{"id":"a9991d9c-3c9f-4013-9eb2-e077bff432fb","arxiv_id":"2607.29472","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"In GdxSm1−xN, the coercive field varies by orders of magnitude with Gd content while the calculated exchange splitting changes only ~20%, and nitrogen vacancies are strongly favored at Sm-rich sites.","lead":"This paper studies how mixing gadolinium and samarium in rare-earth nitride films changes their magnetic behavior and electronic structure. It finds that the coercive field swings by orders of magnitude while the exchange field shifts only slightly, and that nitrogen vacancies form far more easily next to samarium — both properties matter for superconducting spintronics devices.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Table I confounds local vacancy coordination with global Sm concentration, so the 'adjacent to Sm' vacancy-formation claim is not isolated.","rationale":"The paper is a valuable combined experimental/computational study: the coercivity variation is experimentally measured, and the DFT exchange splittings are internally plausible. I credit the random-alloy supercell approach and the ML/MS comparisons as partial evidence. The reader's concern about unstated Hubbard U parameters is real, but it would mainly shift quantitative values; the more load-bearing issue is methodological: Table I varies both global composition and local coordination simultaneously, so the central vacancy-formation claim is confounded. This is a correctable gap rather than a fatal flaw, and the paper should remain CONDITIONAL pending the fixed-composition test. I also note a likely typo in Section III C ('with x=0 the coercive field is ~100 Oe' while x=0 is SmN, which should have a very large coercive field), but this is not load-bearing.","tokens_in":11626,"tokens_out":8877,"duration_ms":92377,"concrete_test":"Using a fixed 54-atom (Gd0.5Sm0.5)N supercell with the same cation arrangement and only the vacancy site moved, compute Ef for coordinations (6,0), (5,1), (4,2), (3,3), (2,4), (1,5), and (0,6), with full relaxation and at least two independent random cation arrangements. If Ef does not decrease monotonically as the number of Sm neighbours around the vacancy increases at fixed x, the 'adjacent to Sm' claim in the abstract and Fig. 5 is unsupported by the data as presented.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section III D explicitly states that each vacancy coordination was 'chosen to reflect the concentration of the host crystal,' so Table I changes two variables at once: the local Gd/Sm coordination and the global x of the supercell. The (6,0) row is GdN, the (0,6) row is SmN, and intermediate rows use different alloys. The monotonic decrease in Ef across Table I could therefore be driven by composition-dependent cohesive energy, strain, or band filling rather than by the identity of the six nearest cations. The only same-x comparisons are the ML/MS pairs, and they do not show a consistent preference: at x=0.7, MS is favored by 0.81 eV; at x=0.48, ML is favored by 0.06 eV; at x=0.33, MS is favored by 0.08 eV. Consequently, the abstract's claim that formation energy is 'significantly reduced for vacancy sites adjacent to Sm ions rather than Gd ions' is not isolated by the data. The Fig. 5 statistical weighting inherits this confound, so its prediction that essentially all vacancies sit at six-Sm sites is not established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper combines experimental optical spectroscopy and magnetometry with DFT+U band-structure and defect calculations for GdxSm1−xN. The authors report that the measured coercive field varies over orders of magnitude with composition while the calculated conduction-band exchange splitting changes by only ~20%, and they attribute a growing mid-infrared absorption in Sm-rich films to nitrogen vacancies whose formation energy is calculated to be lower near Sm ions. The intended application is superconducting spintronics, where hard/soft magnetic pairs with matched exchange fields would be useful.","tokens_in":11885,"tokens_out":6013,"duration_ms":57264,"significance":"The paper addresses a timely and relevant materials-engineering question: whether rare-earth nitride alloys can provide hard/soft magnetic pairs with similar exchange fields for superconducting spintronics. The experimental coercive-field data and the qualitative band-structure trend across the alloy series are valuable, and the defect-formation question is important for thin-film quality and transport. If the quantitative claims were fully supported, this would be a useful contribution to the field. However, the central quantitative claims rest on DFT+U calculations with unreported and experimentally fitted Hubbard parameters, an unreconciled discrepancy between the calculated and measured GdN gap, and a defect table that does not isolate local coordination effects from global composition effects. These issues need to be addressed before the quantitative conclusions can be considered robust.","major_comments":[{"comment":"The Hubbard parameters U_f and U_d are not reported; they are taken from refs. [41,42] and were 'guided by recourse to experimental results'. Since the exchange splitting and defect formation energies are DFT+U outputs, the key numerical results (e.g., E_X = 0.65 eV for GdN, Table I) are not reproducible without these values. Moreover, the calculated GdN direct gap of ~0.9 eV (§III B) is substantially smaller than the measured optical gap of the GdN film, ~1.6 eV (§III A), with no reconciliation. If the 5d band positions are not captured quantitatively, the ~20% exchange-splitting variation across the alloy and the defect energetics may not be quantitatively reliable.","section":"§II B and §III B"},{"comment":"The abstract's claim that the vacancy formation energy is significantly reduced at Sm-coordinated sites is not isolated by the data. As stated in §III D, the coordination was 'chosen to reflect the concentration of the host crystal', so local coordination and global composition change together. The only fixed-composition comparisons are the ML/MS pairs, which show no consistent Sm-coordination preference: (4,2) MS is lower by 0.81 eV at x=0.7, (3,3) ML is lower by only 0.06 eV at x=0.48, and (2,4) MS is lower by only 0.08 eV at x=0.33. The across-composition trend is also non-monotonic, with (2,4) energies exceeding those of (3,3). Figure 5 inherits this confound. To support the local-coordination claim, calculations are needed at fixed global composition with different nearest-neighbour coordinations, or an analysis that separates composition and coordination effects.","section":"§III D and Table I"},{"comment":"The sentence giving example values appears internally inconsistent: 'with x=0 the coercive field is ~100 Oe and the exchange splitting is 0.65 eV, while for x=0.8 the coercive field is ~5000 Oe and the exchange splitting is 0.6 eV'. This conflicts with the paper's own earlier statement that SmN (x=0) has a coercive field >10 T at 2 K and with the trend described for Figure 3. Please check the x values, axis labels, or units in this sentence and in Figure 3.","section":"§III C"}],"minor_comments":[{"comment":"Typo: 'nitrogen vaccines' should read 'nitrogen vacancies'.","section":"§III D"},{"comment":"The exchange splitting is a calculated quantity, not a directly measured one. The abstract and §III C phrase the ~20% change as a finding; please state explicitly that the exchange-field trend is calculated, to avoid implying direct experimental determination.","section":"§III B / §III C"},{"comment":"Coercive field (Oe) and exchange splitting (eV) have different units and likely different scales. Please label the two axes clearly and state that the two quantities are not directly comparable, to avoid visually implying a causal or correlational relation.","section":"Figure 3"},{"comment":"The composition of the optical films was estimated from beam-equivalent pressures rather than from X-ray fluorescence (which was used for the magnetic films). Please state the estimated uncertainty in composition for the optical films.","section":"§II A"},{"comment":"The 'anomalously low' formation energy for the (4,2) MS configuration may reflect a finite-size or strain artifact of the specific supercell. A convergence check with a larger supercell or an explicit discussion of this possibility would strengthen the defect analysis.","section":"Table I"},{"comment":"Reference [26] is an arXiv preprint; if a peer-reviewed version has appeared, it should be cited instead of or in addition to the arXiv version.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of Physical Review Materials and contains useful experimental data, but the quantitative central claims need further support. The computational parameterization and the confounded defect-formation comparison are load-bearing issues; I would like to see them addressed before publication. The paper is not suitable for acceptance in its present form."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Punchline: this paper gives you a useful experimental baseline — coercivity in GdxSm1−xN sweeps orders of magnitude between SmN and GdN while the DFT+U exchange splitting changes by only ~20% — and a plausible qualitative story for why nitrogen vacancies are more abundant in Sm-rich films. The experimental core is solid: optical transmission/reflection on five compositions, SQUID magnetometry, and a clean mid-infrared signature tied to vacancies. The band-structure series across x is new and the 5d hybridization argument is convincing. Credit where due: this is careful, clearly written work from a group that knows this material class.\n\nThe soft spots are in the quantitative claims. The stress-test note is correct: Table I confounds local vacancy coordination with global Sm concentration. The (6,0) row is GdN, (0,6) is SmN, and the intermediate rows use different alloy compositions, so the monotonic decrease in formation energy is not isolating nearest-neighbor identity. The only same-x comparisons (ML vs MS) do not show a consistent preference — at x=0.48, ML wins by 0.06 eV, at x=0.33 MS wins by 0.08 eV. That doesn't support the abstract's wording that formation energy is 'significantly reduced for vacancy sites adjacent to Sm ions.' The Fig. 5 weighting inherits the confound. So the defect-formation trend is plausible but not isolated.\n\nSecond, the Hubbard U values are not stated. They're taken from earlier papers and fitted to experimental results, so the reported exchange splittings and vacancy energetics are not parameter-free predictions. The paper itself notes a gap between the calculated GdN direct gap (0.9 eV) and the measured optical gap (~1.6 eV) without reconciling them. For the qualitative ~20% trend this is probably fine; for absolute numbers it's a real caveat. I'd also flag the data-availability line ('available upon reasonable request') — for a computational study, not shipping inputs and parameters is a missed opportunity.\n\nOverall: the experimental part deserves serious attention, and the exchange-splitting result is a useful design input for superconducting spintronics. The defect-formation analysis needs reworking before the 'preferentially adjacent to Sm' claim is taken at face value. I would not cite the numbers, but I would read the paper and recommend it to someone working on LnN alloys. A serious referee should engage with it.","headline":"A genuinely useful combined study of GdxSm1−xN, but the vacancy-formation claim is confounded and the DFT numbers are less certain than the abstract suggests.","tokens_in":12386,"tokens_out":2370,"would_cite":false,"duration_ms":25588,"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":"This paper shows that in the ferromagnetic semiconductor alloy GdxSm1−xN, replacing gadolinium with samarium changes the coercive field by orders of magnitude while the internal exchange field changes by only about 20%, and that nitrogen va","keywords":["rare-earth nitrides","GdSmN alloy","exchange splitting","coercive field","nitrogen vacancies","DFT+U","superconducting spintronics","ferromagnetic semiconductor"],"falsifier":"Measure, for a series of GdxSm1−xN films of known composition, the optical gap in the ferromagnetic state and the coercive field at 5 K. If the inferred exchange splitting (from the spin-split conduction band) changes by much more than 20% while the coercive field changes by orders of magnitude, the decoupling claim fails. For the vacancy claim, use X-ray absorption fine structure or positron annihilation to map the local coordination of nitrogen vacancies; if vacancies in a Sm-rich film are not predominantly six-Sm-coordinated, the formation-energy prediction is wrong.","tokens_in":11511,"feed_emoji":"🧲","tokens_out":3184,"duration_ms":34581,"temperature":0.7,"pith_summary":"The paper argues that the magnetic hardness of GdxSm1−xN—measured by the coercive field, or the magnetic field needed to reverse the magnetization—can be tuned across orders of magnitude simply by adjusting the ratio of gadolinium to samarium, while the internal exchange field, the energy splitting between spin-up and spin-down conduction states, remains nearly constant, shifting by only about 20%. It further argues that nitrogen vacancies, which degrade these films, form much more readily when surrounded by samarium ions than by gadolinium ions, explaining why samarium-rich films show more vacancy defects. This combination of tunable coercivity with a nearly stable exchange field is precisely what is needed for superconducting spintronics devices, where matching the exchange field across magnetic layers is critical.","feed_headline":"Coercive field swings 1000x while exchange field shifts 20%","feed_subtitle":"GdSmN alloys decouple magnetic hardness from the internal exchange field, opening a route to tunable superconducting spintronics.","key_machinery":"The central object is the 5d conduction-band exchange splitting, EX, the half-difference between the majority- and minority-spin 5d band minima at the X point of the rocksalt Brillouin zone. Band-structure calculations using DFT+U with Hubbard parameters for the 4f and 5d states trace how EX evolves with alloy composition, and the ratio of Sm to Gd spin quantum numbers (5/7) serves as a simple parameter-free estimate for the splitting ratio. For the vacancy energetics, the paper computes formation energies from total-energy differences of pristine and defective supercells, with the key variable being the coordination of the vacancy—how many of its nearest neighbors are Sm versus Gd.","core_discovery":"The central claim is that cation substitution in GdxSm1−xN provides independent control of two magnetic quantities: the coercive field spans several orders of magnitude from GdN to SmN, while the 5d conduction-band exchange splitting varies only from 0.65 eV in GdN to 0.5 eV in SmN, a change of about 20%. This decoupling arises because the exchange splitting is governed by the spin polarization of the 4f states and their hybridization with the 5d bands—a first approximation based on the ratio of spin quantum numbers S_Sm/S_Gd = 5/7 ≈ 0.71, close to the computed splitting ratio of 0.77—whereas the coercive field is controlled by the net magnetization, which nearly vanishes in SmN due to spin-","pith_inferences":["If the near-constancy of the exchange field survives at device-relevant thicknesses and temperatures, GdxSm1−xN layers could be stacked to produce a nearly constant spin-splitting profile while varying coercivity, simplifying the design of superconducting spin valves.","The strong preference of vacancies for Sm coordination implies a possible self-amplifying defect pattern: a vacancy depletes local Gd and leaves a Sm-rich neighborhood, which lowers the formation energy of subsequent vacancies, potentially leading to vacancy clustering or void formation in Sm-rich films.","A direct test of the vacancy claim would be to map the local atomic environment of nitrogen vacancies—for example, with X-ray absorption fine structure or positron annihilation—in films of controlled composition, and check whether the vacancy signal indeed tracks Sm coordination as predicted.","The simple spin-ratio argument (5/7) suggests that other rare-earth nitride alloy pairs, chosen for different 4f spin quantum numbers, could be used to dial the exchange splitting across a wider range; the same computational approach could predict those series."],"forward_implications":["A hard/soft magnetic pair with nearly equal exchange splitting can be made within one alloy system, enabling switchable 0-π Josephson junctions where the relative alignment of layers selects the superconducting phase.","SmN, with its finite exchange splitting but near-zero net moment, could be used in a magnetic Josephson junction with strongly reduced fringe fields.","Nitrogen vacancies will predominantly occupy Sm-coordinated sites, so electron transport in Gd-rich films may come to resemble the localized-state behavior of SmN rather than the extended-state behavior of GdN.","The low formation energy of Sm-coordinated vacancies makes substrate temperature during growth a powerful control knob, as demonstrated by the reduced mid-infrared absorption in SmN grown without passive heating.","The roughly 20% cohesive-energy difference between GdN and SmN may bias the composition of grown films, so stoichiometry control may need to account for this thermodynamic preference."],"fun_headline_variants":["GdSmN alloys: Coercivity swings 1000x, exchange field shifts 20%","Decoupled magnetic control in GdSmN via cation substitution","Nitrogen vacancies form near Sm, tune GdSmN magnetism","Exchange field varies only 20% as coercivity changes 1000x","GdSmN: Rare earth mix tunes magnetic hardness independently"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The quantitative exchange-splitting and vacancy-formation energies come from DFT+U calculations whose Hubbard U parameters for the 4f and 5d electrons are fitted to experimental data on the end members rather than derived from first principles; if these parameters do not transfer faithfully to the alloy, the exact ~20% variation and the absolute defect energies could be off, though the qualitative trend of Sm-favored vacancies may survive.","fun_headline_variants_meta":{"raw":{"variants":["GdSmN alloys: Coercivity swings 1000x, exchange field shifts 20%","Decoupled magnetic control in GdSmN via cation substitution","Nitrogen vacancies form near Sm, tune GdSmN magnetism","Exchange field varies only 20% as coercivity changes 1000x","GdSmN: Rare earth mix tunes magnetic hardness independently"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000576,"raw_usage":{"total_tokens":2561,"prompt_tokens":759,"completion_tokens":1802,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":503,"completion_tokens_details":{"reasoning_tokens":1705}},"tokens_in":503,"tokens_out":1802,"duration_ms":13673,"temperature":1.0,"reasoning_tokens":1705,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T06:25:25.460906+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure, for a series of GdxSm1−xN films of known composition, the optical gap in the ferromagnetic state and the coercive field at 5 K. If the inferred exchange splitting (from the spin-split conduction band) changes by much more than 20% while the coercive field changes by orders of magnitude, the decoupling claim fails. For the vacancy claim, use X-ray absorption fine structure or positron annihilation to map the local coordination of nitrogen vacancies; if vacancies in a Sm-rich film are not predominantly six-Sm-coordinated, the formation-energy prediction is wrong.","supporting_citations":[],"review_version":1}