{"id":"e50796c9-00c7-483c-a81e-30a1f0a8bc92","arxiv_id":"2504.21534","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Spin waves in monolayer and bilayer honeycomb ferromagnets are classified topologically, with edge states described by effective Su-Schrieffer-Heeger chains and nonzero Chern numbers induced by Dzyaloshinskii-Moriya interactions.","lead":"This paper analyzes the topological behavior of spin waves (magnons) in honeycomb ferromagnets and shows it can be captured by simple one-dimensional chain models. It maps which magnetic stackings and interlayer couplings produce topologically protected edge states, which could help design magnonic devices.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"AFM-interlayer bilayer spectra contain non-zero imaginary energies (admitted in text); without a positive-definiteness check, the quoted Chern numbers and |E| edge-state plots for these cases lack physical grounding.","rationale":"The reader identified the AFM-interlayer complex-energy issue as the weakest assumption, and the manuscript itself confirms non-zero imaginary energies. I agree that this is the single most load-bearing concern. It directly targets the central claim for the AFM-interlayer bilayer cases, where the paper computes non-Hermitian Chern numbers and plots |E| edge states without establishing that the underlying bosonic ground state is stable. A positive-definiteness check of HBdG is the natural, decisive test. The monolayer and FM-interlayer cases appear more secure, and the paper's effective SSH-chain explanation for D=0 edge states is a plausible extension of known results, so the reader's CONDITIONAL verdict remains appropriate rather than a full rejection. No code or data is provided, which makes the proposed numerical test particularly important.","tokens_in":24980,"tokens_out":9985,"duration_ms":107795,"concrete_test":"Recompute the eigenvalues of the full Hermitian BdG matrix HBdG(ky) (as defined in Eqs. 50-51 and 60-62) for AA- and AB-stacked AFM-interlayer bilayers at the parameters used in Figs. 8-10 (S=1, J=1 eV, Jin=0.1J, γ=3J, D=0 and 0.1), on both a periodic bulk and a finite strip with the bearded/zigzag open boundaries. Accept the classification only if all eigenvalues of HBdG are non-negative for every ky and k, and if the eigenvalues of σz HBdG have exactly zero imaginary part in the bulk. Report the minimum eigenvalue and the largest imaginary part; if either is negative/non-zero, the AFM-interlayer entries in Table I and the |E| edge-state plots are unphysical.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing weakness is in the AFM-interlayer bilayer sections (IV.A.2 and IV.B.2). The paper derives non-Hermitian effective Hamiltonians ηh(ky) (Eqs. 47, 51, 62) and computes Chern numbers for them (Eqs. 49, 63; Table I), while in Fig. 10 it plots |E| because, in the authors' own words, 'imaginary energies appear (which is pretty small compared to the real part, but non-zero, so we show |E| in Fig. 10 for simplicity)'. For a bosonic BdG system the physical magnon energies are the eigenvalues of σz HBdG. Non-zero imaginary parts mean that the Hermitian matrix HBdG is not positive definite, i.e. the assumed collinear ferromagnetic ground state is dynamically unstable; negative eigenvalues of HBdG would imply that the vacuum is not the ground state. The paper never checks positive definiteness or ground-state stability for the AFM-interlayer cases. If the spectrum is complex, the biorthogonal Chern numbers quoted in Table I are non-Hermitian invariants that need not correspond to any physical bulk-edge correspondence, and the 'straight connection between edge states and non-zero Chern number' asserted in the abstract is unsupported for these systems.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies magnon excitations in monolayer and bilayer honeycomb ferromagnets using linear spin-wave theory and a single-particle bosonic Bogoliubov-de Gennes (BdG) description. It derives the bulk BdG Hamiltonians for monolayer systems with bearded, zigzag, and armchair edge geometries, and for AA- and AB-stacked bilayers with ferromagnetic (FM) or antiferromagnetic (AFM) interlayer coupling, with or without next-nearest-neighbor Dzyaloshinskii-Moriya interaction (DMI). The effective one-dimensional edge Hamiltonians H(ky) are mapped to Su-Schrieffer-Heeger (SSH) chains or SSH ladders, and edge states are analyzed as functions of ky. The main claims are: (i) for DMI-free systems the conventional bulk-edge correspondence based on the 2D Chern number is only partially valid, with edge states still appearing because of the SSH-like topology of H(ky); (ii) when DMI is present the nonzero Chern number is connected to the existence of edge states; and (iii) for AFM-interlayer bilayers the relevant effective Hamiltonian is non-Hermitian, yet a non-Hermitian Chern number can still be defined. The paper summarizes its results in Table I and illustrates the spectra in Figs. 3-10.","tokens_in":25146,"tokens_out":21093,"duration_ms":218149,"significance":"If the results hold, the paper provides a systematic catalog of magnon edge-state behavior and topological invariants for a widely studied family of honeycomb magnetic systems, and it highlights an instructive mapping from bosonic BdG problems to non-Hermitian SSH chains. The explicit analytic derivation of h(k) for the monolayer and FM-interlayer bilayer cases is mostly standard, the internal relations among the Chern numbers in Table I are consistent, and the SSH-chain interpretation gives a useful physical picture. The treatment of the AFM-interlayer cases is the most interesting part but also the least secure: the paper itself admits the presence of imaginary energies, yet it does not establish that the corresponding bosonic ground state is stable or that the quoted non-Hermitian Chern numbers have a well-defined physical meaning.","major_comments":[{"comment":"The AFM-interlayer bilayer cases are central to the paper's claim of a 'straight connection' between edge states and nonzero Chern numbers, but their physical validity is not established. The text explicitly states that 'imaginary energies appear' for the effective Hamiltonian ηh(ky) and that |E| is plotted in Fig. 10 for this reason. In a bosonic BdG problem, the physical magnon frequencies are the eigenvalues of Pz HBdG, and non-real eigenvalues are a standard signature that the Hermitian matrix HBdG is not positive definite, meaning the assumed collinear magnetic state is dynamically unstable. The paper never checks positive definiteness, never identifies a parameter regime in which the spectrum of ηh(ky) is real, and does not discuss whether the biorthogonal Chern numbers C1=±1 listed in Table I for these rows remain well defined when complex eigenvalues are present. Without such a check, the Chern numbers and the edge-state spectra in Fig. 10 for the AFM-interlayer bilayers cannot be interpreted as physical magnon properties, and the corresponding claims in the abstract are unsupported.","section":"Sec. IV.A.2 and IV.B.2, Eqs. (47)-(51) and (61)-(63), Fig. 10"},{"comment":"The phrase 'conventional bulk-edge correspondence is partially valid' is never made precise. For D=0 the 2D Chern number is zero (Eq. (31)), yet edge states exist; the paper attributes these edge states to the SSH-like 1D winding of H(ky), which is a different invariant. For D≠0, the paper claims a 'straight connection' to the nonzero Chern number, but Table I lists NES=4 for Chern number ±1, while the text in Sec. III.B indicates that only one in-gap chiral branch is the Chern-related feature, the other three being boundary-condition-dependent nonchiral states. The manuscript should state explicitly which invariant (Chern number of the 2D bulk, 1D winding of H(ky), or neither) protects each class of edge states and how the entry NES in Table I is computed.","section":"Abstract and Sec. III.B, Table I, Eq. (31)"}],"minor_comments":[{"comment":"The word 'pesudo-Hermiticity' should be 'pseudo-Hermiticity' in Eq. (22) and the surrounding text.","section":"Sec. II.D"},{"comment":"The phrase 'two boardly suitable conclusions' should read 'two broadly suitable conclusions'.","section":"Sec. II.D, around Eq. (24)"},{"comment":"The statement that 'edge states can appear as the ground state' is confusing; the edge states shown in Figs. 3-6 are lowest-energy eigenstates of the single-magnon Hamiltonian, not many-body ground states of the spin system. Please rephrase to avoid the implication of a Bose-condensed or zero-energy ground state.","section":"Abstract and Sec. III.B"},{"comment":"The quantities A, B±, λ1±, and λ2± are used in the expressions for the eigenstates and eigenvalues before they are defined. Please move the definitions ahead of their first use.","section":"Sec. IV.A.2, Eq. (48)"},{"comment":"For the rows involving ηh(k) with AFM interlayer coupling, the table and the text should state explicitly whether the quoted Chern number is computed in a regime where the eigenvalues of ηh(k) are real; otherwise the classification is incomplete.","section":"Table I and Sec. IV.A.2"},{"comment":"The caption says the figure shows |E| for the effective chain, but it does not clarify whether the imaginary energies appear in the bulk spectrum, the edge-state branches, or both. This point should be stated in the main text and the caption.","section":"Fig. 10 caption and Sec. IV.B.2"},{"comment":"The self-referential footnote about the difference between ky and k should be incorporated into the main text or removed, as it reads like an editorial note rather than a regular reference.","section":"Footnote [54]"},{"comment":"The phrase 'straight connection' is likely meant as 'direct connection' and would be less overclaiming if reworded, especially given the additional nonchiral edge states discussed in the paper.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":"The main technical issue is the AFM-interlayer sections (IV.A.2 and IV.B.2), where the paper admits imaginary energies without establishing the stability of the bosonic BdG problem. If the authors cannot show that there is a parameter regime in which the spectrum is real and the Chern numbers are physically meaningful, those rows of Table I and the corresponding edge-state claims should be removed or substantially qualified. The monolayer and FM-interlayer bilayer results appear sound and are likely publishable after the bulk-edge-correspondence statements are made precise. The novelty relative to prior work on SSH-chain mappings (including Ref. [53]) is incremental, but the systematic catalog across stackings and boundary conditions may still be a useful contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a useful, mostly standard application of the bosonic BdG / non-Hermitian classification framework to magnons in honeycomb ferromagnets. The genuinely new content is the systematic bilayer classification and the boundary-dependent distribution of edge states, not the framework itself. The soft joint is the AFM-interlayer bilayer section: complex energies appear, the paper plots |E|, and physical stability is never addressed. That needs to be fixed before the claims there can be trusted.\n\nWhat the paper does well: the derivations are explicit and checkable. The SSH-chain mappings are derived from the BdG Hamiltonians rather than assumed. Table I is internally consistent, and the monolayer DMI limit reproduces the known Chern number ±1. The FM-interlayer AA/AB results, including the AB-stacking C = ±2 and the different edge-state distributions for different boundary combinations, are new and worth having. The citation pattern is fair; the paper builds on the relevant non-Hermitian classification, the symmetry-class literature, and the prior graphene SSH mapping [53]. The paper does not hide its soft spots: it explicitly says imaginary energies appear in the AFM cases and that it shows |E| for simplicity. That candor is good, but it does not resolve the problem.\n\nThe main weakness is the AFM-interlayer bilayer cases. For a bosonic BdG magnon Hamiltonian, the physical spectrum is determined by sigma_z H_BdG, and non-zero imaginary eigenvalues indicate that the assumed collinear ferromagnetic state is not a stable ground state, or at least that the linear spin-wave expansion is on shaky ground. The paper never checks positive definiteness of the BdG matrix for these cases. The non-Hermitian Chern numbers computed for eta h(k) may be mathematically well-defined, but their connection to real magnon edge states is not established. This affects the abstract's claim of a direct connection between edge states and nonzero Chern number for these systems. I would ask the authors to identify parameter regimes where the spectrum is real and positive, and to restrict their physical claims to those regimes.\n\nA lesser issue: the phrase \"partially valid\" bulk-edge correspondence is never made precise. For D = 0 the Chern number is zero yet edge states exist; as the paper shows, those come from the 1D SSH winding of H(k_y). That is fine, but the paper should say explicitly that the relevant invariant is the 1D winding, not the 2D Chern number.\n\nThere are also many typos and no code or data. Given the analytic character, that is a cleanup issue rather than a fatal one.\n\nOverall: I agree with the conditional verdict. The paper deserves a serious referee, because the bilayer classification is a real extension and not a rehash. But the referee should focus on the AFM-interlayer stability question and on whether the non-Hermitian invariants have a physical interpretation there.","headline":"Useful bilayer extension of magnon topology, but the AFM-interlayer cases need a stability check before their Chern numbers can be taken physically.","tokens_in":25779,"tokens_out":2631,"would_cite":true,"duration_ms":34536,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Honeycomb-ferromagnet magnon phases are captured by non-Hermitian SSH chains.","keywords":["magnons","honeycomb ferromagnet","Bogoliubov-de Gennes","non-Hermitian SSH chain","Chern number","Dzyaloshinskii-Moriya interaction","bilayer stacking","edge states"],"falsifier":"Compute the full complex spectrum of $\\eta h(k_y)$ for the AA- and AB-stacked antiferromagnetic-interlayer bilayers at the parameters used in the paper's Figure 10 and scan over $k_y$ and DMI strength; if any imaginary part becomes comparable to the real part, the $|E|$ plot would hide a spectral instability. Separately, diagonalize the real-space bosonic BdG Hamiltonian with open boundaries and check whether any excitation energies become negative or complex, which would settle whether the predicted edge states exist in a stable ground state.","tokens_in":24668,"feed_emoji":"🧲","tokens_out":6719,"duration_ms":68538,"temperature":0.7,"pith_summary":"The paper argues that the topological phases of magnons in honeycomb ferromagnets, including the existence, number, and location of edge states, can be read off from effective one-dimensional non-Hermitian Su-Schrieffer-Heeger (SSH) chain models labelled by momentum $k_y$. For monolayers, the conventional bulk-edge correspondence remains usable when Dzyaloshinskii-Moriya interaction (DMI) is absent, and edge states can even be the lowest-energy states; with DMI, the edge states tie to a nonzero Chern number of the bulk magnon bands. In bilayers, the stacking (AA versus AB), the sign of the interlayer coupling (ferromagnetic or antiferromagnetic), and the edge termination together determine the symmetry class, the Chern numbers, and the edge-state distribution. The point of the reduction is that a two-dimensional magnon problem becomes a set of tractable one-dimensional chains whose topology is already understood. If true, this gives a practical classification recipe for hexagonal ferromagnet magnonics.","feed_headline":"Non-Hermitian SSH chains govern honeycomb magnon edge states","feed_subtitle":"Stacking type, interlayer sign, boundary termination, and DMI fix the Chern number and edge-state count.","key_machinery":"The load-bearing object is the effective one-dimensional SSH chain $H(k_y)$, obtained by Fourier-transforming only along the edge and keeping open boundary conditions in the perpendicular direction; for bilayers it becomes a coupled SSH ladder that can often be decomposed into decoupled chains with a detuned on-site energy $\\pm J_{\\mathrm{in}}$. When pairing terms from antiferromagnetic interlayer couplings are present, the relevant operator $\\eta h(k)$ is non-Hermitian with respect to the bosonic metric $\\eta=\\sigma_z\\otimes I$, and its eigenproblem is equivalent to the original bosonic BdG problem. This machinery transfers topology from the two-dimensional bulk to the boundary bands: Berry curvature and Chern numbers computed on the pseudo-Hermitian $H_{\\sigma\\mathrm{BdG}}(k)$ predict the existence and winding of edge states, while the sign of the DMI sets the sign of the Chern number. For monolayer and ferromagnetic-bilayer cases the chain is Hermitian and conventional SSH winding arguments apply; for antiferromagnetic-bilayer cases the paper notes nonzero imaginary parts in the energies and plots $|E|$, so the chain's topology is presented as a partial, not fully justified, bulk-edge correspondence.","core_discovery":"The central discovery is that the single-particle bosonic Bogoliubov-de Gennes (BdG) Hamiltonians of honeycomb ferromagnets reduce, under open boundary conditions in one direction, to effective one-dimensional SSH-type chains $H(k_y)$ parameterized by the remaining momentum, and the topology of these chains explains the edge states of the original magnon system. In the monolayer case without DMI the BdG Hamiltonian has no pairing block and is Hermitian, so ordinary bulk-edge correspondence applies, producing four edge states for bearded boundaries and two for zigzag boundaries, and these states can sit at the lowest or highest magnon energies rather than near zero energy. Turning on DMI gives the bulk bands a nonzero Chern number (up to magnitude 1 for monolayers, up to magnitude 2 for AB-stacked bilayers), and the edge states are then connected to this Chern number through their winding, with the actual arrangements depending on the boundary. For bilayers, AA versus AB stacking and ferromagnetic versus antiferromagnetic interlayer coupling change the symmetry class and hence the Chern numbers and edge-state counts reported in the paper's classification table.","pith_inferences":["A testable extension would be to perturb a specific edge in a way that breaks the effective chain symmetry: zero-DMI edge states should disappear under such a perturbation, while nonzero-DMI Chern-connected states should remain, because the latter are protected by the bulk invariant rather than by chain fine-tuning.","The results suggest a material-search strategy: honeycomb ferromagnets with DMI, or with suitable stacking and boundary combinations, should show the predicted number of low- or high-energy magnon edge states in spin-resolved or thermal transport, and the sign of the DMI could be inferred from the chirality of those edge bands.","Because the paper works in linear spin-wave theory, a natural next step is to test whether magnon-magnon interactions alter the claim that some edge states can appear as ground states at zero DMI.","For the antiferromagnetic-interlayer bilayers, a cautious follow-up would be to define physical observables biorthogonally before assigning Chern numbers to the complex-energy states; if the imaginary parts grow with parameters, the classification would need revision."],"forward_implications":["The topological phase of a honeycomb-ferromagnet magnon system is fixed by a small parameter set: intralayer exchange, DMI, stacking form, interlayer coupling sign, and boundary termination, with the paper's table giving the resulting class, Chern number, and edge-state count.","Edge states in these systems need not be mid-gap: for bearded and zigzag terminations with zero DMI they can appear at the lowest or highest magnon energies, meaning they could be occupied ground states rather than zero-energy modes.","With nonzero DMI the edge-state count is tied to the bulk Chern number: monolayer cases show four edge states with $|C|=1$, AA ferromagnetic bilayers show eight with $|C|=1$, and AB ferromagnetic bilayers show eight with $|C|=2$, so edge states act as a boundary probe of the bulk invariant.","When DMI is zero, edge states can still exist for certain boundary conditions even though the Chern number vanishes, so the effective one-dimensional SSH or ladder topology, rather than the two-dimensional Chern number, controls the boundary physics.","In AB-stacked ferromagnetic bilayers, swapping the boundary types of the two layers leaves the bulk Hamiltonian and Chern number unchanged but moves edge states to different regions in $k_y$, showing that bulk topology alone does not fix the edge-state location."],"supporting_citations":[{"why":"Supplies the pseudo-Hermitian symmetry classification and the Chern-number formula used for $H_{\\sigma\\mathrm{BdG}}(k)$.","marker":"[40]"},{"why":"Establishes the equivalence between magnon BdG eigenproblems and non-Hermitian systems, which is the basis for the whole classification.","marker":"[35]"},{"why":"Connects bosonic BdG equations to non-Hermitian symmetry classes, supporting the use of non-Hermitian topology in the magnon context.","marker":"[38]"},{"why":"Provides the non-Bloch band framework for bosonic BdG systems used implicitly in the open-boundary chain analysis.","marker":"[39]"},{"why":"Supplies the coupled-SSH-chain correspondence for AB-stacked bilayer graphene that the bilayer magnon edge-state analysis transfers to bosonic chains.","marker":"[53]"},{"why":"Supports the connection between nonzero Chern numbers and the winding of edge states on cylinders, used when DMI is present.","marker":"[45–47]"}],"fun_headline_variants":["Magnon edge states from SSH chains with Chern numbers","Non-Hermitian SSH chains set magnon topology","Honeycomb magnons: edge states from effective SSH model","Topological magnons: SSH chains and Chern numbers","Edge states in honeycomb magnons from SSH chains"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the effective non-Hermitian Hamiltonians for the antiferromagnetic-interlayer bilayer cases describe a stable bosonic system, even though their spectra acquire small nonzero imaginary parts; the paper plots $|E|$ and asserts the imaginary parts are small compared with the real parts, but does not show that the complex energies correspond to a physically admissible bosonic ground state or that the Chern number remains well-defined in their presence.","fun_headline_variants_meta":{"raw":{"variants":["Magnon edge states from SSH chains with Chern numbers","Non-Hermitian SSH chains set magnon topology","Honeycomb magnons: edge states from effective SSH model","Topological magnons: SSH chains and Chern numbers","Edge states in honeycomb magnons from SSH chains"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000576,"raw_usage":{"total_tokens":2726,"prompt_tokens":960,"completion_tokens":1766,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":576,"completion_tokens_details":{"reasoning_tokens":1688}},"tokens_in":576,"tokens_out":1766,"duration_ms":14021,"temperature":1.0,"reasoning_tokens":1688,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T05:01:27.056364+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the full complex spectrum of $\\eta h(k_y)$ for the AA- and AB-stacked antiferromagnetic-interlayer bilayers at the parameters used in the paper's Figure 10 and scan over $k_y$ and DMI strength; if any imaginary part becomes comparable to the real part, the $|E|$ plot would hide a spectral instability. Separately, diagonalize the real-space bosonic BdG Hamiltonian with open boundaries and check whether any excitation energies become negative or complex, which would settle whether the predicted edge states exist in a stable ground state.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the pseudo-Hermitian symmetry classification and the Chern-number formula used for $H_{\\sigma\\mathrm{BdG}}(k)$."},{"cited_title":"Laurell and G","cited_arxiv_id":null,"evidence_quote":"Establishes the equivalence between magnon BdG eigenproblems and non-Hermitian systems, which is the basis for the whole classification."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Connects bosonic BdG equations to non-Hermitian symmetry classes, supporting the use of non-Hermitian topology in the magnon context."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the non-Bloch band framework for bosonic BdG systems used implicitly in the open-boundary chain analysis."}],"review_version":1}