{"id":"7aeee51a-a37a-4fb6-9aec-ad9083c86ba2","arxiv_id":"2608.02531","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"Domain-resolved Raman and infrared spectroscopy on twinned Na3Co2SbO6 reveals a 24% in-plane magnon-energy anisotropy that dominant anisotropic Kitaev/Γ exchange explains, not g-factor anisotropy.","lead":"The paper combines infrared and micro-Raman spectroscopy to separate the magnetic signal of one crystal twin from averaged signals in twinned Na3Co2SbO6 crystals. It shows that the one-magnon energy varies strongly with in-plane field direction, much more than the g-factor variation, pointing to bond-dependent exchange as the dominant source.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Single-domain assignment is the load-bearing link; unresolved micro-twins or an intrinsic middle peak would invalidate the 24.1% anisotropy claim.","rationale":"The reader's weakest_assumption correctly identifies the single-domain assignment as the critical juncture: the headline numbers (39.3, 50.0 cm−1; 24.1% anisotropy) are direct outputs of the Raman-FIRMS correlation, and the conclusion that exchange dominates relies on these numbers being intrinsic. My stress-test agrees with that assessment. The multimodal approach is innovative and the agreement between Raman and FIRMS provides internal consistency, but internal consistency does not break the circularity if both measurements share the same twinned-crystal ambiguity. The paper itself flags the middle peak and the rarity of single-domain crystals, which are exactly the points needing independent verification. The LSWT/CW modeling has additional caveats (Eq. 1 not derived in main text; parameters not uniquely constrained), but those affect the interpretation rather than the primary spectroscopic observation. Therefore the reader's CONDITIONAL verdict is appropriate; no change is recommended. The concrete test—sub-micron domain mapping and reproducibility checks—would either confirm the assignment or expose a domain-averaging artifact.","tokens_in":10594,"tokens_out":6961,"duration_ms":83398,"concrete_test":"Map the exact crystal used for micro-Raman with a sub-micron-resolution structural probe (e.g., X-ray nanodiffraction or SHG microscopy) to confirm the probed area is a single domain with no twin boundaries below the optical resolution. Additionally, repeat the micro-Raman measurement on at least 10 independently selected 'single domains' from different crystals and compare the extracted E_a and E_b at B=14 T; if the peak energies scatter by more than ~1 cm−1 or any sample shows a double-peak structure, the current assignment is unreliable and the 24.1% anisotropy claim would need revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central quantitative claim—E_a=39.3 cm−1, E_b=50.0 cm−1 at 14 T and a 24.1% in-plane anisotropy exceeding the g-factor anisotropy—rests on identifying the single-domain magnon responses in the mosaic FIRMS spectra by comparison with micro-Raman on a 'single domain' (Fig. 3). The weakest link is the assumption that the Raman spot truly probes one crystallographic domain. The paper states that 'single-domain crystals are relatively rare' and that the middle FIRMS peak 'likely arises from slight misalignment among crystals in the mosaic or from magnetic scattering at domain walls'—both are untested assumptions. If the micro-Raman 'single domain' contained unresolved twins (e.g., twinning on a scale below the optical spot) or if the middle peak were actually an intrinsic magnetic mode, then the extracted E_a and E_b would be a weighted average or a misidentified branch, and the 24.1% anisotropy would not be intrinsic. The angular scan in Fig. 3e shows a smooth twofold pattern, but an unresolved twin distribution would also produce a smooth distorted pattern, so this does not rule out the concern. The LSWT modeling is explicitly acknowledged as not sufficient to determine all exchange parameters independently, and the CW fit (Eq. 1) is stated without derivation in the main text, which further underscores that the exchange-domination conclusion is model-dependent. The experimental anisotropy is the primary evidence, so verifying the single-domain purity is the decisive check.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports a multimodal optical study of the Kitaev-Heisenberg candidate Na3Co2SbO6, combining far-infrared magneto-spectroscopy (FIRMS) on a mosaic of 68 twinned crystals with domain-resolved micro-Raman magneto-spectroscopy under in-plane fields up to 14 T. The authors assign the three closely spaced one-magnon peaks in the FIRMS spectra to three twin orientations, use micro-Raman to isolate the response of a single crystallographic domain, and obtain single-domain magnon energies E_a = 39.3 cm^-1 and E_b = 50.0 cm^-1 at 14 T, corresponding to an in-plane anisotropy (E_b-E_a)/sqrt(E_aE_b) = 24.1%, far larger than the 5.7% anisotropy of the measured g-factors. A Curie-Weiss expression (Eq. 1) is used to fit g_a, g_b and two Curie-Weiss temperature differences, and linear spin-wave theory with anisotropic Kitaev and symmetric off-diagonal exchange parameters reproduces the field evolution of the one-magnon modes for B > 2 T. The paper concludes that bond-dependent exchange interactions, particularly K and Gamma, dominate the in-plane magnon anisotropy.","tokens_in":11027,"tokens_out":5083,"duration_ms":62421,"significance":"If the single-domain assignment is reliable, the paper establishes a valuable and broadly applicable strategy for extracting intrinsic magnetic excitations from twinned quantum magnets. The direct single-domain Raman measurement is a genuine methodological advance over spatially averaged FIRMS/INS probes, and the claimed 24.1% magnon anisotropy versus 5.7% g-factor anisotropy is a striking, falsifiable result that would place tight constraints on the spin Hamiltonian of Na3Co2SbO6. The paper is also appropriately cautious in acknowledging that q = 0 magnon data are insufficient to determine all exchange parameters independently. However, the central quantitative claim rests on the assumption that the micro-Raman spot probes a single crystallographic domain; this assumption is currently supported by limited evidence, and the smooth twofold angular dependence alone does not rule out unresolved micro-twins. The experimental claim is plausible but not yet fully load-bearing, and the manuscript therefore requires additional verification before publication.","major_comments":[{"comment":"The central quantitative result—E_a = 39.3 cm^-1, E_b = 50.0 cm^-1, and the 24.1% anisotropy—is obtained from micro-Raman spectra on a spot described as a 'single domain.' The text states that single-domain crystals are relatively rare and that twinning is pronounced, but no evidence is provided that the region probed by the optical spot is a single domain at the scale of the probe volume. The ARPRS structural-domain map (Fig. S5) and the angular scan of Fig. 3e are necessary but not sufficient: an unresolved distribution of micro-twins inside the laser spot would also produce a smooth, apparently twofold response. Please provide spatially resolved maps at the true probe scale, line scans across the crystal, or a second independent single-domain probe, and show that two different spots on the same 'single domain' yield identical magnon energies. This verification is load-bearing for the","section":"Fig. 3 and §3"},{"comment":"The assignment of the weaker middle peak in the triple-peak FIRMS structure is explicitly tentative ('likely arises from slight misalignment ... or from magnetic scattering at domain walls'). This tentative assignment is used to label the open-symbol single-domain branches in Fig. 2a,b and hence to construct the field-dependent data fitted in Fig. 4c. If the middle peak is actually an intrinsic magnetic mode, the branch-tracing in the mosaic FIRMS spectra may be contaminated, even if the Raman-derived energies at 14 T are unaffected. Please provide a more direct test—for example, by comparing spectra from a true twin-free crystal (e.g., Ref. 46) or by using polarization-selective FIRMS—or quantitatively estimate how an unresolved middle peak would shift the extracted E(B) branches and the fitted parameters.","section":"Fig. 2 and Fig. 3c,d"},{"comment":"The Curie-Weiss energy expression used to fit g_a, g_b, Theta_ba, and Theta_bc* is stated without derivation in the main text. This expression is the phenomenological basis for the exchange-anisotropy conclusion, and the manuscript notes that the fitted g-factors differ significantly from low-field THz values in Ref. 46. Please provide a full derivation in the Supporting Information, clarify the range of validity of the high-field approximation, and discuss why the g-factors extracted from Eq. (1) are smaller than those obtained from low-field magneto-THz data. Without this, a reader cannot assess whether the apparent anisotropy is partly imposed by the fitting form rather than being independently constrained by the data.","section":"Eq. (1), §4"},{"comment":"The paper acknowledges that the field evolution of q = 0 magnons 'is not sufficient to determine all exchange parameters independently,' yet it reports a specific LSWT parameter set (K_X = K_Y = -4.50 meV, K_Z = -4.00 meV, Gamma_X = Gamma_Y = 1.04 meV, Gamma_Z = 1.50 meV, Gamma' = 0.56 meV) and states that the data are 'quantitatively captured.' Please provide confidence intervals, parameter correlations, and a discussion of which parameter combinations are actually constrained (e.g., K_Z - K_X, Gamma_Z - Gamma_X) versus those fixed by assumption. In particular, the assumption that C3 symmetry breaking enters only through Gamma while Gamma' is kept isotropic should be justified, since the text notes that C3 breaking in Gamma and Gamma' produces compensating effects. The qualitative conclusion of exchange-dominated anisotropy may be robust, but the specific Hamiltonian should not be prese","section":"Fig. 4d, §4"}],"minor_comments":[{"comment":"Typo: 'challenges posted by crystal twinning' should read 'challenges posed by crystal twinning.'","section":"Abstract/Introduction"},{"comment":"The notation 'B||a'+'-60°' is used to denote a field direction, but the relation between the laboratory frame (a', b') and the crystallographic axes (a, b) is not fully specified in the main text. A brief definition in the figure caption or text would improve clarity.","section":"§3, Fig. 3"},{"comment":"The statement 'g' ∝ δE/δB' is schematic; the proportionality constant depends on the magnon g-factor and the detailed field dependence. Please use a more explicit definition (e.g., dE/dB = g μ_B for a linear branch) and clarify how the 5.7% anisotropy is computed.","section":"Fig. 3e / §3"},{"comment":"The fitted values g_a = 4.75(2), g_b = 5.32(2), Theta_ba = 2.73(5) K, and Theta_bc* = 21.48(18) K are reported without a description of the fitting procedure (e.g., weighting, background, correlated errors). Please include the residual plot or a statement of the reduced chi-square in SI.","section":"§4, Fig. 4c"},{"comment":"The LSWT intensity map is described as 'overlaid' with experimental data, but the plotted experimental guide lacks error bars. Adding representative error bars to the symbols in Fig. 4d would help the reader judge the quality of the agreement.","section":"§4, Fig. 4d"},{"comment":"The citation to Ref. 46 (Li et al., Chinese Physics Letters 2025) is central to the twin-free comparison, and Ref. 21 (Li et al., Phys. Rev. X 2022) is central to the twinning. Please ensure both are quoted with complete author lists and that the discussion of Ref. 46 explicitly states what data are being compared.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is intellectually promising and the multimodal approach is potentially important, but the single-domain purity issue is the decisive experimental point. I would ask the authors to supply the missing spatial/microscopic evidence and the derivation of Eq. (1) before I could recommend acceptance. If the single-domain verification cannot be provided, the 24.1% anisotropy claim and the exchange-anisotropy conclusion would need to be substantially weakened. The paper is not fatally flawed in principle; the required revision is local but load-bearing."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things. First, the paper actually measures single-domain magnon energies in Na3Co2SbO6: 39.3 cm−1 along a and 50.0 cm−1 along b at 14 T, giving a 24.1% in-plane anisotropy. That is a direct measurement, not a fit, and it is the main result. Second, the spin-Hamiltonian modeling attached to it is visibly under-determined, and the paper admits as much. The quantitative exchange parameters should be treated as illustrative constraints, not as a definitive model.\n\nWhat is genuinely new: the multimodal workflow—using micro-Raman to resolve individual twin domains and then mapping those single-domain spectra onto the macroscopic FIRMS mosaic—is a useful and transferable approach. It lets them isolate twin contributions that would otherwise hide the intrinsic anisotropy. The 24% number is new and far exceeds the 5.7% g-factor anisotropy, so the conclusion that anisotropic exchange dominates the magnon anisotropy is well supported by the Raman data alone.\n\nThe writing is clean and the experimental work is careful. They use angle-resolved polarized Raman to map domains before positioning the spot, and they are explicit about the limitations of their LSWT fit—they say that q=0 magnons cannot determine all exchange parameters independently. I also appreciate that they flag the middle FIRMS peak as likely extrinsic.\n\nSoft spots, in proportion. The single-domain purity is the load-bearing assumption. The angular scan in Fig. 3e is consistent with a single domain, but as your stress-test note correctly points out, a pair of unresolved twins with unequal weights also produces a smooth twofold pattern. The paper does not report a direct check—like imaging the same spot after the measurement or stating the domain size relative to the Raman spot. That is a legitimate gap, though not a fatal one; the Raman mode is sharp and single, which is what you would expect from one domain, and the ARPRS mapping gives some confidence. Still, I would want this addressed.\n\nEq. 1 appears without derivation and is used to fit g_a, g_b, and the Θ values, and the LSWT is then constrained by that fit. The paper itself says the parameter set is not unique. So the exchange-anisotropy conclusion is partly model-dependent, but note that the experimental anisotropy claim does not depend on that model. The middle-peak assignment is speculative but not load-bearing for the main claim.\n\nOverall: this is a solid experimental paper that deserves a serious referee. The measurement is likely correct, the method is useful, and the limitations are honestly stated. I would send it to peer review with the request that the single-domain purity be documented more explicitly.\n\nFor your own work, the 24% single-domain anisotropy is now the number to cite when discussing in-plane anisotropy in Na3Co2SbO6.","headline":"A solid spectroscopic paper that delivers a clean single-domain magnon anisotropy measurement in a Kitaev candidate; the Hamiltonian modeling is under-determined, but the measured anisotropy is the headline and it holds up.","tokens_in":11493,"tokens_out":3442,"would_cite":true,"duration_ms":40880,"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":"A Kitaev-Heisenberg magnet's intrinsic spin waves are 24% anisotropic in-plane once twin-domain replica peaks are filtered out.","keywords":["Na3Co2SbO6","Kitaev-Heisenberg quantum magnet","magnon anisotropy","twin domains","magneto-infrared spectroscopy","micro-Raman spectroscopy","linear spin-wave theory","bond-dependent exchange"],"falsifier":"Measure a confirmed twin-free Na3Co2SbO6 crystal (as in previously reported low-field THz work) at 14 T with the field along the in-plane a and b axes: if the spin-polarized one-magnon energies are not 39.3 and 50.0 cm−1 with the same 24% splitting, the domain assignment fails. Alternatively, if a truly single-domain Raman spot ever resolves two closely spaced one-magnon lines, the middle FIRMS peak would be intrinsic rather than extrinsic.","tokens_in":10542,"feed_emoji":"🧲","tokens_out":6703,"duration_ms":65593,"temperature":0.7,"pith_summary":"The paper reports that in the spin-polarized phase of the Kitaev-Heisenberg quantum magnet Na3Co2SbO6, the intrinsic single-domain magnon energy at 14 T is 39.3 cm−1 along a and 50.0 cm−1 along b, giving an in-plane anisotropy (Eb−Ea)/√(EaEb) of 24.1%. This is roughly four times the 5.7% anisotropy of the measured g-factor, so the authors conclude that the magnon anisotropy is dominated by bond-dependent exchange interactions, not the Zeeman term. They reach this conclusion by combining far-infrared magneto-spectroscopy on a 68-crystal mosaic with domain-resolved micro-Raman magneto-spectroscopy on individual twin domains. The far-infrared triple-peak structure, which naively looks intrinsic, is shown to be a superposition of single-domain modes from different twin orientations plus an extrinsic middle peak. If the claim holds, it provides direct optical constraints on the anisotropic spin Hamiltonian of a honeycomb cobaltate and a broadly applicable method for studying multidomain quantum magnets.","feed_headline":"Magnon energy swings 24% between crystal axes in a Kitaev magnet","feed_subtitle":"Infrared plus micro-Raman on twinned crystals isolates single-domain spin waves; bond-dependent exchange sets the anisotropy.","key_machinery":"The central mechanism is the combination of two optical spectroscopies with different beam footprints: far-infrared magneto-spectroscopy (FIRMS) on a millimeter-scale mosaic averages over twin domains, producing a triple-peak 'one-magnon' structure, while angle-resolved polarized micro-Raman magneto-spectroscopy (ARPRS) spatially resolves single domains and, by rotating the field in 30° steps, reproduces the triple peak as the outer modes from three twin orientations plus an extrinsic middle mode. The analytic engine is a Curie-Weiss extension of the one-magnon energy (eq. 1), which expresses the energy as a function of field angle and direction via anisotropic Curie-Weiss temperatures, with","core_discovery":"The paper's core claim is that the intrinsic magnetic excitation spectrum of a twinned Kitaev-Heisenberg crystal can be reconstructed by correlating two optical probes: far-infrared magneto-spectroscopy (FIRMS) averages over a mosaic and shows a triple-peak one-magnon structure, while micro-Raman magneto-spectroscopy with micrometer resolution resolves individual structural domains and reproduces that triple peak as the sum of single-domain modes from twin orientations differing by 60°. The isolated single-domain mode has twofold in-plane symmetry with E_a = 39.3 cm−1 and E_b = 50.0 cm−1 at 14 T, so (E_b−E_a)/√(E_aE_b) = 24.1%. Since the in-plane g-factor anisotropy obtained from the high-fi","pith_inferences":["If the 24% in-plane magnon anisotropy is intrinsic, then models of Na3Co2SbO6 that treat in-plane exchange as nearly isotropic will fail to reproduce spin-polarized dynamics; a direct test would be to measure a confirmed twin-free crystal at 14 T along a and b and check for E_a ≈ 39.3 cm−1 and E_b ≈ 50.0 cm−1.","The same twin-replica mechanism likely applies to other monoclinic honeycomb cobaltates and α-RuCl3-like systems; re-analyzing existing THz data with domain-resolved Raman could reveal that some reported multi-mode structures are twin superpositions rather than intrinsic excitations.","The factor-of-four gap between magnon and g-factor anisotropy implies that the C3-breaking monoclinic distortion affects bond-dependent exchange more strongly than the local g-tensor; if this pattern holds in other cobaltates, the role of monoclinic distortion in stabilizing particular magnetic orders deserves reexamination.","Because the CW-constrained LSWT fit uses only q=0 magnons, the exchange parameters are not unique; extending the measurement to finite momentum via inelastic neutron scattering on aligned single-domain mosaics could break the degeneracy and test whether the K/Γ parameter set survives.",""],"forward_implications":["The prominent triple-peak structure in far-infrared magneto-spectra of Na3Co2SbO6 in the spin-polarized phase is primarily a twinning artifact: the outer peaks are single-domain magnons from differently oriented twins, and the middle peak is extrinsic (mosaic misalignment or domain-wall scattering), so spatially averaged spectra should not be interpreted as intrinsic modes.","The single-domain magnon energies above roughly 2 T can serve as quantitative constraints on the anisotropic Kitaev-Heisenberg-Γ Hamiltonian, pinning down K_X = K_Y = −4.50 meV, K_Z = −4.00 meV, Γ_X = Γ_Y = 1.04 meV, Γ_Z = 1.50 meV, and Γ′ = 0.56 meV, subject to the paper's own caveat that q=0 data alone do not determine all exchange parameters independently.","The high-field g-factors ga = 4.75(2) and gb = 5.32(2) are significantly smaller than previously reported low-field values, implying that field-dependent renormalization of the Zeeman response is substantial in this material.","The two-magnon branch at roughly twice the one-magnon energy with twice the slope, plus a weak replica-like branch merging at high fields, are consistent with ΔS=±1 magnetic-dipole selection rules and with a higher-energy magnon branch made active by interlayer coupling, zone folding, or disorder.","The combined FIRMS-plus-micro-Raman protocol is presented as a generally applicable route to extracting intrinsic spin dynamics in any multidomain quantum magnet where the domain size is small compared with the THz/IR beam.",""],"fun_headline_variants":["Spin-wave anisotropy in Kitaev magnet traced to bond-dependent exchange","Two-probe method reveals 24% magnon anisotropy in twinned Kitaev crystal","Isolating single-domain magnons: 24% in-plane anisotropy in Kitaev magnet","Intrinsic magnon anisotropy in Kitaev magnet: 24% between axes","Domain-resolved Raman + FIRMS expose 24% magnon anisotropy in Kitaev magnet"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The single-domain anisotropy analysis rests on the assumption that the micro-Raman spot truly selects one crystallographic domain and that the middle FIRMS peak is extrinsic misalignment or domain-wall scattering; if the 'single domain' still contains unresolved twins, or if the middle peak is actually an intrinsic magnetic excitation, the extracted E_a = 39.3 cm−1, E_b = 50.0 cm−1, and the 24.1% anisotropy would be misidentified.","fun_headline_variants_meta":{"raw":{"variants":["Spin-wave anisotropy in Kitaev magnet traced to bond-dependent exchange","Two-probe method reveals 24% magnon anisotropy in twinned Kitaev crystal","Isolating single-domain magnons: 24% in-plane anisotropy in Kitaev magnet","Intrinsic magnon anisotropy in Kitaev magnet: 24% between axes","Domain-resolved Raman + FIRMS expose 24% magnon anisotropy in Kitaev magnet"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000866,"raw_usage":{"total_tokens":3592,"prompt_tokens":744,"completion_tokens":2848,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":488,"completion_tokens_details":{"reasoning_tokens":2742}},"tokens_in":488,"tokens_out":2848,"duration_ms":20402,"temperature":1.0,"reasoning_tokens":2742,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T05:16:30.096192+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure a confirmed twin-free Na3Co2SbO6 crystal (as in previously reported low-field THz work) at 14 T with the field along the in-plane a and b axes: if the spin-polarized one-magnon energies are not 39.3 and 50.0 cm−1 with the same 24% splitting, the domain assignment fails. Alternatively, if a truly single-domain Raman spot ever resolves two closely spaced one-magnon lines, the middle FIRMS peak would be intrinsic rather than extrinsic.","supporting_citations":[],"review_version":1}