{"id":"38a00be2-36ad-4c3e-b1ad-b07f506a48df","arxiv_id":"1908.07923","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"75As NMR/NQR measurements show suppressed ferromagnetic spin correlations in SrOs4As12 relative to SrFe4As12 and support conventional s-wave superconductivity.","lead":"This paper reports new nuclear magnetic resonance data on the superconductor SrOs4As12, showing its magnetic fluctuations are much weaker than in the related iron compound. The measurements also indicate the superconducting state is conventional s-wave BCS pairing.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The s-wave claim rests on excluding the 2.2–4 K T1 window as 'extrinsic'; if that window is not modeled, the BCS fit is a selection artifact, and the coherence-peak evidence is untested.","rationale":"The reader's weakest assumption identifies exactly the load-bearing step: the 2.2–4 K T1 window is excluded from the BCS analysis without a quantitative model, even though the sample's broad Tc distribution makes that window the most information-rich test of whether the observed temperature dependence is intrinsic or extrinsic. The s-wave claim is the central conclusion most sensitive to this assumption, because the coherence peak just below Tc and the BCS fit are the only evidence for the pairing symmetry; a fully gapped but non-s-wave state could also produce a large low-temperature drop, so the coherence peak must be trustworthy. The FM-suppression claim, by contrast, is supported by the direct comparison of K and χ values that are much smaller in SrOs4As12 than in SrFe4As12, despite the large absolute error bars on K; even if the quantitative 'strong suppression' wording is softened by the comparable low-temperature 1/T1T values, the qualitative conclusion of suppressed ferromagnetic correlations does not depend on the excluded T1 window. The reader's conditional verdict therefore remains appropriate: the paper should be published with the s-wave conclusion explicitly conditioned on a successful two-component analysis of the excluded temperature range.","tokens_in":15030,"tokens_out":6137,"duration_ms":67160,"concrete_test":"Reanalyze the T1 data from 4.8 K to 0.4 K with a two-component normal/superconducting model using the measured zero-field-cooled susceptibility to fix the normal fraction, and fit all points without excluding 2.2–4 K. If the same BCS parameters reproduce the intermediate-temperature hump and the low-temperature tail within the stretched-exponential fit uncertainty, the exclusion is justified; if not, the s-wave claim should be downgraded to 'fully gapped but pairing symmetry not established.'","verdict_should_be":"UNCHANGED","load_bearing_attack":"The s-wave conclusion in §V.B depends on excluding the 2.2–4 K window as 'extrinsic' and 'more artificially' not discussing it. This is not a tested assumption: the sample has a broad resistive transition (onset 4.8 K, zero resistance 2.2 K), the zero-field-cooled susceptibility shows a two-step superconducting fraction, and the recovery curves are stretched with β varying. The observed 1/T1 in that window is therefore a temperature-dependent superposition of normal and superconducting fractions whose weights are changing. The 'coherence peak just below Tc' is extracted without modeling this mixture, and the BCS calculation is then compared only to the remaining points, while an ad hoc constant term 1/T1 = 0.008 s−1 is added to describe the low-temperature tail. If the 2.2–4 K data contain a real multiband or distribution-induced feature, the BCS fit is a selection artifact and the s-wave claim is not established. The FM-suppression claim is less affected because it relies on the direct K and 1/T1T comparison, although the text's statement that 1/T1T values are 'almost comparable' below 20 K somewhat weakens the summary's 'strong suppression' wording.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports 75As NMR and NQR measurements on the filled skutterudite superconductor SrOs4As12, with the goal of comparing its magnetic and superconducting properties with the isostructural, non-superconducting SrFe4As12, which exhibits ferromagnetic spin correlations. The authors determine the Knight shift K from NQR spectra in small applied fields and measure the spin-lattice relaxation rate 1/T1 in the normal and superconducting states. They find that |K| and 1/T1T are smaller in SrOs4As12 than in SrFe4As12 at most temperatures, and that the temperature dependence of 1/T1T can be reproduced by a simple band model with a nearly flat density of states near the Fermi energy. They interpret these results as evidence that ferromagnetic spin fluctuations are strongly suppressed in SrOs4As12. From the temperature dependence of 1/T1 below Tc, including a coherence peak just below Tc and a large decrease at low temperatures, they conclude that the superconductivity is conventional s-wave BCS type.","tokens_in":15329,"tokens_out":3914,"duration_ms":40044,"significance":"If the conclusions hold, the paper provides a useful comparative NMR study of two isostructural filled skutterudites, showing how replacing Fe (3d) with Os (5d) suppresses ferromagnetic correlations and allows conventional s-wave superconductivity to emerge. The strength of the paper is the direct microscopic comparison between SrOs4As12 and SrFe4As12 using the same experimental techniques and analysis methods, which makes the qualitative suppression of the static spin susceptibility credible. The paper also contains substantial experimental detail, including NQR-based Knight shift determination and a careful discussion of the complications arising from the broad superconducting transition. The main caveat is that the s-wave conclusion depends on excluding a significant temperature interval from the BCS analysis, and the Knight shift data carry large uncertainties, so several quantitative statements are less secure than the qualitative comparison.","major_comments":[{"comment":"The s-wave conclusion rests on excluding the temperature range T = 2.2–4 K as extrinsic. The text states: 'we consider that the temperature dependence of 1/T1 in T = 2.2 − 4 K is not intrinsic but extrinsic and more artificially, we will not discuss it in this paper.' This exclusion is not quantitatively justified. The sample shows a broad resistive transition (onset 4.8 K, zero resistance 2.2 K), a two-step superconducting volume fraction, and a β exponent that varies non-monotonically with a local minimum near 3.5 K. These observations imply that the measured 1/T1 in this window is a temperature-dependent superposition of normal and superconducting fractions. Without modeling this mixture, the BCS fit with Δ(0) = 6 K, r = 5, and an added constant 1/T1 = 0.008 s−1 is compared only to the remaining data, so the coherence peak and the derived 2Δ(0)/kBTc = 2.5 are not robust evidence for s-wave pairing. The claim that 'the temperature dependence of 1/T1 in the superconducting state evidences a conventional s-wave superconductivity' is therefore not fully established. Please either model the two-component response over the full temperature range or present the s-wave conclusion as tentative and limited to temperatures below about 2 K, where the superconducting fraction is largest.","section":"§V.B, Fig. 5"},{"comment":"The Knight shift is reported as K = -1.2 ± 2.5 % at 10 K, and the temperature dependence in Fig. 3 shows variations that are smaller than this uncertainty. Therefore the statement that K is 'nearly independent of temperature, consistent with the temperature dependence of the magnetic susceptibility' is not supported by the K data alone; the consistency is only apparent in the susceptibility curves. The K–χ plot in the inset combines data from both compounds and yields a nearly zero intercept, but given the large error bars on K, the conclusion of K0 ≈ 0 and the derived hyperfine coupling Ahf = -3.84 ± 1.60 kOe/μB are rough estimates. This does not invalidate the qualitative suppression of |K| in SrOs4As12 compared with SrFe4As12, but quantitative statements such as the estimated |Ks| ≈ 0.21% from the Korringa relation should be labeled as order-of-magnitude estimates rather than precise experimental values.","section":"§IV, Fig. 3"},{"comment":"The Summary states: 'The large suppression in the |K| and 1/T1T in SrOs4As12 compared with those in SrFe4As12 indicates no obvious ferromagnetic spin correlations.' However, in §V.A the text states: 'It is interesting to point out that the 1/T1T values for both systems are almost comparable at low temperatures below ∼ 20 K in the normal state.' The suppression of 1/T1T is therefore only present at higher temperatures; below 20 K the values are comparable. The inference about ferromagnetic correlations is based on the different temperature dependence of 1/T1T and on the Korringa ratio, not on a uniform suppression of 1/T1T. Please qualify the summary to avoid overstating the 1/T1T suppression, or explicitly distinguish the temperature ranges over which the suppression is observed.","section":"§V.A, Summary"}],"minor_comments":[{"comment":"The stretched-exponential function is written as 1 − M(t)/M(∞) = e−(3t/T1)β, but the exponent is later said to be β ∼ 0.8 in the paramagnetic state. Since β multiplies (3t/T1), it would be helpful to state explicitly that β is the stretching exponent and to define its range for the normal and superconducting states.","section":"§II"},{"comment":"Equation (6) is written with a proportionality sign, and the normalization constant is not specified. Also, the broadening function is described as triangular with width 2δ, but the relationship between δ and the parameters used in the fit (r = Δ(0)/δ) is not shown in the equation. Please clarify the normalization and the precise definition of δ.","section":"§V.B"},{"comment":"The ad hoc constant term 1/T1 = 0.008 s−1 added to reproduce the low-temperature flattening is introduced without a physical model. The text says its 'origin is not clear at present' but suggests impurity effects. This is acceptable, but it would be more transparent to state explicitly that this term is a free parameter of the fit and to discuss its possible relation to the normal-state fraction or magnetic impurity relaxation.","section":"§V.B"},{"comment":"There are several typographical and wording issues, for example: 'more artificially' is an unusual phrase in §V.B; 'on th other hand' in §V.A should be 'on the other hand'; and in the Figure 5 caption, the pink and red curves are described but the labels in the main text are not always consistent. A careful proofread is recommended.","section":"Various"}],"recommendation":"major_revision","confidential_remarks":"The manuscript presents a useful comparative NMR study, but the two central claims have asymmetric support. The suppression of ferromagnetic correlations is reasonably supported by the direct K and 1/T1T comparison, despite large error bars on K. The s-wave conclusion, however, is presented as an established result even though the analysis excludes a substantial temperature interval without modeling the normal/superconducting mixture. I recommend major revision: the authors should either strengthen the s-wave analysis by explicitly treating the two-component response, or substantially soften the claim. The overstatement of the 1/T1T suppression in the summary should also be corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this is a genuine first look at SrOs4As12 with 75As NMR/NQR, and the main comparison—small, roughly T-independent |K| and 1/T1T versus SrFe4As12—makes a solid case that ferromagnetic spin correlations are strongly reduced. The s-wave claim is plausible but not as firm as the abstract suggests, because the BCS analysis deliberately excludes the 2.2–4 K window.\n\nWhat it does well: the K measurement via field-dependent NQR edge is a clever way around the complex powder NMR spectrum. The K–χ plot and near-zero intercept support the spin-shift interpretation. The 1/T1T data above Tc are useful, and the normal-state band model reproduces the qualitative flat-band trend. They also show a coherence peak just below Tc. The authors are unusually honest: they document the broad resistive transition, the two-step superconducting fraction, and the β variation, and they state plainly why they will not discuss the 2.2–4 K region. That is a real strength, not a hidden flaw.\n\nSoft spots: the Knight shift errors (±2.5%) are larger than the claimed temperature variation, so the T-dependence of K should not be oversold. The 1/T1T \"explanation\" is a parametrized band fit, not an independent prediction; the paper mostly presents it that way. The abstract's \"strong suppression\" is a bit stronger than the text, which says 1/T1T values are almost comparable below ~20 K. The load-bearing issue is the excluded window: the data there are almost certainly a mixture of normal and superconducting fractions, but the BCS fit is compared only to the remaining points plus an added constant relaxation term. That does not prove the s-wave assignment; it shows consistency with it. The coherence peak is real evidence, but it sits right at the edge of the excluded region. A referee should ask for a two-component analysis, or at least a clearly stated uncertainty on the s-wave conclusion.\n\nWho benefits: specialists in filled skutterudites and NMR practitioners. This is a solid first report rather than a definitive symmetry determination. I would engage with it and send it to peer review; the requested revisions are clarifications and tempered claims, not new physics.","headline":"New NMR/NQR data make a credible case that ferromagnetic fluctuations are suppressed in SrOs4As12; the s-wave conclusion is plausible but rests on excluding the 2.2–4 K data without a model.","tokens_in":15914,"tokens_out":2498,"would_cite":true,"duration_ms":25752,"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":"SrOs4As12 suppresses the ferromagnetic spin fluctuations of its Fe counterpart and pairs electrons in a conventional s-wave state, 75As NMR shows.","keywords":["NMR","NQR","filled skutterudite","SrOs4As12","ferromagnetic spin fluctuations","s-wave superconductivity","Knight shift","spin-lattice relaxation"],"falsifier":"Measure 75As $1/T_1$ and the Knight shift on a SrOs4As12 sample with a sharp superconducting transition, for instance a single crystal: persistence of the 2.2-4 K anomaly without a $T_c$ distribution would disprove the s-wave assignment, as would a Knight shift that fails to drop below $T_c$ in a spin-singlet state.","tokens_in":14736,"feed_emoji":"🧲","tokens_out":6600,"duration_ms":63535,"temperature":0.7,"pith_summary":"The paper uses 75As NMR and NQR on the newly synthesized filled skutterudite superconductor SrOs4As12 to ask whether magnetic fluctuations control its superconductivity. Comparing with the isostructural, non-superconducting SrFe4As12, the authors find that both the Knight shift magnitude $|K|$ and the relaxation rate $1/T_1T$ are much smaller in the Os compound and nearly temperature independent above 50 K. That means the ferromagnetic spin correlations that dominate SrFe4As12 are strongly suppressed in SrOs4As12. Below $T_c \\sim 4.8$ K, a tiny coherence peak and a two-order-of-magnitude drop in $1/T_1$ are taken as evidence of conventional s-wave BCS pairing. A simple flat-band model with a small ledge near the Fermi energy reproduces the normal-state $1/T_1T$.","feed_headline":"NMR finds suppressed magnetism and s-wave pairing in SrOs4As12","feed_subtitle":"75As measurements show the skutterudite superconductor lacks ferromagnetic correlations and opens a full BCS-like gap.","key_machinery":"The two measured quantities are the Knight shift $K$, obtained from the field dependence of the lower edge of 75As NQR spectra in fields up to 0.5 T, and the nuclear spin-lattice relaxation rate $1/T_1$, normally written as $1/T_1T$. $K$ tracks the static spin susceptibility while $1/T_1T$ tracks the $q$-summed dynamical susceptibility; the strong suppression of both relative to SrFe4As12 is the evidence against ferromagnetic correlations. In the superconducting state the mechanism is the BCS coherence-factor formula for $1/T_1$ with a triangular broadening function, fitted with gap $\\Delta(0)=6$ K and broadening ratio $r=5$, which yields the coherence peak and the exponential-like drop.","core_discovery":"The paper's central claim is that SrOs4As12 sits on the nonmagnetic side of a magnetic-superconducting divide: the 75As Knight shift $K$ and the relaxation rate $1/T_1T$ are much smaller than in SrFe4As12, and the $K$--$\\chi$ plot has nearly zero intercept, showing that the static spin susceptibility is strongly reduced. The nearly temperature-independent $1/T_1T$ above 50 K is consistent with an almost flat band with a small 40 K ledge near the Fermi energy rather than with magnetic correlations. Below $T_c \\sim 4.8$ K, a Hebel-Slichter coherence peak and a drop of $1/T_1$ by more than two orders of magnitude indicate a fully gapped, spin-singlet $s$-wave superconducting state; the gap ratio $2\\Delta(0)/k_BT_c \\approx 2.5$ comes out slightly below the BCS weak-coupling value, and the residual temperature-independent $1/T_1$ below 0.4 K is assigned to impurity effects.","pith_inferences":["A straightforward extension would be a 75As Knight-shift measurement below $T_c$: a spin-singlet s-wave state should show a clear drop of $K$ in the superconducting state, which the present normal-state data do not test.","Comparing SrRu4As12 with the same NMR protocol would map the 3d/4d/5d trend in this family and test whether the suppression of magnetic fluctuations is monotonic with d-electron delocalization.","The fitted flat-band ledge width of 40 K is a single-band parameter; band-structure calculations or specific-heat measurements could upgrade it into an independent quantitative check."],"forward_implications":["The Fe-to-Os substitution removes ferromagnetic spin correlations and is accompanied by conventional s-wave superconductivity, so magnetic fluctuations and singlet s-wave pairing appear to compete in this family.","Because the low-temperature normal-state $1/T_1T$ is comparable in the two compounds, the effective density of states at the Fermi level is similar; the absence of superconductivity in SrFe4As12 is therefore attributed to the ferromagnetic fluctuations, not to a lack of electronic states.","The more-than-two-orders-of-magnitude drop in $1/T_1$ below $T_c$ establishes that superconductivity in SrOs4As12 is bulk, not filamentary or surface.","If ferromagnetic correlations are the obstacle, suppressing them in SrFe4As12, for example by pressure, should restore superconductivity; the authors report that such experiments are underway."],"supporting_citations":[{"why":"Supplies the high-pressure synthesis, lattice constant, $T_c=4.8$ K, resistivity, and susceptibility data used throughout.","marker":"[31]"},{"why":"Theoretical study pointing to the ferromagnetic nature of SrFe4As12 that motivates the comparison.","marker":"[32]"},{"why":"Provides the SrFe4As12 NMR/NQR baseline, the method for extracting $K$ from NQR edges, and the ferromagnetic-correlation signature being compared.","marker":"[33]"},{"why":"Supplies the flat-band model with a small ledge used to explain the normal-state $1/T_1T$ behavior.","marker":"[47]"},{"why":"Shows that in a related compound the complicated low-temperature $1/T_1$ behavior disappears in higher-quality samples, supporting the extrinsic interpretation of the 2.2-4 K window.","marker":"[55]"},{"why":"Provides the BCS coherence-factor formula used to calculate $1/T_1$ in the superconducting state.","marker":"[56]"},{"why":"Supplies the triangular broadening function used to convolute the BCS density of states in the fit.","marker":"[57]"},{"why":"Gives the isostructural BCS-type superconductor CaOs4P12 as a comparison for conventional pairing in this family.","marker":"[58]"}],"fun_headline_variants":["NMR: SrOs4As12 suppresses ferromagnetic spin fluctuations","SrOs4As12 is nonmagnetic: 75As NMR finds no spin correlations","s-wave superconductor SrOs4As12 shows no ferromagnetic order","75As NMR: SrOs4As12 lacks the magnetism of SrFe4As12","Full BCS-like gap in SrOs4As12 with quenched spin fluctuations"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the complicated $1/T_1$ behavior between 2.2 K and 4 K is an extrinsic artifact of a distribution of superconducting $T_c$ values mixed with normal-state signal; if that window is intrinsic, the BCS fit and s-wave assignment rest on selected data.","fun_headline_variants_meta":{"raw":{"variants":["NMR: SrOs4As12 suppresses ferromagnetic spin fluctuations","SrOs4As12 is nonmagnetic: 75As NMR finds no spin correlations","s-wave superconductor SrOs4As12 shows no ferromagnetic order","75As NMR: SrOs4As12 lacks the magnetism of SrFe4As12","Full BCS-like gap in SrOs4As12 with quenched spin fluctuations"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000286,"raw_usage":{"total_tokens":1764,"prompt_tokens":1110,"completion_tokens":654,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":726,"completion_tokens_details":{"reasoning_tokens":548}},"tokens_in":726,"tokens_out":654,"duration_ms":6514,"temperature":1.0,"reasoning_tokens":548,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:20:29.564835+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure 75As $1/T_1$ and the Knight shift on a SrOs4As12 sample with a sharp superconducting transition, for instance a single crystal: persistence of the 2.2-4 K anomaly without a $T_c$ distribution would disprove the s-wave assignment, as would a Knight shift that fails to drop below $T_c$ in a spin-singlet state.","supporting_citations":[{"cited_title":"Takegaraha, M","cited_arxiv_id":null,"evidence_quote":"Supplies the high-pressure synthesis, lattice constant, $T_c=4.8$ K, resistivity, and susceptibility data used throughout."},{"cited_title":"Nishine, Y","cited_arxiv_id":null,"evidence_quote":"Theoretical study pointing to the ferromagnetic nature of SrFe4As12 that motivates the comparison."},{"cited_title":"Shankar, Sandeep, D","cited_arxiv_id":null,"evidence_quote":"Provides the SrFe4As12 NMR/NQR baseline, the method for extracting $K$ from NQR edges, and the ferromagnetic-correlation signature being compared."},{"cited_title":"Here NA is Avogadro’s number and z = 2 is the number of the nearest neighbor Os ions at the As site","cited_arxiv_id":null,"evidence_quote":"Supplies the flat-band model with a small ledge used to explain the normal-state $1/T_1T$ behavior."},{"cited_title":"Br¨ uckner, R","cited_arxiv_id":null,"evidence_quote":"Shows that in a related compound the complicated low-temperature $1/T_1$ behavior disappears in higher-quality samples, supporting the extrinsic interpretation of the 2.2-4 K window."},{"cited_title":"Matano, K","cited_arxiv_id":null,"evidence_quote":"Provides the BCS coherence-factor formula used to calculate $1/T_1$ in the superconducting state."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the triangular broadening function used to convolute the BCS density of states in the fit."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the isostructural BCS-type superconductor CaOs4P12 as a comparison for conventional pairing in this family."}],"review_version":1}