{"id":"a35ed71c-0211-47de-ab70-0c6cd9ac892a","arxiv_id":"2507.09056","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Polycrystalline, cubic ScSb shows a Nernst thermopower of 128 µV/K at 30 K and 14 T and a Nernst power factor of 240e-4 W m^-1 K^-2, exceeding single-crystal ScSb due to improved electron-hole compensation.","lead":"A powder-sintered form of the semimetal ScSb produces a large Nernst thermopower of 128 µV/K at 30 K in a 14 T field, outperforming its own single-crystal form. The result suggests that cubic, compensated semimetals can deliver strong transverse thermoelectric response in cheap polycrystalline form.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The compensation mechanism is asserted rather than tested: Eq. 3 is never compared with the measured Syx(B), and the phonon-drag contribution to the 30 K peak is not separated. Without that check, the paper's central explanation for the linear non-saturating Nernst effect is unsupported.","rationale":"The reader's weakest assumption identifies the same load-bearing concern: the two-carrier fit parameters are used qualitatively to explain the linear non-saturating Syx, but Eq. 3 is never checked against the measured field dependence. That is precisely where the argument is least secure. The raw measured values (Syx ≈ 128 μV/K at 30 K and 14 T, MR ≈ 940% at 2 K) are internally consistent with PFN and zN, so I do not see grounds to reject the headline data. The conditional verdict is appropriate: the numbers can stand while the mechanism claim requires an additional quantitative test. I therefore recommend no change to the reader's CONDITIONAL verdict. The additional inconsistencies in the single-crystal baseline (47 vs 35 μV/K) and the unseparated phonon-drag contribution support the same conclusion rather than a stronger one. A direct recomputation of Eq. 3 from the published fit parameters would settle whether the compensation model is compatible with the observed linear Syx(B), or whether the paper's central explanatory claim is unsupported.","tokens_in":12607,"tokens_out":7516,"duration_ms":88337,"concrete_test":"At 30 K, compute the normalized two-carrier Nernst curve y(B) = [σ_e σ_h (μ_e + μ_h) B] / [(σ_e + σ_h)^2 + (σ_e μ_e B − σ_h μ_h B)^2] using the fitted carrier densities and mobilities from Fig. 2(d), and overlay y(B)/y(B0) on the measured Syx(B)/Syx(B0) from Fig. 3(d) over 2–14 T. If the model curve bends over while the measured data remains linear beyond the experimental scatter, the compensation mechanism fails the paper's own two-carrier test. As a secondary check, fit the data as Syx(B) = A y(B) + c B to allow a linear phonon-drag channel, and test whether A is independent of temperature; a strongly temperature-dependent or field-dependent A would indicate an additional carrier or phonon-drag contribution.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central explanation—that better electron-hole compensation causes the large linear Nernst signal—is never quantitatively checked. Section III introduces Eq. 3 as the two-carrier expression for Syx and then uses the fitted n_h/n_e ≈ 1.1 and μ_h/μ_e ≈ 1.2 to argue that ScSb is nearly ideal and therefore Syx stays linear up to 14 T. But Eq. 3 is never evaluated against the measured Syx(B) in Fig. 3(d). The B dependence of Eq. 3 is fixed by the fitted densities and mobilities up to the unknown prefactor (S_h−S_e), so this is a straightforward, falsifiable test. The imbalance term in the denominator is not negligible for the quoted parameters: with μ_e B of order 2–4 at 14 T, the term (σ_e μ_e B − σ_h μ_h B)^2 is comparable to (σ_e + σ_h)^2, so the model generically predicts curvature or saturation in Syx(B) at the highest fields. The observed linear, non-saturating behavior therefore must be shown to be consistent with Eq. 3, not assumed. In addition, the paper attributes the 30 K peak in both Sxx and Syx to phonon drag, but no separation of the drag contribution is attempted, so the peak PFN and zN values cannot unambiguously be assigned to the compensation mechanism. Finally, the single-crystal baseline is internally inconsistent (47 μV/K in Table I vs 35 μV/K in the text), which weakens the 'improved' claim even if the raw polycrystalline data are correct.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports transport and thermomagnetic measurements on polycrystalline ScSb, a cubic, topologically trivial compensated semimetal. The headline results are a Nernst thermopower of about 128 µV/K at 30 K and 14 T, a Nernst power factor of about 240 × 10^-4 W m^-1 K^-2, a Nernst figure of merit of about 11 × 10^-4 K^-1, and a non-saturating magnetoresistance of about 940% at 2 K and 14 T. The authors attribute the enhanced Nernst and magnetotransport properties to improved electron-hole compensation, with fitted carrier density ratio nh/ne ≈ 1.1 and mobility ratio μh/μe ≈ 1.2, obtained from a two-carrier fit to the longitudinal and Hall conductivities. They compare the results with their previous single-crystal ScSb study and with other thermomagnetic materials.","tokens_in":12772,"tokens_out":4271,"duration_ms":52869,"significance":"If the central claim holds, the paper is significant: it shows that a cubic, topologically trivial semimetal in polycrystalline form can exhibit a Nernst response comparable to or better than that of many topological semimetal single crystals, despite the usual degradation expected in polycrystalline samples. The reported PFN and zN values are internally consistent with the plotted resistivity and thermal conductivity data, and the use of standard PPMS/TTO protocols with antisymmetry checks for the Nernst signal adds credibility to the measurements. The main weakness is that the paper's mechanistic explanation—that near-perfect compensation produces the linear non-saturating Nernst signal—is asserted rather than quantitatively tested against the measured Syx(B) curves. The manuscript also contains an unresolved inconsistency in the single-crystal baseline. These issues are fixable within the scope of the paper, so the work warrants revision rather than rejection.","major_comments":[{"comment":"The central mechanism claim is never tested quantitatively. Equation (3) gives the two-carrier expression for Syx whose magnetic-field dependence is fixed by the fitted densities and mobilities up to the prefactor (S_h - S_e). The manuscript could directly compare Eq. (3) with the measured Syx(B) in Fig. 3(d) by normalizing the prefactor at one field, but no such comparison is shown. This is a straightforward and falsifiable check. With the quoted nh/ne ≈ 1.1 and μh/μe ≈ 1.2, and μB values of order 2-4 at 14 T, the imbalance term (σ_e μ_e B - σ_h μ_h B)^2 in the denominator of Eq. (3) is not negligible, so the model does not automatically give linear non-saturating Syx. The observed linear behavior must be shown to be consistent with Eq. (3), not assumed.","section":"Section III, Eq. (3) and Fig. 3(d)"},{"comment":"The 30 K peak in Syx is attributed to phonon drag, but no quantitative separation of the phonon-drag and diffusion contributions is attempted. Because the maxima of PFN and zN occur at this same 30 K peak, the claim that the enhanced Nernst performance originates from better electron-hole compensation is not uniquely supported; a phonon-drag contribution could carry part or most of the enhancement. The manuscript should either separate the drag contribution using a standard analysis or explicitly bound its magnitude before assigning the peak to the compensation mechanism.","section":"Section III, phonon drag attribution (Fig. 3(c))"},{"comment":"The single-crystal baseline is internally inconsistent: Table I lists Syx ≈ 47 µV/K (12 K, 14 T) for single-crystal ScSb, while the text and Conclusions state Syx ≈ 35 µV/K at the same conditions. Since the paper's central claim is that the polycrystalline sample shows improved Nernst performance compared with the single crystal, this discrepancy directly affects the claimed improvement factor and must be resolved. The same issue affects the statement in the Introduction that the PFN is 'nearly seven times' the single-crystal value at the same temperature and field.","section":"Table I and Sections III/IV"}],"minor_comments":[{"comment":"The caption for Fig. 2(a) says 'Longitudinal resistivity (ρyy)', but the panel axis and the text in Section III describe it as magnetoresistance (MR %); please correct the caption to match the plotted quantity.","section":"Fig. 2 caption"},{"comment":"The phrase 'nearly seven times the value found in the single crystal samples at the same temperature and magnetic field' is misleading because Table I reports the single-crystal maximum at 12 K, while the polycrystalline maximum occurs at 30 K; please specify the exact comparison conditions.","section":"Introduction"},{"comment":"The phrase 'in comparison to that was observed' should be rewritten, for example as 'compared with that observed in single-crystal ScSb previously.'","section":"Abstract"},{"comment":"There is a typo in 'in a a JEOL JSM-7600F scanning electron microscope'; delete the duplicated article.","section":"Methods"},{"comment":"Reference [42] lacks author names and a full title; please provide the complete citation.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within scope for a condensed-matter physics journal. The referee's main concern is not the raw data but the unsupported mechanism claim. During revision, I would ask the authors to provide the direct comparison of Eq. (3) with the measured Syx(B), to address the phonon-drag decomposition, and to reconcile the single-crystal baseline values. It would also be helpful if the authors made the raw transport data available so that independent checks of the two-carrier fit and the Nernst normalization could be performed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper reports the first Nernst measurements on polycrystalline ScSb: a peak Nernst thermopower of about 128 µV/K at 30 K and 14 T, a Nernst power factor near 240 × 10^-4 W m^-1 K^-2, and a magnetoresistance of about 940% at 2 K. I checked the headline algebra and it is internally consistent—PFN = Syx^2/rho_yy and zN = PFN/kappa_xx follow from the plotted rho_yy (~70 µΩcm) and kappa_xx (~22 W m^-1 K^-1) at 30 K and 14 T. The measurement protocol looks standard, and the antisymmetry check on Syx is a good sign. The material result is genuinely new; nothing in the cited literature has Nernst data on polycrystalline ScSb.\n\nThe soft spots are real but not fatal. The central mechanism claim—that better electron-hole compensation causes the linear non-saturating Syx—is asserted rather than tested. The authors fit a two-carrier model to Hall and longitudinal conductivity, extract nh/ne ≈ 1.1 and μh/μe ≈ 1.2, then invoke Eq. 3 to explain the linear field dependence without ever plugging those parameters back into Eq. 3 and comparing with the measured Syx(B) in Fig. 3(d). The stress-test note is correct that with μeB of order 2–4 at 14 T, the denominator in Eq. 3 is not in the simple linear regime, so the model may actually predict curvature. This is a fixable omission, and the data needed for the check are already in the paper. Also, the 30 K peak in both Sxx and Syx is attributed to phonon drag, but no drag separation is attempted, so the peak PFN and zN values cannot be unambiguously assigned to the compensation mechanism. The single-crystal baseline is inconsistently quoted (47 µV/K in Table I vs 35 µV/K in the text), and the \"nearly seven times\" PFN claim compares peaks at different temperatures (12 K vs 30 K) despite the text implying the same temperature. These are errors, not fraud, but they need cleaning up.\n\nThe Wiedemann–Franz violation side claim rests on two-parameter fits to four field points per temperature; I would present that as preliminary.\n\nWho is this for? The thermomagnetic transport community—people tracking Nernst power factors in semimetals will want to know that a cubic, topologically trivial, compensated semimetal in polycrystalline form can reach numbers comparable to Mg3Bi2 and NbP. The paper deserves a serious referee. The headline measurements are credible, and the mechanism can be tested with data already in hand. I would send it to review with a request for a quantitative comparison of Eq. 3 with the measured Syx(B), error bars on the headline numbers, a consistent single-crystal baseline, and an explicit note that the PFN comparison is peak-to-peak at different temperatures.","headline":"Credible new Nernst data on polycrystalline ScSb, but the compensation mechanism is asserted rather than tested quantitatively.","tokens_in":13609,"tokens_out":3986,"would_cite":false,"duration_ms":39590,"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":"Polycrystalline ScSb reaches a Nernst thermopower of about 128 µV/K at 30 K and 14 T, outperforming its single crystal via better electron-hole compensation.","keywords":["Nernst thermopower","magnetoresistance","compensated semimetal","ScSb","polycrystalline","electron-hole compensation","transverse thermoelectric effect","two-carrier model"],"falsifier":"Using the fitted $n_h$, $n_e$, $\\mu_h$, $\\mu_e$ at 30 K in equation (3), compute the predicted $S_{yx}(B)$ and compare it with the measured linear curve that reaches 128 µV/K at 14 T; a large mismatch would show that the fitted compensation parameters do not actually explain the Nernst response. A second decisive check is to vary the Sc:Sb stoichiometry: if compensation is the cause, the Nernst peak should be largest near $n_h/n_e = 1$ and diminish as the ratio moves away from unity.","tokens_in":12128,"feed_emoji":"🧲","tokens_out":9711,"duration_ms":98311,"temperature":0.7,"pith_summary":"Polycrystalline ScSb—a cubic, topologically trivial semimetal—is reported to produce a Nernst thermopower of about 128 µV/K at 30 K and 14 T, a Nernst power factor of about 240 × 10⁻⁴ W m⁻¹ K⁻², and a magnetoresistance of about 940% at 2 K. The field dependence of the Nernst signal is linear and non-saturating up to 14 T. The paper's central claim is that these enhancements come from better electron-hole compensation in the sintered sample, with hole-to-electron density ratio $n_h/n_e \\approx 1.1$, compared with single-crystal ScSb where the ratio is about 0.8. A reader would care because polycrystalline samples are cheap and scalable to bulk sizes, and this result indicates that cubic, topologically trivial materials can deliver transverse thermoelectric performance comparable to single crystals and to topological semimetals such as NbP and Mg3Bi2.","feed_headline":"Polycrystalline ScSb delivers 128 µV/K Nernst thermopower at 30 K","feed_subtitle":"A cubic, topologically trivial semimetal reaches this because its electrons and holes nearly balance.","key_machinery":"The working machinery is the semiclassical two-carrier model, expressed in equations (1) and (2), which writes the longitudinal and Hall conductivities as sums of hole and electron contributions with densities $n_h$, $n_e$ and mobilities $\\mu_h$, $\\mu_e$. Simultaneous fits to the measured $\\sigma_{yy}(B)$ and $\\sigma_{xy}(B)$ produce the near-compensation ratios that carry the argument. The companion expressions, equations (3) and (4), give the Nernst thermopower $S_{yx}$ in terms of the same two-carrier parameters and show that when the electron and hole Hall terms cancel, $S_{yx} = (S^h_{xx}-S^e_{xx})\\mu B/2$, a linear non-saturating field dependence.","core_discovery":"On the paper's own terms, the central discovery is that polycrystalline ScSb outperforms its single-crystal form on transverse thermoelectric metrics: $S_{yx} \\approx 128$ µV/K at 30 K and 14 T versus $\\sim 47$ µV/K at 12 K and 14 T for the single crystal, and a Nernst power factor $PF_N \\approx 240 \\times 10^{-4}$ W m⁻¹ K⁻², about seven times the single-crystal value. The authors attribute the improvement to better electron-hole compensation: a two-carrier fit to Hall and longitudinal conductivity gives $n_h/n_e \\approx 1.1$ and $\\mu_h/\\mu_e \\approx 1.2$, closer to perfect balance than the single crystal's $n_h/n_e \\approx 0.8$. They further use the two-carrier Nernst expression to explain why $S_{yx}(B)$ stays linear without saturating: when electron and hole Hall contributions nearly cancel, the Nernst signal grows as $\\mu B$ even in fields where the usual saturation would set in.","pith_inferences":["Because ScSb is cubic and topologically trivial, it provides a cleaner benchmark than Weyl or Dirac semimetals for testing whether electron-hole compensation alone can drive a large Nernst signal; a quantitative reconstruction of $S_{yx}(B)$ from the fitted parameters would settle the mechanism.","The 30 K phonon-drag peak is invoked for both $S_{xx}$ and $S_{yx}$ but never separated from the compensation term, so part of the 128 µV/K could be phonon drag; an isotope or grain-size series could separate the two contributions.","The same spark-plasma-sintering route could be applied to other members of the Sc/Y/La monopnictide family to map Nernst thermopower versus $n_h/n_e$ and test whether the maximum always occurs at perfect compensation."],"forward_implications":["Polycrystalline ScSb becomes a practical low-temperature transverse thermoelectric without single-crystal growth or crystallographic alignment.","The near-perfect compensation obtained in a sintered sample suggests that bulk synthesis with exact stoichiometry can tune the Fermi level of other cubic MX monopnictides toward $n_h/n_e \\approx 1$.","Because $S_{yx}$ remains linear and unsaturated at 14 T, still higher applied fields should yield proportionally larger Nernst signals if the compensation mechanism holds.","A Nernst power factor of about $240 \\times 10^{-4}$ W m⁻¹ K⁻² exceeds the Seebeck power factors of Bi2Te3, PbTe, and SnSe, making transverse devices competitive for cryogenic cooling and waste-heat recovery."],"supporting_citations":[{"why":"Prior single-crystal ScSb study that supplies the baseline Nernst values and compensation ratio this work compares against.","marker":"[28]"},{"why":"Single-crystal ScSb magnetoresistance and Fermi-surface study establishing the family's compensation-driven large MR and the 500–28000% variability.","marker":"[14]"},{"why":"Sources of the two-carrier conductivity equations (1) and (2) used to extract carrier densities and mobilities.","marker":"[36, 37]"},{"why":"Origin of equations (3) and (4) relating the Nernst thermopower to two-carrier parameters and compensation.","marker":"[12, 39]"},{"why":"Prior observations of non-saturating Nernst thermopower and phonon-drag peaks in NbSb2 used as comparison and motivation.","marker":"[9, 34]"},{"why":"Polycrystalline NbP data showing the sharp Nernst drop from single crystal to polycrystal, the contrast that cubic ScSb is meant to overcome.","marker":"[13]"},{"why":"Polycrystalline Mg3Bi2 Nernst value that ScSb is reported to rival.","marker":"[40]"}],"fun_headline_variants":["Polycrystalline ScSb delivers 128 µV/K Nernst thermopower","ScSb polycrystal: non-saturating Nernst at 30 K","Polycrystalline ScSb beats single crystal in Nernst","Electron-hole balance boosts Nernst in ScSb polycrystal"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the two-carrier fit to one polycrystalline sample's Hall and longitudinal conductivity uniquely captures the electron and hole densities and mobilities, and that these same carriers, rather than phonon drag or grain-boundary effects, produce the measured Nernst signal.","fun_headline_variants_meta":{"raw":{"variants":["Polycrystalline ScSb delivers 128 µV/K Nernst thermopower","ScSb polycrystal: non-saturating Nernst at 30 K","Polycrystalline ScSb beats single crystal in Nernst","Electron-hole balance boosts Nernst in ScSb polycrystal"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000655,"raw_usage":{"total_tokens":3051,"prompt_tokens":1050,"completion_tokens":2001,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":666,"completion_tokens_details":{"reasoning_tokens":1920}},"tokens_in":666,"tokens_out":2001,"duration_ms":16953,"temperature":1.0,"reasoning_tokens":1920,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:08:57.775801+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Using the fitted $n_h$, $n_e$, $\\mu_h$, $\\mu_e$ at 30 K in equation (3), compute the predicted $S_{yx}(B)$ and compare it with the measured linear curve that reaches 128 µV/K at 14 T; a large mismatch would show that the fitted compensation parameters do not actually explain the Nernst response. A second decisive check is to vary the Sc:Sb stoichiometry: if compensation is the cause, the Nernst peak should be largest near $n_h/n_e = 1$ and diminish as the ratio moves away from unity.","supporting_citations":[{"cited_title":"Nernst power factor and figure of merit in compensated semimetal ScSb","cited_arxiv_id":"2504.15450","evidence_quote":"Prior single-crystal ScSb study that supplies the baseline Nernst values and compensation ratio this work compares against."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Single-crystal ScSb magnetoresistance and Fermi-surface study establishing the family's compensation-driven large MR and the 500–28000% variability."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Polycrystalline NbP data showing the sharp Nernst drop from single crystal to polycrystal, the contrast that cubic ScSb is meant to overcome."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Polycrystalline Mg3Bi2 Nernst value that ScSb is reported to rival."}],"review_version":1}