulk–edge correspondence for second-order topological insulators and superconductors."],"supporting_citations":[{"why":"Supplies the dimensional-reduction scattering formalism for strong topological insulators that the paper extends to spatial symmetries.","marker":"[20]"},{"why":"Defines the BBH model, the primary testbed for the intrinsic scattering invariant.","marker":"[25]"},{"why":"Provides the two-dimensional $C_4\\mathcal{T}$-symmetric network model used to demonstrate the scattering invariant in a system without a bulk Hamiltonian.","marker":"[39]"},{"why":"Gives the Pfaffian formalism and the three-dimensional $C_4\\mathcal{T}$-symmetric model used for the hinge-mode example.","marker":"[41]"},{"why":"Supplies the topological invariants for crystalline phases with glide symmetries, applied to the axion insulator calculation.","marker":"[49]"},{"why":"Establishes the inversion-protected higher-order topological insulator and its chiral surface mode, the basis of the axion insulator example.","marker":"[48]"},{"why":"Studied the delocalization transition of a disordered axion insulator, which the scattering proof in this paper addresses.","marker":"[6]"},{"why":"Reported the flux cancellation of finite-size splitting in annular geometries, used to justify the necessity of the threaded flux.","marker":"[32]"},{"why":"Provided the scattering formula for the topological quantum number, the basis for reading the index from the reflection matrix.","marker":"[19]"}],"fun_headline_variants":["Reflection matrix flux winding exposes hidden higher-order topology","Symmetric annulus makes flux-dependent reflection a bulk invariant","Flux through a hole: scattering reveals higher-order topology","Flux-dependent reflection matrix: a robust probe of higher-order topology"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that for every intrinsic second-order higher-order topological phase one can choose a symmetric annulus or cylinder geometry in which the two leads are separated by the bulk and the threaded flux acts as a faithful momentum parameter, with a gauge-independent flux-phase convention for the symmetry operators; if a phase cannot be probed this way, the reflection-matrix invariant would not see the bulk topology.","fun_headline_variants_meta":{"raw":{"variants":["Reflection matrix flux winding exposes hidden higher-order topology","Symmetric annulus makes flux-dependent reflection a bulk invariant","Flux through a hole: scattering reveals higher-order topology","Flux-dependent reflection matrix: a robust probe of higher-order topology"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000671,"raw_usage":{"total_tokens":3032,"prompt_tokens":897,"completion_tokens":2135,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":513,"completion_tokens_details":{"reasoning_tokens":2067}},"tokens_in":513,"tokens_out":2135,"duration_ms":12695,"temperature":1.0,"reasoning_tokens":2067,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:31:18.836994+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the scattering invariant of an intrinsic second-order HOTI in an annulus and check whether $\\det h(\\phi)$ (or the Pfaffian) winds while the reflection matrix has no zero eigenvalue in the topological phase, or whether it changes when only the surface gap closes; a single such example, or a demonstration that the flux-phase convention $V^n_{C_n} = C^n(0) e^{i\\phi}$ depends on gauge, would falsify the claim that the reflection matrix captures the bulk index.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the two-dimensional $C_4\\mathcal{T}$-symmetric network model used to demonstrate the scattering invariant in a system without a bulk Hamiltonian."},{"cited_title":"Shiozaki, M","cited_arxiv_id":null,"evidence_quote":"Supplies the topological invariants for crystalline phases with glide symmetries, applied to the axion insulator calculation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reported the flux cancellation of finite-size splitting in annular geometries, used to justify the necessity of the threaded flux."}],"review_version":1}