{"id":"5c33b756-36e6-40d7-aa4c-bafce3fb2ca5","arxiv_id":"1908.11617","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"The high-field phase of α-RuCl3 is a partially polarized quantum disordered state with a single spin-flip mode, a two-particle bound state, and a gapped continuum.","lead":"Raman and THz measurements of α-RuCl3 in magnetic fields up to 33 T show that the high-field phase hosts a gapped continuum, a two-particle bound state, and a sharp single-particle excitation. The results identify the high-field phase as a partially polarized quantum disordered state and resolve a controversy about steep excitation slopes.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Concern: the asymptotic slope g* = 2.51 ± 0.18 is extracted from a fit form shown only in the Supplemental, over a field range where the slope is still decreasing; if this extrapolation is not robust, the identification of m1α as the single-particle ΔS=1 mode and the QDS conclusion lose support.","rationale":"The paper is a high-quality experimental study, and the mode assignments are plausible given the agreement between Raman and THz data. However, the QDS conclusion is explicitly tied to the infinite-field limit: the text states that the asymptotic slope 'confirms m1α as the mode that evolves into the |ΔS|=1 spin-flip excitation,' and the abstract and conclusion generalize to 'firmly establishes the partially-polarized quantum disordered character.' The reader correctly identified that this asymptotic slope is the load-bearing premise. I agree. The main text provides no functional form or goodness-of-fit details for the extrapolation, and the data only reach 33 T while the slope is still far from the asymptote. Additionally, the ED calculation used to support the single-particle interpretation uses a model whose parameters were adjusted to reproduce the same g*(B), making it a weaker independent check. This does not invalidate the paper; it warrants a CONDITIONAL verdict pending an independent, transparent extrapolation and preferably higher-field data (e.g., pulsed fields above 40 T) or an independent gab measurement to anchor the asymptotic slope. The reader's verdict should remain CONDITIONAL.","tokens_in":12120,"tokens_out":7033,"duration_ms":64924,"concrete_test":"Retrieve the numerical m1α peak energies vs field from the Raman and THz data (Fig. 2 and Fig. 3, and the underlying tables if available). Refit the combined data without the Supplemental [39] ansatz, e.g., using E(B) = a + bB + c/B + d/B^2 for B ≥ 10 T, and separately using only B ≥ 20 T, with the Raman and THz data sets fit independently. If the extrapolated b differs from 2.51 ± 0.18 by more than 0.3 or falls outside 2–2.8, the asymptotic g* claim is not robust. Then publish the fitted intercepts and goodness-of-fit for each ansatz as a supplement, so the sensitivity of the extrapolation can be assessed directly.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central conclusion that the high-field phase is a partially polarized quantum disordered state rests on assigning m1α as the |ΔS|=1 single-particle spin-flip excitation. The key evidence is the asymptotic slope g*|B→∞ = 2.51 ± 0.18, obtained by fitting the m1α peak positions over the full field range (up to 30–33 T) with a functional form referenced only to the Supplemental [39]. The main text does not display the fit function, the number of free parameters, or the data range used. This matters because at 30 T the measured slope is still ~3, substantially above the claimed asymptote, so the extrapolation must cover a large remaining field decrease with no data beyond 33 T to anchor it. The fit is further expected to incorporate level repulsion near Bc (slope ~8 at 10 T), and the 'adjusted model' in Fig. 3(b) was selected specifically to reproduce the measured g*(B), so the ED demonstration of single-particle character is not independent of the claim. If a different extrapolation form (or restricting the fit to B > 20 T) shifts the intercept outside the physical range 2 ≤ gab ≤ 2.8, or if the fit only converges through the Supplemental's specific ansatz, the evidence for the |ΔS|=1 assignment falls away, and the resolution of the steep-slope controversy (QSL vs bound-state vs large g) would remain unresolved.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports Raman and THz spectroscopic measurements of α-RuCl3 in in-plane magnetic fields up to 33 T. The authors observe that the low-energy magnon modes in the zero-field zig-zag ordered phase are suppressed at a field-induced transition around Bc = 7.5 T, where the Raman continuum intensity is maximized. In the high-field phase they identify a sharp single-particle mode m1α, a satellite m1β, a two-particle bound state m2γ, and a gapped multi-particle continuum with a broad maximum m2α ≈ 2m1α. They assign m1α to the |ΔS| = 1 spin-flip excitation of a partially polarized quantum disordered state (QDS) based on an asymptotic slope g* = 2.51 ± 0.18 extracted from a fit over the full field range. Exact diagonalization of a C3-broken Kitaev-Heisenberg-Γ model reproduces the main spectral features and shows that Kitaev and off-diagonal terms in the Fleury-Loudon operator are required to observe the single-particle mode in Raman scattering. The paper concludes that the high-field phase is neither a quantum spin liquid nor a fully polarized state but a partially polarized QDS.","tokens_in":12491,"tokens_out":5560,"duration_ms":47984,"significance":"If the interpretation holds, the paper provides a resolution of the controversy over the steep intermediate-field slopes (g* ≈ 8) by attributing them to level repulsion and anisotropic-coupling effects rather than to fractionalized excitations. The combined Raman and THz dataset up to 30–33 T, including the observation of a two-particle bound state and a gapped continuum, is a substantial experimental contribution. The ED calculations underline the importance of Kitaev and off-diagonal exchange in the Raman response. However, the central mode assignment rests on an asymptotic-slope extraction whose details are not given in the main text, and the supporting ED model is adjusted to reproduce the measured slope; these issues currently prevent the paper from 'firmly establishing' the QDS interpretation. The paper is nevertheless a strong candidate for publication after the analysis is made more transparent and the claims appropriately qualified.","major_comments":[{"comment":"The asymptotic slope g*|B→∞ = 2.51 ± 0.18 is obtained from a fit whose functional form is described only in the Supplemental Material [39]. The main text does not report the fit function, the number of free parameters, the field range included in the fit, or any measure of fit quality. This is a load-bearing result: it is the basis for identifying m1α as the |ΔS| = 1 single-particle spin-flip mode, which in turn anchors the QDS conclusion. Because the slope at 30 T (~3) is still substantially above the claimed asymptotic value, the extrapolation to infinite field is nontrivial. Please present the fit and demonstrate its robustness, for example by varying the fitting window (e.g., using only B > 20 T) and by testing alternative functional forms, and discuss how the uncertainty in the THz field orientation affects the extracted g*.","section":"p.4, paragraph beginning 'An important quantity of interest...'"},{"comment":"The exact diagonalization model parameters (J1, Kx/y, Kz, Γ1, J3) = (−0.5, −7.5, −5, 2.5, 0.5)·1.5 meV are explicitly described as 'adjusted' so that the model captures the strong field dependence of g*. This makes the numerical support for the single-particle assignment partially circular: the model has been tuned to reproduce the very quantity it is used to interpret. The experimental observation of the sharp mode and continuum is independent, but the claim that the ED calculations 'demonstrate' the single-particle character is overstated. The authors should clarify which spectral features are generic to the parameter region derived from ab initio studies and which depend specifically on the fit to g*(B).","section":"Fig. 3(b) and the text describing the 'adjusted model'"},{"comment":"The claim that the study 'firmly establishes' the partially-polarized quantum disordered character of the high-field phase is stronger than the evidence supports. The mode assignments are plausible, and the spectroscopic observations are consistent with a QDS, but the identification of m1α as the |ΔS|=1 mode relies on the extrapolation discussed above, and no direct measurement of spin correlations or static magnetization is presented. I recommend either providing the missing fit details and robustness checks or softening the conclusion to state that the results strongly support, rather than firmly establish, the QDS interpretation.","section":"Conclusions, first paragraph"}],"minor_comments":[{"comment":"The 0.6 meV offset between the Raman and THz peak positions of m1α is attributed to in-plane anisotropy, but the in-plane field orientation was not determined for the THz measurements. Since the THz data are used in the g* extraction, this systematic offset should be included in the uncertainty budget or analyzed as a separate systematic.","section":"Fig. 2(a)-(b)"},{"comment":"The fit used to extract g* is referenced only as [39]. Please include the fit expression and parameter values in the main text or in a table in the Supplemental so that readers can assess the extrapolation without consulting an external reference.","section":"Supplemental Material [39]"},{"comment":"The laser power of 100 µW used for high-field Raman measurements causes sample heating that suppresses the low-energy magnon modes and weakens m1α at 8 T; this power dependence should be acknowledged as a limitation when comparing intensities across field regimes.","section":"Experimental setup, Fig. 2(a)"},{"comment":"The statement 'No clear evidence is found for an intermediate phase around 7.5 T' is based on the field dependence of the Raman and THz features; a narrow intermediate phase could be missed by these measurements. A more cautious phrasing would be 'no evidence for an intermediate phase was resolved in our measurements.'","section":"Conclusion, sentence about intermediate phase"}],"recommendation":"major_revision","confidential_remarks":"The paper is a strong experimental contribution, and the main concern is the transparency and robustness of the g* extrapolation, which is load-bearing. If the Supplementary material already contains a complete description of the fit, the revision could be minor; however, as the manuscript currently stands, the central claim is not fully supported in the main text. I also note that the ED model adjustment is presented honestly, but the word 'firmly' in the abstract and conclusion should be softened unless the analysis is strengthened."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThis one is worth a close look. The genuinely new piece is the high-field Raman study of α-RuCl3 up to 33 T inside the field-induced phase—previous Raman work stopped around zero or low field. The authors resolve a sharp lower mode m1α, a satellite m1β, a two-particle bound state m2γ, and a gapped multi-particle continuum, tracking these in both Raman and THz. That is a solid spectroscopic fingerprint of the high-field state. The observation of a single-particle spin-flip in Raman is unusual, and the proposed mechanism—Kitaev and off-diagonal terms in the Fleury-Loudon operator—is plausible and backed by ED on a 24-site cluster. The experimental care is also evident: 10 µW low-power runs, two crystals, two probes, consistent peak positions.\n\nThe soft spot is the asymptotic slope g* = 2.51 ± 0.18. The fit form is only in the supplemental; the main text gives no equation, parameter count, or robustness check. At 30 T the local slope is still about 3, so the extrapolation has to cover a long remaining decrease with no data beyond 33 T. That said, this is a transparency problem, not a fatal one. 2.5 sits inside the physical gab range 2–2.8 and comfortably excludes the gab ≈ 10 alternative. But 'firmly establishes' is too strong when the reader cannot see the fit or judge its stability. A referee should ask for the fitting function, the fitted field range, and a test of alternative functional forms.\n\nRelatedly, the ED model in Fig. 3(b) is explicitly adjusted to reproduce the measured g*(B). That makes the numerical support partly a fit, so the theory is illustrative rather than an independent confirmation. The core experimental observations stand on their own: the gapped continuum, the bound state, the single sharp mode, and the overall field evolution are all consistent with a partially polarized quantum disordered state. The steep-slope controversy is real, and this paper is a credible resolution if the slope analysis is robust.\n\nThe citation pattern is fine; self-citations are to the earlier work being extended. I would send this to peer review. The authors can fix the transparency in revision, and the data will be worth having in the literature.\n\nBest","headline":"New high-field Raman/THz data give a credible QDS picture for α-RuCl3, but the 'firmly establishes' relies on an under-displayed slope extrapolation and an adjusted ED model.","tokens_in":13073,"tokens_out":2455,"would_cite":true,"duration_ms":25823,"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":"High-field spectroscopy identifies α-RuCl3's field-induced phase as a partially polarized quantum disordered state, not a spin liquid.","keywords":["α-RuCl3","Kitaev spin liquid","quantum disordered state","Raman spectroscopy","terahertz spectroscopy","spin-flip excitation","two-particle bound state","exact diagonalization"],"falsifier":"If Raman or THz data above 33 T reveal m1α splitting or a slope discontinuity, or if inelastic neutron scattering shows that m1α carries a different spectral weight than a single spin flip, the central identification fails.","tokens_in":11991,"feed_emoji":"🧲","tokens_out":7103,"duration_ms":59274,"temperature":0.7,"pith_summary":"This paper sets out to settle what happens in the layered magnetic material α-RuCl3 when a magnetic field above about 7.5 tesla suppresses its antiferromagnetic order. Using Raman and terahertz spectroscopy up to 33 tesla, it resolves three distinct magnetic excitations: a sharp single-particle spin-flip mode, a two-particle bound state, and a gapped multi-particle continuum. The authors argue this spectrum identifies the high-field phase as a partially polarized quantum disordered state, neither the long-sought Kitaev spin liquid nor the fully polarized magnet. They also use the single-particle mode's energy-versus-field slope to resolve a controversy about anomalously steep slopes reported near the critical field.","feed_headline":"High-field α-RuCl3 is neither spin liquid nor fully polarized","feed_subtitle":"Raman and THz spectra up to 33 T reveal a partially polarized quantum disordered state and a sharp single-particle mode.","key_machinery":"The central object is the Fleury-Loudon scattering operator $F \\propto \\sum_{ij} \\mathbf{S}_i \\cdot \\hat{J}_{ij} \\cdot \\mathbf{S}_j (\\boldsymbol{\\delta}_{ij} \\cdot \\mathbf{E}_{\\mathrm{in}})(\\boldsymbol{\\delta}_{ij} \\cdot \\mathbf{E}_{\\mathrm{out}}^*)$, which generates Raman intensity from exchange-mediated light scattering. In a conventional magnet this operator creates two-magnon excitations; the paper shows that in a Kitaev magnet with bond-dependent and off-diagonal exchange, terms with $S_i^{\\mu} S_j^{\\nu}$, $\\mu \\neq \\nu$, create single spin-flip ($|\\Delta S| = 1$) excitations, making the single-particle mode m1α Raman-active. Exact diagonalization on a 24-site cluster with an adjusted model $(J_1, K_1^{x/y}, K_1^{z}, \\Gamma_1, J_3) = (-0.5, -7.5, -5, 2.5, 0.5) \\times 1.5$ meV, with broken $C_3$ symmetry, reproduces the measured field dependence and identifies the Kitaev and off-diagonal terms as responsible for the observed single-particle peak and continuum.","core_discovery":"The paper's central claim is that the field-induced phase above Bc = 7.5 T in α-RuCl3 is a quantum disordered state with partial field alignment of the spin-orbital moments, not a quantum spin liquid and not a fully field-polarized state. The evidence is spectroscopic: a well-defined single-particle excitation m1α whose high-field slope g* approaches 2.51 ± 0.18, a two-particle bound state m2γ slightly below a gapped multi-particle continuum, and the absence of any additional phase transition up to 33 T. The steep intermediate-field slopes reported earlier are attributed to level repulsion between the single-particle mode and the continuum, so they do not indicate fractionalization or an enormous g-factor. Exact diagonalization of a realistic C3-broken Kitaev model reproduces the field-dependent Raman response, including the single-particle mode, only when Kitaev and off-diagonal exchange terms enter the Fleury-Loudon scattering operator.","pith_inferences":["If the quantum-disordered-state assignment is right, the same single-particle mode should appear in other spectroscopic probes, such as inelastic neutron scattering under field, and its dispersion would directly test the adjusted model's parameters.","The asymptotic g* ≈ 2.5 constrains the in-plane g-factor; a full in-plane angular dependence of m1α would discriminate among the C3-broken parameter sets.","The observed two-particle bound state slightly below the continuum invites the question of whether it condenses or hybridizes at even higher fields; extending measurements to pulsed fields beyond 33 T could test this.","The strategy of fitting the full field range to extract an asymptotic slope, rather than focusing near the critical field, could be applied to other Kitaev candidates where steep intermediate slopes have been interpreted as evidence for fractionalization."],"forward_implications":["The high-field phase of α-RuCl3 above 7.5 T is a partially polarized quantum disordered state; no sign of a separate intermediate phase appears up to 33 T.","The asymptotic slope g* = 2.51 ± 0.18 confirms m1α as the mode that grows into the infinite-field |ΔS| = 1 spin flip, settling the steep-slope debate.","Kitaev and off-diagonal exchange terms in the Fleury-Loudon operator are essential: conventional two-magnon Raman would not show the sharp single-particle mode.","One expects similar high-field behavior in other Kitaev candidate materials, notably the iridates."],"supporting_citations":[{"why":"Reported the intermediate-field THz mode and its steep slope, which this paper reinterprets through a full-field analysis.","marker":"[12]"},{"why":"Theoretical prediction of the high-field quantum disordered state and bound states that this work confirms spectroscopically.","marker":"[20]"},{"why":"An ESR study reporting a steep mode slope, part of the controversy resolved by the asymptotic g* analysis.","marker":"[22]"},{"why":"Neutron scattering showing out-of-plane dispersion of the single-particle mode, used to explain the m1β satellite.","marker":"[29]"},{"why":"Established the Raman response of a Kitaev spin liquid as a continuum, the background against which the high-field features are interpreted.","marker":"[30]"},{"why":"The Fleury-Loudon scattering formalism used to compute and interpret the Raman spectra.","marker":"[38]"},{"why":"Ab-initio parameters for the C2/m structure that motivate the C3-broken model used in exact diagonalization.","marker":"[45]"},{"why":"A realistic (J,K,Γ,J3) model whose Raman response the exact diagonalization captures.","marker":"[49]"}],"fun_headline_variants":["α-RuCl3 high-field phase: quantum disordered, not a spin liquid","Partially polarized quantum disorder in α-RuCl3, not a spin liquid","High-field α-RuCl3: single sharp mode, bound state, not spin liquid","Quantum disordered, not spin liquid: α-RuCl3's high-field phase","α-RuCl3 above 7.5 T: quantum disorder, not full polarization"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the sharp mode m1α is truly a single spin-flip excitation whose high-field slope flattens to the reported 2.51 ± 0.18, because the entire resolution of the slope controversy and the quantum-disordered-state assignment rests on that identification and extrapolation.","fun_headline_variants_meta":{"raw":{"variants":["α-RuCl3 high-field phase: quantum disordered, not a spin liquid","Partially polarized quantum disorder in α-RuCl3, not a spin liquid","High-field α-RuCl3: single sharp mode, bound state, not spin liquid","Quantum disordered, not spin liquid: α-RuCl3's high-field phase","α-RuCl3 above 7.5 T: quantum disorder, not full polarization"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000989,"raw_usage":{"total_tokens":4179,"prompt_tokens":918,"completion_tokens":3261,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":534,"completion_tokens_details":{"reasoning_tokens":3151}},"tokens_in":534,"tokens_out":3261,"duration_ms":22717,"temperature":1.0,"reasoning_tokens":3151,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:09:21.020115+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"If Raman or THz data above 33 T reveal m1α splitting or a slope discontinuity, or if inelastic neutron scattering shows that m1α carries a different spectral weight than a single spin flip, the central identification fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"A realistic (J,K,Γ,J3) model whose Raman response the exact diagonalization captures."},{"cited_title":"Baek, S.-H","cited_arxiv_id":null,"evidence_quote":"Theoretical prediction of the high-field quantum disordered state and bound states that this work confirms spectroscopically."},{"cited_title":"Janˇ sa, A","cited_arxiv_id":null,"evidence_quote":"An ESR study reporting a steep mode slope, part of the controversy resolved by the asymptotic g* analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The Fleury-Loudon scattering formalism used to compute and interpret the Raman spectra."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Ab-initio parameters for the C2/m structure that motivate the C3-broken model used in exact diagonalization."}],"review_version":1}