{"id":"ab0d0748-6e39-4dc7-9d32-34b257f46c35","arxiv_id":"1908.00724","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In BPCB, high-resolution neutron scattering resolves a 40 to 50 micro-electronvolt splitting of the triplet band and attributes it to weakly anisotropic leg exchange while ruling out rung anisotropy.","lead":"High-resolution neutron scattering reveals that the tiny magnetic anisotropy in the spin-ladder compound BPCB comes from the weak ladder-leg bonds, not the much stronger rung bonds. This gives an unusually direct experimental check on how anisotropic superexchange behaves in a model quantum magnet.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Rung-Ising contribution is never fit in a combined model; the conclusion that rung Ising anisotropy is negligible is not directly established.","rationale":"The reader's weakest assumption concerns the convergence of the third-order strong-coupling expansion. I agree that is a caveat, but I do not think it is the most load-bearing issue: the rung-only vs leg-only distinction is qualitative and stable between second and third order, and the authors explicitly disclaim reliance on the small Δmax differences between leg models. The deeper gap is logical: the paper claims rung Ising anisotropy is 'negligible' without ever testing a Hamiltonian that contains both rung and leg anisotropy. The rung-only model being excluded only proves that rung Ising cannot be the sole source; it does not bound its magnitude when leg terms are present. Since a rung term produces an almost constant-in-k contribution and the leg terms produce a k-varying one, the observed ratio 50/40 does not by itself rule out a substantial constant rung offset. A two-parameter shape can easily accommodate a constant offset by rescaling the leg anisotropy; depending on the offset, 'leg dominant' may still hold but 'rung negligible' is a separate, unsupported claim. The remedy is a combined fit with C free, which is a standard and inexpensive check. Until then, ACCEPT is slightly too strong; CONDITIONAL on this fit (or on softening the conclusion) is the appropriate verdict.","tokens_in":9778,"tokens_out":24976,"duration_ms":256268,"concrete_test":"Fit the extracted dispersions of Fig. 5 to the combined Hamiltonian with C (rung Ising), A and B (leg Ising/DM) all free, using the same strong-coupling expressions (or a DMRG cross-check), and compute the 95% confidence interval on the rung contribution c = (λ C/2)J⊥ to the band-minimum splitting. If c is constrained to <5 μeV, the 'negligible rung' claim holds; if c can exceed ~10 μeV while still fitting Δmin/Δmax = 50/40 μeV, the central claim must be weakened to 'leg terms are a necessary source of the k-dependent part,' not the unique dominant source.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's central conclusion—that the anisotropy is dominated by weakly anisotropic leg interactions and that rung Ising anisotropy is negligible—is reached by comparing three leg-only models (cases b, c, d) against the rung-only model (case a). Case a is excluded because its triplet splitting is nearly k-independent, contrary to the measured Δmin=50 μeV vs Δmax=40 μeV. But no model containing both rung Ising (C in Eq. 3) and leg anisotropy (A, B in Eq. 6) is ever fitted. Because a rung term contributes an approximately k-independent splitting c while leg terms contribute a k-dependent splitting, a combined model can reproduce the observed 50/40 μeV pattern with c up to several tens of μeV, as long as the leg parameters are rescaled: the data ratio does not by itself force c to be small. The statement in Sec. V that rung Ising is 'negligible' therefore goes beyond the fitted model space; it assumes the leg-only parametrization is complete. Consequently, the abstract's 'dominant source' claim is only established if a fit with C free actually bounds the rung contribution to a small fraction of the ~0.6 K splitting, which is not shown.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports high-resolution inelastic neutron scattering on the S=1/2 spin ladder compound BPCB, resolving a splitting of the triplet band of 50(1) μeV at the band minimum and 40(2) μeV at the band maximum. The authors derive a strong-coupling expansion for the triplet dispersion of a spin ladder with anisotropic rung or leg exchange and compare four models: rung Ising, leg Ising, leg DM, and both DM plus Ising on the legs. The rung-Ising model is excluded because it yields a nearly k-independent splitting, whereas the data show a clear k-dependence. All three leg-anisotropy models reproduce the data, and the authors conclude that weakly anisotropic leg interactions dominate the magnetic anisotropy in BPCB, with the rung Ising anisotropy 'negligible'. This conclusion is used to support the theory of anisotropic superexchange.","tokens_in":9995,"tokens_out":24232,"duration_ms":223352,"significance":"The experimental achievement is significant: the 50/40 μeV splittings are well resolved given the reported 19/36 μeV resolution, and the Gaussian-fit analysis is standard. The strong-coupling calculation for anisotropic ladders is a useful contribution that will be of interest to the quantum magnetism community. The paper is honest about the degeneracy of the three leg-anisotropy models and about the residual perturbation-order dependence (footnote 31). However, the central interpretive claim that rung Ising anisotropy is negligible is not established by the presented fits, because no combined model is fitted. As a result, the paper's strongest conclusion goes beyond its model space. With a combined fit bounding the rung contribution, or with appropriately softened claims, the paper would be a valuable case study in identifying exchange anisotropy in quantum magnets.","major_comments":[{"comment":"The conclusion that rung Ising anisotropy is 'negligible' is not supported by the fitted model space. Only the rung-only model (case a) and leg-only models (cases b, c, d) are compared; no fit with both the rung Ising parameter C in Eq. (3) and the leg parameters A, B in Eq. (6) is reported. From Eqs. (4)-(5), a rung Ising term contributes a splitting that is k-independent to first order and has only a weak k-dependence at third order, while leg terms produce a strongly k-dependent splitting. A combined model can reproduce the observed 50/40 μeV pattern with the rung term providing most of the splitting: for example, in the notation of Sec. III, values around C≈0.33 and A≈0.02 give a rung contribution of roughly 50 μeV and a leg contribution that accounts for the ~10 μeV difference between the two extrema. The data alone therefore do not bound C. The statement that rung Ising anisotropy is 'negligible' in Sec. IV.D and V, and the superexchange-support argument built on it, assumes the leg-only parametrization is complete. I recommend fitting a combined model with C free and reporting a bound on C, or explicitly restricting the conclusion to 'leg anisotropy is required to explain the k-dependence of the splitting' and removing the 'negligible' wording.","section":"Sections IV.C, IV.D, V and the abstract"},{"comment":"The quantitative comparison to the leg-anisotropy models is affected by the truncation of the strong-coupling expansion. The footnote states that the predicted Δmax values (37, 32, 35 μeV for cases b, c, d) differ from the measured 40(2) μeV, and that the dispersion still changes from second to third order. Since the 'excellent agreement' of the leg models is part of the evidence for the central conclusion, the authors should either demonstrate that the second-to-third-order shift does not affect the discrimination between rung and leg models, or present the comparison with a clear statement of the theoretical uncertainty. The qualitative exclusion of the rung-only model is likely robust because its splitting is k-independent already at first order, but the quantitative agreement claim for the leg models should be calibrated against the perturbative uncertainty.","section":"Footnote 31 and Sec. IV.C"}],"minor_comments":[{"comment":"The third-order coefficient for cos(3k) is printed as '1/8 cos(3k)', but in the limit A=B=0 Eq. (9) should reduce to Eq. (2), which contains 'cos(3k)' with coefficient 1. The printed '1/8' is inconsistent with the isotropic limit and is presumably a typo; it should be corrected because the dispersion expression is central to the model comparison.","section":"Equation (9)"},{"comment":"The text states that all three leg-anisotropy models give 'excellent agreement' but shows only the fit for case d. Showing the fits or residual plots for cases b and c would substantiate this claim.","section":"Sec. IV.C, Fig. 5"},{"comment":"The statement in the appendix that 'all H′ commute with Sz' is not immediately obvious for the DM term HLeg,DM in Eq. (7); a one-sentence demonstration that e_z·(S×S) preserves total Sz would help the reader.","section":"Appendix and Sec. III.C"},{"comment":"For case d, the anisotropy is specified by two parameters A and B, whereas the caption refers to 'the anisotropy parameter' singular. The caption should state how A and B are related via Eq. (11) when plotting case d.","section":"Figure 4 caption"},{"comment":"The sentence 'inter-ladder interactions were previously estimated on the order of a few 10s of mK 18 (a few μeV)' would be clearer as 'a few tens of mK'.","section":"Sec. IV.A"}],"recommendation":"major_revision","confidential_remarks":"The experimental data and resolution analysis are convincing, and the paper fits the journal's scope. The main issue is the combined-model degeneracy: the conclusion that rung Ising anisotropy is negligible is not proven by the reported fits, and the superexchange-supporting argument in Sec. IV.D inherits that gap. If the authors can add a combined fit with C free (or a bound), the paper would likely be acceptable. Alternatively, softening the 'negligible' and 'dominant' claims to 'leg anisotropy is required for the k-dependence' would make the paper publishable as a measurement and model study, though with reduced interpretive strength."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead this one. The experiment is the real deal: they resolve a 50/40 micro-eV triplet splitting in BPCB, something no one had seen before, and the data are deposited. The strong-coupling formulas for the anisotropic ladder are useful extensions of Reigrotzki et al. and worth having on record. The paper also makes a nice qualitative case: a pure rung-Ising term gives an almost k-independent splitting, so it can't produce the observed difference between band minimum and maximum. That exclusion is solid.\n\nThe soft spot is the stronger claim. The authors conclude that rung Ising anisotropy is \"negligible\" and that leg interactions are the \"dominant\" source. What they actually fit are four separate models: three leg-only and one rung-only. No combined model with both rung C and leg A/B is ever fitted. Since a rung term adds a roughly k-independent constant to the splitting while a leg term contributes a k-dependent piece, a combined model could reproduce 50/40 with C as large as tens of micro-eV, just by rescaling the leg parameters. The data alone do not bound C to a small fraction of the ~0.6 K splitting. The theoretical symmetry argument that rung Ising should vanish is plausible, but the paper presents it as supported by the data, and the data don't actually constrain it. That should be fixed, either by a two-parameter fit or by softening the abstract and conclusion.\n\nThe perturbation expansion convergence is honestly flagged in footnote 31, so I don't hold the 32–37 vs 40 micro-eV discrepancy against them. The three leg models being indistinguishable is also acknowledged up front. These are minor.\n\nWhere does that leave the paper? The experimental observation is new and the qualitative exclusion of pure rung Ising is robust. The \"dominant source\" framing overstates what the fits establish. I'd send it to a good referee: the analysis is careful, the limitations are mostly acknowledged, and the one missing fit is easy to request. With that addition or a toned-down claim, it's a solid paper. For a condensed-matter audience it's worth reading.\n\nRecommendation: peer review, with the combined-model issue raised.","headline":"A genuinely new high-resolution neutron result with a clear qualitative message, but the claim that rung Ising anisotropy is negligible outruns the models actually fitted.","tokens_in":10548,"tokens_out":3303,"would_cite":true,"duration_ms":31920,"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":"The splitting of BPCB's spin-triplet band comes from the weak leg bonds, not the strong rungs.","keywords":["spin ladder","magnetic anisotropy","triplet excitations","inelastic neutron scattering","Dzyaloshinskii-Moriya interaction","strong-coupling expansion","superexchange","BPCB"],"falsifier":"Measure the triplet splitting as a function of the direction and magnitude of a small applied magnetic field: the Ising-only, DM-only, and superexchange-combined leg models predict distinct field-orientation responses, so a dataset that matches none of them would falsify the claim that leg exchange anisotropy is the dominant source.","tokens_in":1611,"feed_emoji":"🧲","tokens_out":2321,"duration_ms":97163,"temperature":0.7,"pith_summary":"This paper identifies the microscopic origin of magnetic anisotropy in the spin-ladder compound BPCB. High-resolution neutron data resolve a splitting of the triplet excitation band of 50(1) \\mu eV at the band minimum and 40(2) \\mu eV at the maximum, and the paper shows that a rung-Ising anisotropy model gives a dispersion that disagrees qualitatively with the data. Three models with weakly anisotropic leg interactions (Ising, Dzyaloshinskii-Moriya, and the combination required by superexchange theory) all reproduce the measured dispersion, so the authors conclude that weakly anisotropic leg exchange is the dominant source of anisotropy in BPCB. If correct, this resolves an earlier unresolved ESR finding and means the anisotropy axis is a property of the weak legs rather than the strong rungs.","feed_headline":"Weak leg bonds, not strong rungs, set BPCB's spin anisotropy","feed_subtitle":"High-resolution neutron data show a 40-50 μeV triplet split that rules out rung anisotropy and points to leg exchange.","key_machinery":"The carrying object is the strong-coupling expansion of the one-triplet dispersion for an $S=1/2$ ladder with anisotropic exchange, built on the small parameter $\\lambda = J_{\\parallel}/J_{\\perp} \\approx 0.28$. Starting from isolated rung singlets and triplets, the calculation produces analytic expressions $\\epsilon_{\\sigma}(k)$ for the three triplet branches in each anisotropy scenario; the $\\sigma=0$ branch and the $\\sigma=\\pm$ branches respond differently to leg versus rung anisotropy. This turns the small measured splitting into a fingerprint that excludes rung Ising anisotropy and attributes the anisotropy to the legs.","core_discovery":"The central claim is that the very small (about 0.6 K) magnetic anisotropy of BPCB lives on the ladder legs. Treating the ladder as strongly coupled rung dimers and adding anisotropic exchange perturbatively, the authors calculate the triplet dispersion up to third order in $\\lambda = J_{\\parallel}/J_{\\perp}$. The rung-Ising case is qualitatively incompatible with the measured band shape, while all leg-anisotropy cases give the observed pattern: a doubly degenerate band with smaller bandwidth, a non-degenerate band with larger bandwidth, and a splitting slightly larger at the band minimum than at the band maximum. Because the Cu$^{2+}$ ions are $S=1/2$, single-ion anisotropy is absent on symmetry grounds, and the estimated dipolar contribution is below 1 \\mu eV, so the observed splitting must be exchange anisotropy. Since only the combined Ising-plus-DM leg model is consistent with the microscopic superexchange picture, the paper concludes that weakly anisotropic leg interactions dominate, with the rungs contributing negligibly.","pith_inferences":["If leg-bond anisotropy sets the anisotropy in BPCB, then in other strong-rung ladders the anisotropy axis may be determined by the weakest exchange paths, so estimates based on the strongest bonds could be misdirected.","The near-degeneracy of the Ising-only, DM-only, and combined leg dispersions suggests that many published 'pure DM' fits to neutron data may in fact be compatible with the full superexchange ratio; field-orientation measurements could separate them.","A testable extension is to compare the predicted $\\lambda$-dependence of the splitting ratio with a series of isostructural ladders in which $J_{\\parallel}$ is varied systematically."],"forward_implications":["The anisotropy axis and the ratio of the band-minimum to band-maximum splitting in BPCB are determined by the leg exchange bonds, not by the much stronger rung bonds.","Ising-type anisotropy on the rungs is negligible, consistent with the inversion symmetry at rung centers and with superexchange theory.","Zero-field neutron data cannot distinguish Ising-only, DM-only, or combined leg anisotropy; the paper's combined model is the only one with microscopic superexchange justification.","A fit to the leg-anisotropy model gives $D_{\\parallel}/k_B = 1.44(2)$ K, but this value should not be read as the anisotropy magnitude; the physical scale is the roughly 0.6 K band splitting.","Measuring the triplet dispersion under a small applied magnetic field is the proposed route to identify which leg-anisotropy term is actually present."],"supporting_citations":[{"why":"Supplies the strong-coupling expansion method and the isotropic triplet dispersion that the anisotropic calculations extend.","marker":"Ref. 27"},{"why":"Provides the rung and leg exchange constants $J_{\\perp}$ and $J_{\\parallel}$ used as the starting point for fitting.","marker":"Ref. 21"},{"why":"Previous ESR study that detected the same anisotropy scale and an anisotropy axis but could not identify its origin.","marker":"Ref. 20"},{"why":"Earlier inelastic neutron scattering on BPCB that resolved the triplet band without the resolution to see the splitting.","marker":"Ref. 16"},{"why":"Theory stating that symmetric and antisymmetric superexchange anisotropy appear together at a fixed ratio, used to argue rung Ising anisotropy should be negligible.","marker":"Refs. 4-6"},{"why":"Earlier study where the full superexchange anisotropy Hamiltonian beat a pure-DM model, used as comparative evidence.","marker":"Ref. 32"}],"fun_headline_variants":["Spin ladder's tiny anisotropy lives on the legs","Leg bonds set spin anisotropy in BPCB ladder","Tiny 50 μeV split traces to leg exchange","Rung anisotropy ruled out; leg anisotropy pinned","Leg exchange, not rung, sets BPCB's spin anisotropy"],"cache_read_input_tokens":12672,"weakest_assumption_plain":"The argument depends on the third-order strong-coupling expansion in $\\lambda = J_{\\parallel}/J_{\\perp}$ being quantitatively reliable for the anisotropic triplet dispersions; the paper itself notes that the predicted maximum splitting shifts from second to third order, so the expansion may not be fully converged at the experimental precision.","fun_headline_variants_meta":{"raw":{"variants":["Spin ladder's tiny anisotropy lives on the legs","Leg bonds set spin anisotropy in BPCB ladder","Tiny 50 μeV split traces to leg exchange","Rung anisotropy ruled out; leg anisotropy pinned","Leg exchange, not rung, sets BPCB's spin anisotropy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000775,"raw_usage":{"total_tokens":3427,"prompt_tokens":939,"completion_tokens":2488,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":555,"completion_tokens_details":{"reasoning_tokens":2411}},"tokens_in":555,"tokens_out":2488,"duration_ms":19207,"temperature":1.0,"reasoning_tokens":2411,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:35:25.248379+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the triplet splitting as a function of the direction and magnitude of a small applied magnetic field: the Ising-only, DM-only, and superexchange-combined leg models predict distinct field-orientation responses, so a dataset that matches none of them would falsify the claim that leg exchange anisotropy is the dominant source.","supporting_citations":[],"review_version":1}