{"id":"1979a2ac-fe9f-438d-aca9-b083608b4667","arxiv_id":"2507.06372","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"SmN films grown on MgO can be tuned from ferromagnetic insulating to ferromagnetic metallic by changing substrate temperature, with no superconductivity observed down to 0.35 K.","lead":"This study shows that growth temperature, not nitrogen flow, is the main lever for doping epitaxial samarium nitride films, switching them from ferromagnetic insulators to ferromagnetic metals. The most conductive films still show no superconductivity down to 0.35 K, a null result that complicates earlier reports of superconducting SmN and points to defects as the likely origin.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The insulator-metal tuning and the superconductivity null both assume the 3 nm CrN cap is electrically passive; the paper gives no CrN conductivity data or parallel-channel correction, so measured transport may not be intrinsic to SmN.","rationale":"The reader's weakest-assumption analysis correctly identifies the CrN cap as the most load-bearing assumption. The paper's unique positive result is the growth-parameter tuning of transport, and the no-superconductivity bound is secondary; both depend on the measured electrical response being from SmN. Without cap characterization, a metallic or semiconducting CrN layer of 3 nm could contribute a sheet conductance comparable to the 30 nm SmN film, especially for the high-resistivity samples. The paper's inclusion of a GaN-capped sample shows that a control is feasible but the comparison is not made, so the omission is a real gap rather than a fundamental impossibility. I agree with the reader's CONDITIONAL verdict: the issue is concrete, addressable, and not fatal on its own, but the transport conclusions cannot be fully accepted until the cap's electrical role is quantified. No alternative concern is more central; the structural and magnetic data are consistent and the superconductivity discussion is appropriately caveated.","tokens_in":7471,"tokens_out":7569,"duration_ms":87379,"concrete_test":"Co-grow a set of SmN films under identical conditions (e.g., Tsub = 813 C, 0.1 sccm) with the standard 3 nm CrN cap and with a GaN cap, and separately grow a 3 nm CrN film on MgO(001) under the same capping conditions. Measure van der Pauw resistivity and Hall carrier density for each at room temperature and down to 2 K. If the CrN-only film's sheet conductance is below a few percent of the SmN stack's, and if the GaN-capped SmN values agree with the CrN-capped sample to within measurement error, the cap is passive and the reported trend is intrinsic. If either condition fails, the transport data must be corrected for parallel conduction or the central claim is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that SmN is tuned from insulating-ferromagnetic to metallic-ferromagnetic, and the null observation of no superconductivity, is inferred from resistivity and room-temperature Hall data on 30 nm SmN films capped with 3 nm CrN (Methods; Fig. 2). This inference requires the cap to be a passive, non-conductive layer. The paper does not report the resistivity or Hall coefficient of CrN grown under the same conditions, does not correct for parallel conduction, and does not discuss the CrN/SmN interface. The situation is aggravated by Fig. 1(d), which shows an additional strained interfacial layer that could itself carry current. If the CrN cap or the interfacial layer contributes carriers, the stated carrier densities (4.14e19 to 5.51e20 cm^-3) and the magnitude of the insulator-metal crossover are distorted, possibly changing the classification of the lower-temperature films. The paper actually contains a GaN-capped SmN sample (Fig. 2(f)), which could serve as a control, but it is not quantitatively compared with the CrN-capped samples.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports molecular-beam epitaxy growth of 30 nm SmN films on MgO(001) with a 3 nm CrN cap, varying substrate temperature (576–813 °C) and nitrogen mass flow (0.1–2.0 sccm). The authors claim that these growth parameters tune the electronic properties from an insulating ferromagnetic state to a metallic ferromagnetic state, with room-temperature Hall carrier densities increasing from 4.14×10^19 to 5.51×10^20 cm^-3 as substrate temperature increases. Structural characterization shows highly crystalline, epitaxial films with a possible strained interfacial Sm-O-N layer. Resistivity measurements show a ferromagnetic transition near 27.5 K, negative magnetoresistance below TC, and, in the most conductive samples, no superconductivity down to 0.35 K. The paper discusses possible reasons for the absence of superconductivity, including carrier concentration, disorder, strain, and quantum confinement.","tokens_in":7641,"tokens_out":3527,"duration_ms":44467,"significance":"If the central claims hold, the paper establishes substrate temperature as a primary synthesis knob for controlling the electronic phase of epitaxial SmN and provides a useful experimental bound on superconductivity in high-quality thin films. The systematic series of 14 samples, the structural characterization (XRD, RSM, pole figures), and the transport and Hall measurements are valuable empirical contributions. The authors also explicitly connect their observations to the proposed triplet-pairing scenario, which gives the null result interpretive weight. The main weakness is that the transport conclusions rely on the uncharacterized electrical behavior of the CrN cap and a possible interfacial layer, and the superconductivity null is not yet fully documented in terms of which samples reached 0.35 K.","major_comments":[{"comment":"The 3 nm CrN cap grown at 650 °C is implicitly assumed to be electrically passive, but the paper provides no resistivity or Hall data for CrN grown under the same conditions, no parallel-conduction correction, and no discussion of possible conduction through the CrN/SmN interface. Since the central insulator-to-metal tuning claim and the reported carrier densities (4.14×10^19–5.51×10^20 cm^-3) are extracted from transport and Hall measurements on capped films, a conductive cap or the strained interfacial layer seen in Fig. 1(d) could materially distort these results. Please provide a CrN-only control, a two-layer transport model, or at least an estimate of the cap's sheet conductance; the GaN-capped sample in Fig. 2(f) is a promising control but is not quantitatively compared.","section":"Methods; Fig. 2(a)–(e)"},{"comment":"The statement that no superconductivity is observed down to 0.35 K is broader than the measurements justify. The Methods say the PPMS He3 option was used only for selected samples, and the text does not state how many films were measured to 0.35 K or which growth conditions they correspond to. Fig. 2(f), cited in the superconductivity discussion, is the GaN-capped sample, not necessarily the most conductive CrN-capped sample. Please restrict the null claim to the specific samples measured and specify their growth parameters.","section":"Methods; Results and Discussion; Fig. 2(f)–(h)"},{"comment":"Room-temperature Hall carrier densities are reported as a function of substrate temperature and nitrogen flow for 14 samples, but no error bars, measurement reproducibility, or Hall analysis assumptions are given. In a ferromagnetic, potentially multi-band semiconductor, the single-band Hall formula may not give the true carrier density. The comparison with the superconducting polycrystalline samples (n_i = 2×10^21 cm^-3) therefore needs an explicit statement of the Hall factor and its uncertainty.","section":"Fig. 2(e); Results and Discussion"},{"comment":"The semiconducting samples are said to have small energy gaps of 4–8 meV in the paramagnetic state, but no equation, fitting procedure, or definition of the gap is provided. Because this estimate contributes to classifying the low-temperature films as insulating, the extraction method (e.g., activation energy from ln ρ versus 1/T) and the fitting range should be stated.","section":"Results and Discussion"}],"minor_comments":[{"comment":"The main text refers to phi scans as 'seen in Fig. 1(g)', but the Fig. 1 caption has no panel (g); the phi-scan panel appears to be (b). Please correct the cross-references and label all panels consistently.","section":"Fig. 1 caption; main text"},{"comment":"The word 'Deaulaney' should be 'Delaunay' in the description of the contouring triangulation algorithm.","section":"Methods"},{"comment":"The phrase 'in agreements with previous estimates' should be 'in agreement with previous estimates', and 'we did not observed' should be 'we did not observe'.","section":"Results and Discussion"},{"comment":"The text says 'As can be seen from Figs. 2(f), no sign of a superconducting state is observed down to 0.35 K', but the panel shown is a GaN-capped sample. Please clarify in the caption and text which sample is being displayed and whether it is representative of the most conductive films.","section":"Fig. 2(f)"},{"comment":"The notation SmN_x is introduced but then the text reverts to SmN without stating the measured nitrogen stoichiometry or its uncertainty. Please state whether x was quantified or is used only as a formal notation.","section":"Abstract; main text"},{"comment":"The carrier mobility values (25.4 and 52.68 cm^2/V-s) are reported without units typeset consistently; use cm^2 V^-1 s^-1 and state the measurement temperature (presumably room temperature).","section":"Results and Discussion"}],"recommendation":"major_revision","confidential_remarks":"The central empirical result is plausible and the transport data are valuable, but the uncharacterized CrN cap is a load-bearing gap that could change the insulator-to-metal classification. If the authors provide a CrN-only control or a parallel-conduction bound, the paper could become acceptable. The superconductivity null also needs to be stated with the exact sample count and conditions. The self-citation to the prior synthesis paper is normal and not problematic for this journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The real news here is practical and clean: over a 576–813 °C substrate temperature range, the authors move epitaxial SmN on MgO(001) from a ferromagnetic insulator to a ferromagnetic metal, with Hall carrier density rising from 4.14×10^19 to 5.51×10^20 cm^-3. The tuning map (Fig. 2e) is a systematic extension of their earlier MgO synthesis work and of the nitrogen-vacancy doping picture, and it establishes substrate temperature as the dominant knob, more so than nitrogen flux. The structural quality is well documented, with narrow XRD FWHM across the series. The superconductivity null, down to 0.35 K, is honestly reported and appropriately caveated: the carrier density is an order of magnitude lower than in the prior superconducting polycrystalline samples, so the result constrains but does not kill the pairing discussion. The authors also give a sensible list of possible reasons for the absence, from defect sensitivity to strain and quantum confinement.\n\nThe soft spots are real but not fatal. The biggest one is exactly what the stress-test flags: all transport and Hall data come from 30 nm SmN capped with 3 nm CrN, and the paper nowhere establishes that the cap is electrically passive. CrN can be a decent conductor depending on stoichiometry, and the interfacial Sm-O-N layer seen in Fig. 1(d) could also carry current. Since the cap is identical across all samples, the relative trend across growth conditions probably survives, but the absolute carrier densities and the assignment of the low-Tsub films as insulating could shift if the cap contributes significantly. The authors even have a GaN-capped sample (Fig. 2f) that could serve as a control, but they never compare it quantitatively to the CrN-capped ones. That should be fixed: either measure the cap's conductivity under identical conditions, do a parallel-channel correction, or at least discuss why a 3 nm CrN layer is expected to be negligible. Minor issues: no error bars on the Hall carrier densities, XPS degradation prevents direct nitrogen-vacancy quantification (the paper admits this), and there are some inconsistent figure references that a careful referee should catch.\n\nOverall, this is a useful contribution for anyone working on rare-earth nitride heteroepitaxy or on the ferromagnetic semiconductor/superconductor debate in SmN. The central empirical claim—that growth temperature controls the electronic phase—is credible and internally consistent. I'd send it to peer review, but I'd ask for cap characterization and error bars before accepting. It deserves serious referee time.","headline":"A solid experimental report showing substrate temperature tunes SmN from insulating to metallic ferromagnet on MgO, with an honest superconductivity null—just needs to address the CrN cap as a possible parallel conduction path.","tokens_in":8208,"tokens_out":1615,"would_cite":true,"duration_ms":20601,"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":"Growth temperature tunes epitaxial SmN from a ferromagnetic insulator to a ferromagnetic metal, with no superconductivity down to 0.35 K.","keywords":["samarium nitride","molecular beam epitaxy","ferromagnetic semiconductor","metal-insulator transition","nitrogen vacancies","superconductivity","thin films","transport properties"],"falsifier":"Measure the resistivity and Hall effect of a SmN film grown without the CrN cap (or of a CrN-only film on MgO under identical growth conditions). If the CrN-only film shows comparable conduction or if the uncapped SmN film does not reproduce the metallic crossover and carrier densities, the claim of an intrinsic tuning of SmN's electronic phase fails.","tokens_in":7266,"feed_emoji":"⚡","tokens_out":5269,"duration_ms":51416,"temperature":0.7,"pith_summary":"This paper establishes that the electronic phase of epitaxial samarium nitride (SmN) can be tuned during molecular beam epitaxy by growth parameters, chiefly substrate temperature. Raising the substrate temperature from 576 °C to 813 °C at fixed low nitrogen flux transforms the film from a ferromagnetic insulator with a small gap (4–8 meV) to a ferromagnetic metal, with carrier density increasing from 4.14×$10^{19}$ to 5.51×$10^{20}$ $cm^{-3}$ while crystallinity stays high. The tuning is attributed to nitrogen vacancies acting as electron donors. The paper also reports that no superconductivity appears down to 0.35 K in any film, including the most conductive, which places new constraints on earlier claims of superconductivity in SmN and on the pairing mechanism.","feed_headline":"One temperature knob tunes SmN from insulator to metal","feed_subtitle":"Carrier density jumps tenfold and films stay ferromagnetic, yet no superconductivity appears down to 0.35 K.","key_machinery":"The central knob is the density of nitrogen vacancies (N$_V$) in the rocksalt SmN lattice, controlled during growth by substrate temperature and nitrogen flux. Each vacancy releases electrons from the three Sm atoms it coordinates; Sm 4f states drop to the Fermi level and hybridize with Sm 5d and N 2p states, converting the material from a small-gap semiconductor (gap 4–8 meV) to a degenerate metal. The transport signatures — positive temperature coefficient of resistivity, $T^2$ resistivity below ordering, negative magnetoresistance — all track this carrier density.","core_discovery":"The paper shows that substrate temperature during molecular beam epitaxy is the dominant synthesis knob for the electronic phase of SmN on MgO(001). By raising substrate temperature from 576°C to 813°C at low nitrogen flux, with all other parameters fixed, the carrier density rises from 4.14×$10^{19}$ $cm^{-3}$ to 5.51×$10^{20}$ $cm^{-3}$ and the resistivity changes from semiconducting (negative temperature coefficient) to metallic (positive coefficient above 50 K), while the ferromagnetic transition near 27.5 K persists in all films. The authors interpret this as nitrogen-vacancy doping: higher temperatures and lower N flux increase nitrogen vacancies, releasing electrons into Sm 5d and N 2p states that hybridize with Sm 4f states. No superconductivity is observed down to 0.35 K in any film, including the most conductive, which the authors contrast with earlier reports of superconductivity in polycrystalline SmN and attribute to lower carrier density, disorder sensitivity of triplet pairing, strain, or quantum confinement in thin films.","pith_inferences":["If the CrN cap is electrically passive, then the reported Hall carrier densities imply the metal-insulator crossover is intrinsic; a natural test is gating or annealing a single film to sweep carrier density continuously and check whether superconductivity emerges at higher density than the 5.5×10^20 cm^-3 reached here.","The strain-free growth on a highly mismatched substrate via an interfacial Sm-O-N or Sm-O layer suggests a route to integrate rare-earth nitride phases with transition-metal nitride platforms; one could probe whether the interfacial layer also contributes a parallel conductive channel, which would affect the interpretation of low-temperature resistivity.","The $T^2$ resistivity and negative magnetoresistance below $T_C$ are consistent with electron-magnon scattering in a ferromagnetic metal; measuring the anomalous Hall effect would test whether the orbital-dominated magnetic character reported elsewhere in SmN persists across this insulator-metal crossover.","The absence of superconductivity even at 0.35 K in the most conductive film suggests that if triplet pairing is intrinsic to SmN, the clean thin-film limit actually suppresses it — the opposite of the usual expectation that disorder suppresses unconventional pairing more than conventional pairing — a point worth checking in other rare-earth nitride systems."],"forward_implications":["Epitaxial SmN on MgO(001) can host both insulating and metallic ferromagnetic states depending on growth temperature, enabling monolithic integration of distinct electronic phases without strain from dissimilar lattice-matched layers.","Substrate temperature is a much stronger control than nitrogen flux for carrier density, so device-relevant doping can be set by one synthesis parameter while preserving crystallinity (FWHM around 0.39°).","The absence of superconductivity down to 0.35 K in high-quality films places an experimental constraint: any pairing mechanism must be suppressed by the combination of lower carrier density, thin-film disorder, strain, or confinement, or must be a property of polycrystalline inhomogeneity rather than intrinsic SmN.","Carrier concentration in these films is roughly an order of magnitude below the value reported for superconducting samples (~2×10^21 cm^-3), suggesting carrier density is the controlling parameter for superconductivity.","The ferromagnetic transition remains robust (~27.5 K) across the insulator-metal crossover, so the magnetic order does not require metallicity."],"supporting_citations":[{"why":"Supplies the MBE synthesis procedure on MgO(001) that this work varies.","marker":"[17]"},{"why":"The prior report of superconductivity in SmN that this work tests and does not reproduce.","marker":"[2]"},{"why":"Provides the carrier density (~2×10^21 cm^-3) of reported superconducting SmN films, the benchmark one order of magnitude higher than the present films.","marker":"[12]"},{"why":"Theoretical account of how nitrogen vacancies release electrons and bring Sm 4f states to the Fermi energy, the doping mechanism assumed here.","marker":"[27]"},{"why":"Establishes the spin/orbit moment imbalance and orbital-dominated ferromagnetism in SmN used to interpret the magnetic transition.","marker":"[10]"},{"why":"Prior determination of near-zero-moment ferromagnetism and Curie temperature in SmN, confirming the transition at roughly 27.5 K.","marker":"[26]"}],"fun_headline_variants":["Hotter growth turns SmN from insulator to ferromagnetic metal","SmN films: one heat knob raises carrier density tenfold, no superconductivity","SmN's electronic phase set by growth temperature, superconductivity absent","Insulator-metal switch in SmN by growth temperature alone","Growth heat tunes SmN to metallic, still no superconductivity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The measurements that show the insulator-to-metal crossover assume the 3 nm CrN capping layer is electrically passive; if CrN conducts in parallel, the reported resistivities and Hall carrier densities would not be intrinsic to SmN.","fun_headline_variants_meta":{"raw":{"variants":["Hotter growth turns SmN from insulator to ferromagnetic metal","SmN films: one heat knob raises carrier density tenfold, no superconductivity","SmN's electronic phase set by growth temperature, superconductivity absent","Insulator-metal switch in SmN by growth temperature alone","Growth heat tunes SmN to metallic, still no superconductivity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000639,"raw_usage":{"total_tokens":2917,"prompt_tokens":892,"completion_tokens":2025,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":508,"completion_tokens_details":{"reasoning_tokens":1935}},"tokens_in":508,"tokens_out":2025,"duration_ms":14453,"temperature":1.0,"reasoning_tokens":1935,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T19:06:44.806085+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the resistivity and Hall effect of a SmN film grown without the CrN cap (or of a CrN-only film on MgO under identical growth conditions). If the CrN-only film shows comparable conduction or if the uncapped SmN film does not reproduce the metallic crossover and carrier densities, the claim of an intrinsic tuning of SmN's electronic phase fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the MBE synthesis procedure on MgO(001) that this work varies."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the carrier density (~2×10^21 cm^-3) of reported superconducting SmN films, the benchmark one order of magnitude higher than the present films."},{"cited_title":"Holmes-Hewett, K","cited_arxiv_id":null,"evidence_quote":"Theoretical account of how nitrogen vacancies release electrons and bring Sm 4f states to the Fermi energy, the doping mechanism assumed here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the spin/orbit moment imbalance and orbital-dominated ferromagnetism in SmN used to interpret the magnetic transition."},{"cited_title":"Meyer, B","cited_arxiv_id":null,"evidence_quote":"Prior determination of near-zero-moment ferromagnetism and Curie temperature in SmN, confirming the transition at roughly 27.5 K."}],"review_version":1}