{"id":"45dfa36c-aa46-4617-8eca-95c2a30614a3","arxiv_id":"2501.03887","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"An expansion in the anisotropy ratio α=Jxy/Jzz gives the exact second-order magnon dispersions in the 1/3 plateau for any spin, shows slow 1/S convergence for S=1/2, and improves agreement with KCSO neutron data over linear spin-wave theory.","lead":"For a triangular-lattice quantum magnet, the authors compute magnon energies in the one-third magnetization plateau by expanding in the small ratio of transverse to Ising exchange, rather than in 1/spin. They find that the usual spin-wave expansion converges very slowly for spin-1/2 and that their result matches neutron-scattering data on K2Co(SeO3)2 better than linear spin-wave theory.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The experimental claim rests on the α^2 truncation being converged at α=0.08; the authors' own residual W+/W− mismatch admits higher-order terms could be significant, which would make the 'much better agreement' coincidental.","rationale":"The reader's weakest assumption—quantitative convergence of the α^2 truncation for KCSO—is exactly the load-bearing point for the experimental conclusion (ii). The paper's own bandwidth-ratio comparison (W+/W− = 0.77(3) theory vs 0.67(4) experiment) shows a residual discrepancy that the authors attribute to higher-order terms; this admission makes the truncation concern concrete rather than speculative. The theoretical result that Eq. (2) is exact at O(α^2) is well supported for S=1/2 (equivalence to Ref. [8] and a clean derivation skeleton in App. A), so the critique is not about internal correctness but about whether the second-order formula is the right diagnostic for a material with α=0.08 and momenta where |fk|≈3. A direct third-order calculation or a numerical DMRG check on a moderate cluster would settle the matter. The arbitrary-S formulas (Eqs. (6)-(9)) are also asserted without derivation, and an independent re-derivation for S=1 or S=3/2 would strengthen the slow-convergence claim, but the S=1/2 experimental comparison would survive even if those formulas had an error. Thus the single most load-bearing concern is the truncation convergence, and the reader's conditional verdict already captures it. I see no reason to move the verdict; the paper is honest about the residual mismatch and the theoretical parts are credible, but the strong 'much better agreement' statement should be treated as provisional until the higher-order check is done.","tokens_in":10702,"tokens_out":5457,"duration_ms":55199,"concrete_test":"Compute the O(α^3) correction to the S=1/2 magnon dispersions (Eqs. (2) and (3)) using degenerate perturbation theory in α, or perform a linked-cluster expansion to order α^4/α^6. Evaluate W+/W− and the Γ-K-M-Γ dispersion at α=0.08; if the O(α^3) term shifts W+/W− by more than ~0.05 or produces changes comparable to the LSWT difference, the experimental claim is not robust. Alternatively, extract the single-magnon branches from DMRG on a 36-site triangular uud cluster and compare with Eq. (2).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central experimental conclusion—that Eq. (2) computed with bare Jzz=3.1 meV, α=0.08 reproduces the KCSO magnon data much better than LSWT—requires the second-order expansion in α to be quantitatively converged at α=0.08. The true small parameter is not α alone but α|fk|/2, which reaches about 0.12 near the Γ and K points, and the α^2 corrections themselves are large enough to change the bandwidth ratio W+/W− by about 0.1 relative to first order. The authors explicitly attribute the residual mismatch between their predicted W+/W−=0.77(3) and the experimental 0.67(4) to 'higher-order terms in the α expansion' (Sec. 2). If O(α^3) corrections are comparable to the O(α^2) terms that distinguish Eq. (2) from LSWT, the agreement with experiment is not a robust test of the model. Nothing in the paper bounds the third-order contribution; the S=1/2 formulas are exact as perturbation theory, but their usefulness for KCSO is exactly the uncontrolled truncation. The slow-convergence claim (i) depends on the arbitrary-S formulas, which are also asserted without a displayed derivation, but the S=1/2 limit is supported by equivalence to Ref. [8].","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the triangular-lattice XXZ model in the up-up-down (1/3-plateau) phase. The authors develop a perturbation expansion in α = Jxy/Jzz, the ratio of transverse to longitudinal exchange, and compute the magnon dispersions through second order in α for arbitrary spin S. For S = 1/2 the result is exact at order α^2. They compare this with the linear spin-wave (LSWT) small-α expansion and show that LSWT misses the order-α^2 coefficients by a factor of about 3. They then compare the S = 1/2 dispersions, evaluated with the bare exchange parameters of KCSO (Jzz = 3.1 meV, α = 0.08, gz = 7.9), with published neutron data [22], finding better agreement than LSWT without renormalized parameters. They also analyze the convergence of the 1/S expansion, concluding that it converges very slowly for S = 1/2. The paper is a Letter with appendices giving the LSWT diagonalization and the single-particle hopping solution.","tokens_in":10972,"tokens_out":8031,"duration_ms":71674,"significance":"The arbitrary-S second-order dispersion formulas, if correct, are a useful benchmark for spin-wave calculations in Ising-like triangular magnets. The demonstration that the 1/S expansion converges slowly for S = 1/2 is an important caution for semiclassical methods. The comparison to KCSO is valuable and uses parameters determined independently from magnetization data, with no fitting to the neutron dispersions it claims to explain. However, the experimental conclusion is limited by the lack of an explicit derivation of the effective hopping amplitudes and by the absence of any estimate of higher-order terms in α.","major_comments":[{"comment":"The perturbative derivation of the effective hopping amplitudes t1u, t2u, t3u, Vu, Vd, and t1d is not shown. The text states 'by an explicit calculation' and cites equivalence to Ref. [8] only for S = 1/2. Since the arbitrary-S formulas are new and underpin both the slow-convergence claim (Fig. 3) and the experimental comparison, the authors should provide the derivation in an appendix or supplementary material, or at least present the key intermediate steps so the S-dependence can be checked.","section":"Main text near Eq. (2) and Eqs. (6)-(7)"},{"comment":"The experimental conclusion that Eq. (2) with bare parameters reproduces the KCSO data 'much better' than LSWT depends on the α^2 truncation. The residual bandwidth-ratio discrepancy W+/W− = 0.77(3) versus 0.67(4) is attributed to 'higher-order terms in the α expansion' without any estimate of their size. Near Γ and K, |fk| ≈ 3, so the effective expansion parameter α|fk|/2 ≈ 0.12, and the α^2 corrections themselves change W+/W− by about 0.24 relative to first order. Without a bound on α^3 contributions, the agreement could be partly coincidental. The authors should either estimate the next-order correction (for instance, by evaluating a third-order contribution to a key observable) or explicitly qualify claim (ii) as qualitative.","section":"Experimental comparison near Fig. 2 and W+/W- discussion"}],"minor_comments":[{"comment":"The arXiv identifier is given as '2402.077730' but the DOI indicates the correct identifier is '2402.07730'; please correct this typo.","section":"Reference list, Ref. [19]"},{"comment":"The statement that next-to-leading corrections appear 'multiplied with factors of order |fk|' is imprecise: the α^2 terms actually involve |fk|^2. The wording could be clarified to avoid implying a linear growth in |fk|.","section":"Text before Eq. (5)"},{"comment":"The phrase 'plotted in superposition to Fig. 4b of Ref. [22]' is understandable but would read more clearly as 'plotted superposed on Fig. 4b of Ref. [22].'","section":"Fig. 2 caption"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know about this paper if you care about the triangular XXZ model or the KCSO experiments: it gives exact-to-order-α² magnon dispersions for the uud plateau for arbitrary spin S, and shows that the 1/S expansion converges painfully slowly for S=1/2. The S=1/2 results are not new—they reproduce the hard-core boson results of Zhang et al. (Ref. [8])—but the arbitrary-S formulas and the slow-convergence analysis are.\n\nThe paper is honest and carefully done. The effective hopping problem for u and d excitations is clearly set up, and the final dispersions [Eqs. (8),(9)] are clean. The comparison to KCSO uses bare exchange parameters from a magnetization fit, not a fit to the neutron data, so the improved agreement with the measured magnon dispersion is a genuine prediction rather than a back-fit. The authors also flag the residual mismatch in the bandwidth ratio W+/W− = 0.77(3) versus 0.67(4) and do not pretend it is perfect.\n\nThe main soft spot is the derivation: the hopping amplitudes and energy shifts in Eq. (6) are introduced with \"by an explicit calculation\" and the details are not shown. For a Letter that is acceptable, but a serious referee should ask for the derivation or a clear appendix. Second, the experimental conclusion depends on the α² truncation being converged at α=0.08. Near Γ and K, α|fk|/2 ≈ 0.12, so higher-order terms could in principle shift the bandwidth ratio by an amount comparable to the experimental error. The authors acknowledge this possibility; it means the \"much better agreement\" should be read as strong but not definitive evidence for the pure nearest-neighbour XXZ model. The theory itself—the order-α² dispersion—does not depend on that truncation and stands on its own.\n\nOverall: solid theoretical contribution with a useful warning about spin-wave theory in frustrated magnets. It deserves peer review. I would recommend publishing after the derivation details are provided or referenced.","headline":"Exact-to-order-α² magnon dispersions for the uud plateau with arbitrary S; shows the 1/S expansion converges very slowly, and gives an improved but not airtight account of the KCSO neutron data.","tokens_in":11515,"tokens_out":2129,"would_cite":true,"duration_ms":20207,"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":"In the 1/3-plateau phase of the easy-axis triangular XXZ antiferromagnet, magnon dispersions are exact to second order in the anisotropy ratio $\\alpha=J_{xy}/J_{zz}$, and for $S=1/2$ these disagree sharply with spin-wave theory—resolving…","keywords":["triangular lattice","XXZ antiferromagnet","up-up-down phase","1/3 magnetization plateau","magnon dispersion","spin-wave theory","easy-axis anisotropy","K2Co(SeO3)2"],"falsifier":"Compute the third-order ($\\alpha^3$) contribution to the u-magnon dispersion at $S=1/2$ and check whether, at $\\alpha=0.08$ and near $\\Gamma$ or $K$ where $|f_k|\\approx 3$, it is small compared with the few-percent gap between Eq. (2) and the KCSO data; alternatively, measure the bandwidth ratio $W_+/W_-$ with higher precision, since Eq. (2) predicts $0.77(3)$ for $\\alpha=0.08(1)$, and a clear experimental value outside that range would force corrections beyond second order or longer-range exchanges.","tokens_in":10491,"feed_emoji":"🧲","tokens_out":11243,"duration_ms":88181,"temperature":0.7,"pith_summary":"The paper studies the excitation spectrum of the nearest-neighbour triangular XXZ model in the up-up-down (1/3-plateau) phase near the Ising limit. It performs a perturbation expansion in the anisotropy ratio $\\alpha = J_{xy}/J_{zz}$ rather than in $1/S$, and derives magnon dispersions that are exact at second order in $\\alpha$ for arbitrary spin $S$ (Eqs. (8) and (9)). For $S=1/2$, those exact formulas differ from linear spin-wave theory already in the order-$\\alpha^2$ coefficients, exposing a very slow convergence of the $1/S$ expansion. The paper then shows that the second-order formula, evaluated with the bare exchange parameters $J_{zz}=3.1$ meV, $\\alpha=0.08$, $g_z=7.9$, agrees with the KCSO neutron data much better than linear spin-wave theory, removing the need for a strongly renormalized $J_{zz}$.","feed_headline":"An α-expansion beats spin-wave theory for the KCSO 1/3 plateau","feed_subtitle":"It matches KCSO neutron data using bare exchange constants, no renormalized Jzz needed.","key_machinery":"The machinery is a perturbation expansion in $\\alpha=J_{xy}/J_{zz}$ around the Ising point, which maps each magnon species onto an effective single-particle hopping problem: u-excitations on the honeycomb lattice formed by the two up-sublattices (with first-, second-, and third-neighbour hoppings plus an energy shift) and d-excitations on the triangular C sublattice. The exact second-order hopping amplitudes of Eqs. (6) and (7) come from summing virtual spin-flip processes, and diagonalizing the resulting tight-binding Hamiltonians in momentum space yields the closed-form dispersions (8) and (9). The single ratio $\\gamma = t_{2u}/t_{1u} = t_{3u}/t_{1u} \\simeq -S\\alpha/(3S-1)$ controls the spectral shape at this order, and because the $S$-dependence enters through rational functions like $1/(3S-1)$, the slow convergence of the $1/S$ series is explicit.","core_discovery":"The central claim is that, on top of the up-up-down state, the magnon dispersions are exactly given at order $\\alpha^2$ by Eqs. (8) and (9) for arbitrary spin $S$, reducing to Eqs. (2) and (3) for $S=1/2$. The u-magnons obey a single-particle hopping problem on the honeycomb lattice with hoppings $t_{1u}=-S\\alpha(1-2S\\alpha/(6S-1))J_{zz}$ and $t_{2u}=t_{3u}=S^2\\alpha^2J_{zz}/[2(3S-1)]$, while the d-magnons hop on the triangular C sublattice with $t_{1d}=S^2\\alpha^2J_{zz}/(3S-1)$. Comparing with the small-$\\alpha$ expansion of linear spin-wave theory, the order-$\\alpha^2$ coefficients differ from the exact $S=1/2$ result—e.g., the bandwidth asymmetry is $W_+/W_-=1-3\\alpha$ from Eq. (2) versus $1-\\alpha$ from LSWT—so spin-wave theory misses the corrections by up to a factor of three. Since $|f_k|$ reaches about 3 near the $\\Gamma$ and $K$ points, the corrections are large even for $\\alpha=0.08$, and Eq. (2) reproduces the KCSO magnon bands without renormalizing $J_{zz}$.","pith_inferences":["If the slow convergence of the $1/S$ series seen here is generic for up-up-down phases with easy-axis anisotropy, then spin-wave-based interpretations of other frustrated triangular magnets—especially compounds with larger $\\alpha$—may need similar re-examination.","The same $\\alpha$-expansion machinery could be applied to other gapped collinear phases of the XXZ model, such as the zero-field supersolid or the high-field phases, to test whether their linear-spin-wave predictions fail in the same way.","The single-parameter $\\gamma$ scaling of the spectral shape suggests a collapse test: plot the normalized u-magnon dispersion for different compounds or fields; data that fail to collapse would indicate physics beyond the nearest-neighbour XXZ model."],"forward_implications":["Spin-wave fits to the 1/3-plateau spectrum of an easy-axis S=1/2 triangular antiferromagnet are quantitatively unreliable, because LSWT underestimates the order-$\\alpha^2$ corrections by up to a factor of three and therefore biases any exchange constants extracted from such fits.","For KCSO, Eq. (2) evaluated at the bare parameters $J_{zz}=3.1$ meV, $\\alpha=0.08$, $g_z=7.9$ matches the measured magnon bands, indicating that the strongly renormalized $J_{zz}\\approx0.68$ meV required by the LSWT fit is an artifact of the spin-wave approximation rather than a physical coupling.","The underestimated bandwidth asymmetry in LSWT would be misattributed to longer-range exchanges; with the correct second-order formula, only a small ferromagnetic $J_{2xy}\\approx -0.015(10)J_{xy}$, or higher-order terms in $\\alpha$, is needed to explain the residual discrepancy.","The general-$S$ dispersions (8) and (9) interpolate between exact $S=1/2$ and the $S\\to\\infty$ spin-wave limit, providing a quantitative benchmark for how many orders of the $1/S$ expansion are needed before its error becomes small.","The agreement supports the nearest-neighbour XXZ model with the measured bare couplings as the minimal description of KCSO's plateau-phase excitations, while giving a controlled way to bound possible small longer-range couplings."],"supporting_citations":[{"why":"Supplies the KCSO neutron-scattering data and the bare parameters Jzz = 3.1 meV, alpha = 0.08, gz = 7.9 against which Eq. (2) is compared.","marker":"[22]"},{"why":"Contains the LSWT fit whose strongly renormalized Jzz = 0.68 meV and bandwidth ratio W+/W- = 0.67(4) define the discrepancy the paper resolves.","marker":"[25]"},{"why":"Provides independent KCSO measurements confirming the phase diagram and the d-magnon branch at 7 T, with LSWT-derived exchange constants compared to the bare values.","marker":"[23]"},{"why":"Gives an equivalent hard-core boson calculation whose effective parameters match the S = 1/2 hoppings, validating the perturbation theory.","marker":"[8]"},{"why":"Shows that corrections beyond LSWT are strong in the uud phase of Ba3CoSb2O9, motivating the need for an expansion beyond 1/S.","marker":"[24]"},{"why":"Reports corrections to LSWT in the uud phase of Na2BaCo(PO4)2, cited as further evidence that spin-wave corrections matter in plateau phases.","marker":"[19]"}],"fun_headline_variants":["α-expansion fixes magnon dispersion in KCSO 1/3 plateau","New theory beats spin-wave for K2Co(SeO3)2 plateau","Exact α² magnons match KCSO data without renormalization","Spin-wave fails, α-expansion wins for KCSO plateau","Magnon dispersion solved: α-expansion beats 1/S for KCSO"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the second-order expansion in $\\alpha$ is quantitatively converged for KCSO, where $\\alpha=0.08(1)$, despite $|f_k|$ reaching about 3 near the $\\Gamma$ and $K$ points; if the neglected $\\alpha^3$ and higher terms are large there, the close agreement with the neutron data could be coincidental.","fun_headline_variants_meta":{"raw":{"variants":["α-expansion fixes magnon dispersion in KCSO 1/3 plateau","New theory beats spin-wave for K2Co(SeO3)2 plateau","Exact α² magnons match KCSO data without renormalization","Spin-wave fails, α-expansion wins for KCSO plateau","Magnon dispersion solved: α-expansion beats 1/S for KCSO"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000589,"raw_usage":{"total_tokens":2817,"prompt_tokens":1047,"completion_tokens":1770,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":663,"completion_tokens_details":{"reasoning_tokens":1668}},"tokens_in":663,"tokens_out":1770,"duration_ms":11086,"temperature":1.0,"reasoning_tokens":1668,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T21:44:54.207318+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the third-order ($\\alpha^3$) contribution to the u-magnon dispersion at $S=1/2$ and check whether, at $\\alpha=0.08$ and near $\\Gamma$ or $K$ where $|f_k|\\approx 3$, it is small compared with the few-percent gap between Eq. (2) and the KCSO data; alternatively, measure the bandwidth ratio $W_+/W_-$ with higher precision, since Eq. (2) predicts $0.77(3)$ for $\\alpha=0.08(1)$, and a clear experimental value outside that range would force corrections beyond second order or longer-range exchanges.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Contains the LSWT fit whose strongly renormalized Jzz = 0.68 meV and bandwidth ratio W+/W- = 0.67(4) define the discrepancy the paper resolves."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides independent KCSO measurements confirming the phase diagram and the d-magnon branch at 7 T, with LSWT-derived exchange constants compared to the bare values."},{"cited_title":"Zhang, R","cited_arxiv_id":null,"evidence_quote":"Gives an equivalent hard-core boson calculation whose effective parameters match the S = 1/2 hoppings, validating the perturbation theory."},{"cited_title":"Kamiya, L","cited_arxiv_id":null,"evidence_quote":"Shows that corrections beyond LSWT are strong in the uud phase of Ba3CoSb2O9, motivating the need for an expansion beyond 1/S."}],"review_version":1}