{"id":"cf5181a7-d49e-46c2-917b-4783f1b1c938","arxiv_id":"2607.14893","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Neutron spin-wave data plus spin-wave fits show NiI2 has a sizable Kitaev interaction K=3.33 meV and off-diagonal Gamma=0.37 meV, attributed to iodine ligand spin-orbit coupling.","lead":"A new neutron-scattering study of the layered magnet NiI2 finds large bond-direction-dependent (Kitaev) magnetic interactions even though the nickel ions have quenched orbital angular momentum. The authors attribute the anisotropy to iodine-ligand spin-orbit coupling, which would widen the search for Kitaev quantum-spin-liquid materials.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Omission of the symmetry-allowed Γ′ term biases the extracted K and Γ; the ligand-Kitaev claim is conditional on an untested model restriction.","rationale":"The reader's weakest assumption—that the in-plane easy axis cannot be explained by SIA and must arise from bond-dependent terms—is a valid starting point, and the SIA part of that argument is actually well founded. However, the more load-bearing and concrete gap is the omission of Γ′, a symmetry-allowed anisotropic exchange term that the paper itself acknowledges but never quantitatively excludes. The phase diagram handles only K and Γ; the reported parameter uncertainties cover statistical fluctuations, not model ambiguity. Therefore the fitted K = 3.33 meV is not as secure as the abstract implies. This does not overturn the paper—it strengthens the CONDITIONAL verdict rather than moving it. The proposed refit would directly test whether the Kitaev claim is robust to including Γ′. Independent support exists: the LSWT calculation with SpinW reproduces many measured spectral features (flat band top, kink at 5.2 meV, 3.1-meV gap, and 53.4° canting angle), which indicates the fit is high quality. But quality within a restricted model does not resolve the model-choice question.","tokens_in":13387,"tokens_out":5815,"duration_ms":61745,"concrete_test":"Refit the measured INS spectra and the magnetic-structure constraints (canting angle, in-plane propagation direction, period) with Eq. (1) extended to include the symmetry-allowed Γ′ term (and, in a second pass, a single-ion D(Sz)^2 term), keeping J1, J2, J3, and Jp2 free. Compare the best-fit values of K, Γ, and Γ′ and the global χ². The claim survives if the extended model still gives K ≳ 2 meV and Γ′ ≲ 0.2 meV with statistically indistinguishable χ²; it fails if a fit with Γ′ ~ 1 meV and K < 1 meV reproduces the gap and canting equally well.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—a sizable ligand-driven Kitaev interaction—rests on the fitted minimal model, but that model drops a symmetry-allowed term without quantitative justification. The text explicitly lists 'off-diagonal interactions Γ and Γ′' as allowed anisotropic terms, yet Eq. (1) contains only K and Γ. The exclusion argument for SIA is sound: for S=1 Ni2+ in the D3d site, the only SIA easy axis is z, so SIA alone cannot produce the observed in-plane easy axis. However, that argument does not constrain Γ′. Γ′ can also contribute to in-plane anisotropy, to the 3.1-meV spin-wave gap, and to the canting angle that sets K. The phase diagram in Fig. 3 varies only K/J1 and Γ/J1, and the reported uncertainties (e.g., K = 3.33(0.032) meV) are statistical only, not systematic. If Γ′ were non-negligible, K and Γ could be substantially renormalized. Thus the 'compelling experimental evidence' for the ligand-Kitaev mechanism is not uniquely established; the paper's conclusion is conditional on Γ′ = 0, a restriction that is stated but not tested.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports inelastic neutron scattering, magnetization, magnetic-structure, and linear-spin-wave studies of the S=1 triangular-lattice van der Waals antiferromagnet NiI2. The authors construct a minimal spin Hamiltonian containing nearest-neighbor Heisenberg, Kitaev, and off-diagonal Γ exchange, together with in-plane J2 and J3 and interlayer Jp2. Fitting the spin-wave dispersions and the canted proper-screw ground state yields J1=-5.86 meV, K=3.33 meV, Γ=0.37 meV, J2=0.28 meV, J3=1.99 meV, and Jp2=0.78 meV. They argue that these bond-dependent interactions explain the 3.1 meV spin-wave gap, the 5.2 meV kink, the flat band top, and the in-plane easy axis, and they conclude that NiI2 provides compelling experimental evidence for a ligand-driven Kitaev mechanism.","tokens_in":13594,"tokens_out":3573,"duration_ms":37713,"significance":"If the extracted parameters are reliable, this is an important result: it would establish a sizable Kitaev-type interaction in a high-spin triangular-lattice magnet where the magnetic Ni2+ ion has quenched orbital angular momentum, thereby supporting the ligand-SOC route to bond-dependent anisotropy. The manuscript has notable strengths: high-quality single-crystal INS data that resolve the band top and a 3.1 meV gap not seen in previous work, a phase diagram that connects K and Γ to the canted proper-screw state, and a clear comparison with prior transport and thermodynamic studies. However, the central quantitative claim is based on a constrained fit to a model that excludes a symmetry-allowed Γ′ term, and the reported uncertainties are purely statistical. No sensitivity or identifiability analysis is presented to show that K and Γ are uniquely determined by the data rather than by the model restriction. The claim of 'compelling experimental evidence' is therefore not yet fully established.","major_comments":[{"comment":"The symmetry analysis explicitly lists 'Kitaev interaction K, and off-diagonal interactions Γ and Γ′' as allowed anisotropic terms, but Eq. (1) contains only K and Γ. The exclusion argument for single-ion anisotropy is sound—only the z-axis easy axis is symmetry-allowed for SIA—but that argument does not constrain Γ′. Γ′ can contribute to the in-plane anisotropy, to the 3.1 meV gap, and to the canting angle, all of which are used to fix K and Γ. As written, the extracted K=3.33 meV and Γ=0.37 meV are conditional on Γ′=0, a restriction that is neither derived from symmetry nor tested against the data. I request a quantitative benchmark: fit the same INS and magnetic-structure constraints with Eq. (1) plus Γ′ (and, if desired, a symmetry-breaking in-plane SIA associated with the small lattice distortion at TN2), and report the resulting Γ′ and the renormalized K and Γ, or a bound such as |","section":"Model, Eq. (1) and preceding paragraph"},{"comment":"The paper states that parameters were obtained by 'jointly and self-consistently constraining the spin-wave spectra and the magnetic structure—specifically the canting angle, in-plane propagation direction, and period.' However, no identifiability analysis is shown. The reported uncertainties (e.g., K=3.33(0.032) meV) are statistical only, and no correlation matrix or leave-one-out tests are given. The canting angle that is central to constraining K is reported experimentally as 55±10 degrees, a relative uncertainty of about 18%; it is not clear how this propagates into K. Moreover, the statement that 'K/Γ≈9 will give a proper canting angle' is not accompanied by a plot of the low-energy gap, spin-wave dispersions, or canting angle as functions of K and Γ with other parameters fixed. Please provide a sensitivity analysis for at least the pairs (K, Γ), (K, Γ′), and (K, SIA), showing that","section":"Parameter extraction and reported uncertainties after Eq. (1)"},{"comment":"The paper openly acknowledges that the LSWT intensity near the K point is 'significantly weaker in the experimental data than in the LSWT calculation,' and attributes this to magnon decay/anharmonicity and omitted Umklapp processes. This is a relevant discrepancy because the affected region includes the high-energy constraints used to determine J2 and J3. The two-magnon density of states shown in Fig. S6 is offered as support, but no quantitative calculation of the one-magnon spectral function with decay is presented, nor is the effect of the Umklapp processes estimated. If the high-energy part of the fit is unreliable, the extracted J2 and J3—and indirectly the low-energy parameters through the global fit—could be affected. Please quantify the spectral-weight mismatch, e.g., by convoluting LSWT with an energy-dependent broadening or by showing that the fitted parameters are insensitive","section":"Fig. 2(b) and high-energy spectral weight discussion"},{"comment":"The manuscript notes that a biquadratic term B(Si·Sj)2 'may be present' and that INS is only weakly sensitive to it, so it is not included. For S=1 Ni2+ with strong Hund coupling, ligand-mediated biquadratic exchange can be non-negligible, as the authors themselves cite (Ref. [47] and [57]). Since a biquadratic term can affect the magnetic ground state, the spin-wave gap, and the intensity distribution, the assertion that the minimal model 'captures the essence' should be accompanied by at least an order-of-magnitude estimate of B from first principles or a statement of the INS constraint on B. This is a supporting point rather than a block on the central claim, but it adds to the concern that the reported K is model-dependent.","section":"Biquadratic term, last paragraph before Fig. 4"}],"minor_comments":[{"comment":"Typos and grammar: 'stabalizes' for 'stabilizes' (twice), 'experimentlly' in Fig. 4 caption, 'the our model parameters', 'first-principle calculations' instead of 'first-principles calculations', and 'Lattice Bragg peaks' should be 'nuclear Bragg peaks' or simply 'Bragg peaks'.","section":"Throughout"},{"comment":"The notation '3.33(0.032)' is not standard in this journal. Please state explicitly that the parentheses contain one standard deviation or 95% confidence intervals, and clarify whether these are from the LSWT fit to INS only or from the joint fit including magnetic-structure constraints.","section":"Parameter list after Eq. (1)"},{"comment":"The reference list contains several very recent items (e.g., Refs. [34], [44], [45], [49], [55], [58], [59]) that are appropriate for the field, but the main text does not cite any reference for the 'first-principles calculations' that are said to guide the model. If those calculations are reported in the SM, the main text should point the reader to them explicitly.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"This is a technically competent paper with potentially important implications for Kitaev physics in high-spin magnets. The central issue is model identifiability: the omission of the symmetry-allowed Γ′ term and the absence of a sensitivity analysis leave open the possibility that the extracted K is substantially renormalized once Γ′ is allowed. The requested Γ′ benchmark and sensitivity study are feasible within the scope of a revision and would strengthen the paper. I do not see a basis for rejection, but the conclusion 'compelling experimental evidence' should be softened unless the omitted-term issue is addressed quantitatively."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead the NiI2 paper by Xu et al. It's the most complete spin-wave measurement of this compound so far—band top at 37.5 meV, a clean 3.1 meV gap, and a fit that reproduces the dispersion with K=3.33 meV and Γ=0.37 meV. That's a genuine step beyond Ref. 49, which missed the gap and the full bandwidth. The INS work is careful, the LSWT calculation is appropriate for the incommensurate structure, and the gap explanation is plausible. They also released the neutron coverage simulation code on Zenodo, which is good practice.\n\nThe main thing to know: the abstract's 'compelling experimental evidence' is too strong. The model in Eq. (1) omits Γ′, which the text itself lists as symmetry-allowed. The argument that single-ion anisotropy can't produce the in-plane easy axis is fine, but it doesn't constrain Γ′. Γ′ can also open the gap, tilt the canting angle, and renormalize the fitted K and Γ. The authors don't report a fit with Γ′ included, so the error bars are conditional on Γ′=0. That's the kind of model restriction you have to test before drawing a mechanism conclusion.\n\nAlso, the 'self-consistent' constraint from the canting angle and propagation direction means those outputs are not independent predictions—they were used as input. That's not fatal, but it weakens the claim of agreement. And the high-energy spectral weight mismatch near K is waved off with magnon decay; plausible, but not quantified.\n\nThe paper deserves to be sent to referees. A good referee should ask for a sensitivity analysis: vary Γ′, add SIA, maybe a biquadratic term, and show how K and Γ shift. If K stays around 3 meV, the case is much stronger. As it stands, this is a solid experimental paper with an overinterpreted conclusion. I'd still cite it for the data, and I'd use it in a reading group to talk about parameter identifiability in spin wave fits.\n\nMy recommendation: send to peer review, with the expectation that the authors will need to either do the extra fits or soften the abstract.","headline":"Good new data, careful fit, but the headline overclaims: the model drops Γ′ without a test, so the Kitaev parameters are conditional, not confirmed.","tokens_in":14249,"tokens_out":3872,"would_cite":true,"duration_ms":38055,"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":"Experiments show that the spin-1 triangular magnet NiI2 hosts a sizable Kitaev interaction driven by the spin-orbit coupling of iodine ligands rather than by the nickel ions themselves.","keywords":["Kitaev interaction","ligand spin-orbit coupling","triangular lattice","NiI2","spin-1 magnet","inelastic neutron scattering","spin-wave gap","bond-dependent anisotropy"],"falsifier":"Refit the same inelastic neutron scattering data with a model that includes in-plane single-ion anisotropy terms (or Gamma-prime) while allowing K and Gamma to be zero; if an equally good fit to the gap, the kink, and the canting angle emerges, the claim that the anisotropy is bond-dependent collapses. A polarized-neutron measurement that directly resolves the anisotropy contribution would settle the question.","tokens_in":13212,"feed_emoji":"🧲","tokens_out":7449,"duration_ms":58415,"temperature":0.7,"pith_summary":"The paper aims to establish that NiI2, a spin-1 triangular-lattice antiferromagnet, is a genuine platform for Kitaev physics — bond-dependent exchange anisotropy where the coupling strength depends on the spatial direction of the bond — even though its magnetic Ni2+ ions have no orbital angular momentum. By combining inelastic neutron scattering, magnetization measurements, magnetic-structure analysis, first-principles calculations, and linear spin-wave fits, the authors extract a minimal Hamiltonian with a large Kitaev term K = 3.33 meV, a second bond-dependent term Gamma = 0.37 meV, and a ferromagnetic nearest-neighbor coupling J1 = -5.86 meV. The claimed origin is the strong spin-orbit coupling of the iodine ligands, which transfers bond-dependent anisotropy through the Ni-I-Ni exchange paths. If this is correct, it provides a concrete route to Kitaev materials in high-spin systems with weak magnetic-ion spin-orbit coupling and extends Kitaev physics to the triangular lattice.","feed_headline":"Iodine's spin-orbit coupling drives Kitaev interaction in NiI2","feed_subtitle":"Neutron spin-wave data put K at 3.33 meV, showing ligand-mediated bond anisotropy in a spin-1 triangular magnet.","key_machinery":"The key object is a minimal spin Hamiltonian on the triangular lattice written in the Kitaev basis, in which each nearest-neighbor bond is labeled X, Y, or Z by the normal to its Ni-I-Ni bond plane: nearest-neighbor Heisenberg exchange J1 plus the Kitaev interaction K and off-diagonal Gamma, further-neighbor Heisenberg exchanges J2 and J3, and an interlayer coupling Jp2. Within the model, the Gamma term produces the spin-wave gap, the Kitaev term controls the canting angle of the spin-rotation plane out of the layer, and the competition between ferromagnetic J1 and antiferromagnetic J3 sets the incommensurate period. The parameter set is obtained by iteratively constraining linear spin-wave","core_discovery":"The central claim is that the observed canted proper-screw magnetic ground state — a spiral whose rotation plane is tilted out of the layer — and the 3.1 meV gap in the spin-wave spectrum of NiI2 require substantial bond-dependent Kitaev and Gamma interactions. The optimal minimal model gives K = 3.33(0.032) meV, Gamma = 0.37(0.007) meV, J1 = -5.86(0.122) meV, J2 = 0.28(0.089) meV, J3 = 1.99(0.024) meV, and interlayer Jp2 = 0.78(0.005) meV. These parameters reproduce the measured dispersion, the gap, the 5.2 meV kink, the flat band top, and the magnetic structure's canting angle of 53.4 degrees versus the reported 55 degrees. Because Ni2+ is an S = 1 ion with quenched orbital moment, the pap","pith_inferences":["Inference: If the ligand-driven mechanism is the dominant source of bond anisotropy, isostructural nickel dihalides with lighter halogens (chloride, bromide) should show systematically weaker Kitaev terms, a testable trend across the series.","Inference: The extracted K/Gamma ratio of about 9 may fingerprint the microscopic coupling route; first-principles calculations that tune ligand spin-orbit coupling strength could refine or falsify the mechanism without new neutron data.","Inference: Because the paper explicitly excludes biquadratic and Gamma-prime terms from the minimal model, future resonant inelastic x-ray scattering or high-field torque experiments could reveal additional anisotropy that shifts the quoted values, so the fitted K is best read as a working estimate within the minimal model."],"forward_implications":["NiI2 becomes a concrete spin-1 triangular-lattice system with sizable K and Gamma, showing that ligand-only spin-orbit coupling can generate Kitaev-like bond anisotropy.","The minimal model quantitatively reproduces the canted proper-screw ground state and the 3.1 meV gap, including the 53.4-degree canting angle and the propagation direction along [1-10].","The flat band top and the 5.2 meV kink in the neutron data are assigned to J2 and J3 respectively, giving direct spectral handles for exchange parameters.","The mechanism broadens the search for Kitaev materials beyond Jeff = 1/2 ions with strong magnetic-ion spin-orbit coupling to high-spin systems with heavy ligands.","The discrepancy with a previously reported sub-0.3 meV gap is explained by different magnon branches and gap definitions, reconciling the two datasets."],"fun_headline_variants":["Ligand-driven Kitaev interaction confirmed in NiI2","Iodine ligands generate Kitaev bonds in NiI2","Spin-1 triangular NiI2 shows ligand Kitaev","NiI2 Kitaev gap from iodine spin-orbit","Ligand-mediated bond anisotropy in NiI2"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The whole parameter extraction rests on the assumption that the observed in-plane easy-axis anisotropy cannot come from single-ion anisotropy, which symmetry allows only along the z-axis, so the anisotropy must be dominated by the bond-dependent K and Gamma terms; a finite in-plane single-ion term, anisotropic g-factor, Dzyaloshinskii-Moriya interaction, or omitted Gamma-prime term at even a modest level would shift the fitted values.","fun_headline_variants_meta":{"raw":{"variants":["Ligand-driven Kitaev interaction confirmed in NiI2","Iodine ligands generate Kitaev bonds in NiI2","Spin-1 triangular NiI2 shows ligand Kitaev","NiI2 Kitaev gap from iodine spin-orbit","Ligand-mediated bond anisotropy in NiI2"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000354,"raw_usage":{"total_tokens":1798,"prompt_tokens":819,"completion_tokens":979,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":563,"completion_tokens_details":{"reasoning_tokens":897}},"tokens_in":563,"tokens_out":979,"duration_ms":8003,"temperature":1.0,"reasoning_tokens":897,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T00:46:43.393346+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Refit the same inelastic neutron scattering data with a model that includes in-plane single-ion anisotropy terms (or Gamma-prime) while allowing K and Gamma to be zero; if an equally good fit to the gap, the kink, and the canting angle emerges, the claim that the anisotropy is bond-dependent collapses. A polarized-neutron measurement that directly resolves the anisotropy contribution would settle the question.","supporting_citations":[],"review_version":1}