{"id":"eff6e6df-3725-496a-9542-3a0991870b4c","arxiv_id":"2607.27419","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Sputtered epitaxial Mn4N on MgO shows strong perpendicular magnetic anisotropy, and Mn4N/Pt bilayers show current-induced switching, supporting a scalable rare-earth-free spin-orbitronic platform.","lead":"Researchers grew thin films of the magnetic compound Mn4N by industry-friendly sputtering and mapped the growth conditions that give stable perpendicular magnetization. The result strengthens the case for a rare-earth-free, scalable material for low-power spintronic memory and logic.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Current-induced switching evidence is partial and thermally assisted; 'efficient' and 'strong spin transparency' are not supported (Fig. 7f, SI S5).","rationale":"The reader's stated weakest assumption concerns the strain-mediated PMA mechanism, but their rationale also lists 'overstatement of switching efficiency and interfacial spin transparency' as a weakness. I focus on the switching claim because it is the most load-bearing for the paper's advertised spintronic platform: if the switching is partial and thermally assisted, the conclusion that sputtered Mn4N/Pt enables efficient SOT control and has strong interfacial spin transparency collapses, even though the growth and PMA results may remain valid. The paper's own SI S5 provides direct evidence of partial switching and significant Joule heating, making this a concrete, internally documented gap rather than a speculative concern. A single quantitative torque-efficiency measurement would settle whether the spin transparency conclusion is warranted. The structural and magnetic characterisation otherwise appears solid, so the conditional verdict remains appropriate.","tokens_in":16720,"tokens_out":4365,"duration_ms":48959,"concrete_test":"Perform harmonic Hall measurements on the same Mn4N(30)/Pt(3) Hall bars at low AC current (≤1 mA, negligible Joule heating) and extract the damping-like spin-torque efficiency ξ_DL from the second-harmonic signal using the standard method (e.g., Ref. [52] methodology). If ξ_DL is comparable to Pt/ferromagnet references (~0.05–0.1), the 'strong spin transparency' claim is supported; if ξ_DL is below ~0.01 or the harmonic signal is dominated by thermoelectric artifacts, the current-induced switching is likely thermally assisted and the claim fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central spintronic claim — 'efficient current-induced magnetization switching in Mn4N/Pt bilayers, confirming strong interfacial spin transparency' (Abstract, Conclusions) — is not supported by the presented data. In Fig. 7(f), the current-induced Hall voltage step is far smaller than the field-driven AHE loop, and SI S5 explicitly states that 'the current induced switching is only partial' and that Joule heating 'plays an important role in facilitating it.' The authors estimate that the device temperature during pulses exceeds the calibrated 189°C and may approach ~380°C (upper bound), which reduces coercivity and AHE amplitude. Since the switched volume cannot be quantified and no torque-efficiency measurement (e.g., harmonic Hall or ST-FMR) is provided, the observed reversal could be dominated by thermal assist rather than efficient spin-torque transfer from the Pt interface. Therefore, the conclusion about 'strong interfacial spin transparency' is an overstatement; the data demonstrate only that some current-induced reversal occurs under strong heating, not that the interface has high spin transparency.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports a systematic optimization of reactive sputtering growth of Mn4N thin films on MgO(100) and SrTiO3(100), correlating deposition temperature, N2 partial pressure, and film thickness with structural, magnetic, and magnetotransport properties. The central growth claim—that epitaxial, single-crystalline Mn4N films with strong PMA can be obtained on MgO, while STO yields textured polycrystalline films—is well supported by XRD, STEM, SQUID, and AHE data. The authors further use DFT to argue that tensile tetragonal strain promotes PMA and that interface effects contribute, and they demonstrate current-induced Hall switching in Mn4N(30)/Pt(3) bilayers. However, the abstract and conclusions go beyond the evidence by calling the switching 'efficient' and by claiming 'strong interfacial spin transparency,' whereas the SI acknowledges only partial, Joule-heating-assisted switching. The strain interpretation is also weakened by the largely relaxed state of the 30 nm films.","tokens_in":16990,"tokens_out":6717,"duration_ms":71409,"significance":"If the growth and PMA results stand, the paper provides a valuable scalable route to rare-earth-free ferrimagnetic Mn4N for spin-orbitronic devices. The strengths are the systematic parameter map, the combination of structural/magnetic/transport probes, the first-principles DFT MAE calculations that are not fitted to the measured anisotropy, and the transparency of the SI, which explicitly reports the partial switching and heating limitations. The remaining gap is quantitative proof of SOT efficiency (or switched volume) and a direct link between the measured anisotropy and the DFT strain states. With those issues addressed, the manuscript would be a solid contribution.","major_comments":[{"comment":"The abstract and Conclusions state 'efficient current-induced magnetization switching' and 'strong interfacial spin transparency'. The data and the authors' own SI do not support these terms. SI S5 explicitly states that 'the current induced switching is only partial' and that 'Joule heating plays an important role in facilitating it'; the Hall steps in Fig. 7(f) are much smaller than the field-driven AHE loop, the two switching polarities have different amplitudes, and no torque-efficiency measurement (harmonic Hall or ST-FMR) or switched-volume quantification is provided. The paper can demonstrate deterministic, symmetry-correct current-induced reversal, but not efficiency or interfacial spin transparency. Please rephrase the claims (e.g., 'thermally assisted current-induced switching') and either add a torque-efficiency measurement or remove the efficiency/transparency statements from","section":"Abstract; Conclusions; Fig. 7(f); SI S5"},{"comment":"The DFT MAE calculation is presented as explaining the measured BK difference (≈700 mT on MgO vs ≈300 mT on STO) via epitaxial strain. However, the DFT uses bulk tetragonal cells coherently strained to the substrate lattice constants, whereas the actual 30 nm films are largely relaxed: Table S1 gives c≈3.86 Å and a≈3.8 Å for MgO and c≈3.8 Å, a≈3.8 Å for STO, all within ~1% of bulk, and STEM shows misfit dislocations on MgO and textured polycrystalline growth on STO. Thus the MgO-like and STO-like strain states in Fig. 6(c) are not the film states. The measured anisotropy difference could therefore be dominated by microstructure, defects, interface quality, or off-stoichiometry (the latter acknowledged as difficult to exclude in SI S1). To make the strain claim load-bearing, the authors should either relate MAE to experimentally measured strain states (e.g., the thickness series where XRD","section":"Section III.D; Fig. 6(c); Table S1"},{"comment":"The reported critical current density jC ≈ 1.8×10^7 A/cm^2 is quoted for a process that is explicitly only partial switching. Because the switched volume is not determined, this value is not a material figure of merit for efficient SOT switching; it is an upper bound on the current needed to initiate reversal under strong Joule heating. The same issue affects the comparison with previous Mn4N/Ta bilayers in Ref. [24]: the sign comparison is meaningful, but the efficiency claim requires a measurement that separates the spin-torque contribution from heating. Please clarify this limitation in the main text, not only in the SI.","section":"Section III.E; SI S5"}],"minor_comments":[{"comment":"The unit '2.9 µΩ/cm' should be '2.9 µΩ·cm' (micro-ohm centimeter), as used elsewhere for resistivity.","section":"Section II (Methods)"},{"comment":"The phrase 'tensile strain of approximately −8%' is confusing; a negative value usually denotes compressive strain. Please state the sign convention explicitly or use a magnitude with a clear direction.","section":"Section III.A"},{"comment":"The background color map representing 'relative elastic energy cost' has no colorbar or scale. Without a quantitative axis, the statement that the substrate-relevant strain states remain 'energetically accessible' is not verifiable from the figure.","section":"Fig. 6(c)"},{"comment":"The inference of spin Hall magnetoresistance from the Pt layer is based on a finite MR during the β rotation. This contribution is not unambiguously separated from other MR mechanisms; please clarify the assumption or provide supporting data.","section":"Section III.E"},{"comment":"Several typographical issues: 'AHE loop measured at that highest possible temperature reached by our setup' has 'ne' for 'one'; Fig. S5(c) label 'up tp 45 mA' should be 'up to'; and in Fig. S5(b) the field axis is labeled HZ while the text uses BZ. Please harmonize notation.","section":"SI S5 / Fig. S5"},{"comment":"The caption contains 'M−Bloops' without a space; it should read 'M−B loops'.","section":"Fig. 4 caption"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a solid systematic growth study and I would not reject it. The main obstacle is the overstatement in the abstract and conclusions about efficient switching and strong interfacial spin transparency, which is contradicted by the authors' own SI. The strain-PMA link is also weakened because the 30 nm films are largely relaxed. Both issues are fixable with tempered claims and, ideally, a torque-efficiency measurement. The growth and PMA results themselves are convincing and valuable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things. First, the growth study is real: a systematic map of temperature, N2 ratio, and thickness for sputtered Mn4N on MgO, with epitaxy and PMA documented by XRD, STEM, SQUID, and AHE. Second, the switching claim is softer than the abstract suggests—the SI admits partial switching and Joule heating, but the abstract and conclusions still say 'efficient' and 'strong interfacial spin transparency.' That gap is the main problem.\n\nWhat's genuinely new: the deposition-parameter sweep, the clear epitaxial-vs-textured distinction between MgO and STO, the BK values extracted from angular Hall fits, and the first current-induced switching in sputtered Mn4N/Pt layers. The DFT is first-principles and not fitted to the measured BK; the slab calculations in the SI show they thought about finite-size effects. Credit where due: the structural, magnetic, and transport data are cross-checked and consistent, and the paper is clearly written.\n\nThe soft spots are real but not fatal. The switching evidence in Fig. 7(f) and SI S5 shows partial reversal, with estimated device temperatures above the 189°C calibration and possibly up to ~380°C. The switched volume is not quantified, and there is no torque measurement (harmonic Hall or ST-FMR), so 'efficient' and 'strong spin transparency' are not supported. The authors themselves describe the switching as partial and thermally assisted, so this is an overstatement in the summary rather than a hidden flaw. Second, the strain mechanism: the DFT uses idealized strained bulk to explain anisotropy in films that STEM shows are largely relaxed (c ≈ 3.86 Å, a ≈ 3.8 Å), and the STO films are textured polycrystalline. The authors hedge ('not the sole contributor'), but the strain interpretation remains qualitative. Third, the Pt-vs-V interface comparison relies on a sample from a prior paper rather than a controlled set in this study—minor but worth noting.\n\nThese are addressable. The growth and PMA results are solid and useful on their own. This paper is for anyone working on Mn4N or ferrimagnetic SOT materials. It deserves a serious referee; the growth section is publishable, and the switching result, if reframed as partial and thermally assisted, is a useful data point. I would send it to peer review with a request for revision—mainly to tone down the abstract and add either a torque measurement or a careful discussion of the switched fraction. As written, the load-bearing spintronic claim is overreaching, but the underlying work is credible.","headline":"Solid growth-optimization work on sputtered Mn4N, but the abstract overclaims 'efficient switching' and 'strong spin transparency' when the authors' own SI shows partial, thermally assisted reversal.","tokens_in":17493,"tokens_out":1904,"would_cite":true,"duration_ms":21433,"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":"Reactive sputtering produces epitaxial Mn4N films with strong perpendicular anisotropy and efficient spin-orbit torque switching, establishing a scalable rare-earth-free platform.","keywords":["Mn4N","ferrimagnets","perpendicular magnetic anisotropy","spin-orbit torques","anomalous Hall effect","magnetotransport","reactive magnetron sputtering","rare-earth-free spintronics"],"falsifier":"Grow a series of Mn4N films on MgO with systematically varied in-plane strain (for example, by thickness or buffer layers) while keeping the Pt interface and microstructure fixed, and measure the perpendicular anisotropy field; if BK does not increase with increasing in-plane strain, or if a fully relaxed film still shows BK≈700 mT, the strain-driven anisotropy trend would fail.","tokens_in":16633,"feed_emoji":"🧲","tokens_out":4380,"duration_ms":51282,"temperature":0.7,"pith_summary":"The paper tries to show that reactive magnetron sputtering, an industrially compatible deposition method, can produce high-quality Mn4N films for spintronics. It claims that epitaxial, single-crystalline Mn4N with strong perpendicular magnetic anisotropy forms on MgO(100) under optimized sputtering conditions, while SrTiO3(100) yields textured polycrystalline films. The authors argue that epitaxial strain tunes the magnetic anisotropy but is not the whole story: PMA persists even in largely relaxed films, and interfacial effects, especially with Pt capping, also stabilize the perpendicular easy axis. They demonstrate deterministic current-induced magnetization switching in Mn4N(30)/Pt(3) bilayers, which they take as evidence of strong interfacial spin transparency. A sympathetic reader would care because this points to a rare-earth-free, scalable material platform for energy-efficient spin-orbitronic devices.","feed_headline":"Sputtered rare-earth-free Mn4N switches with current","feed_subtitle":"Epitaxial films on MgO show strong perpendicular anisotropy and efficient current-induced spin-orbit torque switching.","key_machinery":"The central object is Mn4N, a ferrimagnetic antiperovskite in which nitrogen sits at the body center of a face-centered cubic Mn lattice and the two Mn sublattices align antiparallel, giving a low net magnetization and a strong anomalous Hall response. The growth control relies on reactive sputtering parameters, especially substrate temperature and Ar:N2 ratio, which suppress the competing α-Mn phase and stabilize the (100) orientation. The quantitative anisotropy argument uses DFT-computed magnetic anisotropy energy as a function of tetragonal a/c ratio: MgO-matching tensile strain gives a larger out-of-plane MAE than STO-matching strain, and the same trend appears in finite-slab calculatio","core_discovery":"The paper's central claim is that Mn4N, a ferrimagnetic antiperovskite with low net magnetization and a large anomalous Hall response, can be grown by reactive sputtering into device-grade films. On MgO(100), 30-nm films are epitaxial and single-crystalline with square hysteresis loops, high remanence, tunable coercivity, and an anomalous Hall resistance of about 40 mΩ. On SrTiO3, growth is textured and polycrystalline, yet still ferrimagnetic with perpendicular anisotropy, though weaker. Combining transport measurements with density functional theory, the authors argue that tensile epitaxial strain, especially the MgO-matching strain state, raises the magnetic anisotropy energy and favors o","pith_inferences":["Beyond the paper: if the interfacial spin-orbit contribution is as strong as indicated, then capping-layer engineering (e.g., Pt alloys or heavy-metal/oxide stacks) could push the anisotropy field higher and lower the switching current, potentially extending PMA to thinner films.","Beyond the paper: the MgO-versus-STO anisotropy comparison is partly confounded by microstructure, since STO films are textured polycrystalline rather than epitaxial; a cleaner test would compare epitaxial films on substrates with matched interface chemistry but different lattice constants.","Beyond the paper: the reported current density is an upper bound for partial, thermally assisted switching, not a clean measure of intrinsic spin-orbit torque efficiency; pulse-width and temperature-dependent measurements would separate thermal effects from torque-driven reversal.","Beyond the paper: explicit DFT models of the Pt/Mn4N interface, including the heavy metal, would likely quantify an additional interfacial PMA contribution that the current bulk- and slab-only calculations leave implicit."],"forward_implications":["Epitaxial Mn4N with strong PMA can be grown by sputtering, a scalable and industrially compatible method, removing the need for molecular beam epitaxy in this material class.","PMA persists in films thicker than the fully strained limit, so device stacks do not depend on maintaining extreme epitaxial strain.","The anisotropy field can be tuned by substrate and capping layer: MgO gives BK≈700 mT versus ≈300 mT on STO, and Pt capping gives a larger BK than V capping.","The large anomalous Hall response (~40 mΩ) is sufficient for electrical readout of the magnetization state in sputtered Mn4N.","Deterministic SOT switching works in a 30-nm-thick ferrimagnetic layer with Pt, and the switching polarity reverses with in-plane field, matching the symmetry of damping-like SOT.","The opposite switching polarity relative to Ta capping is consistent with the opposite spin Hall angle of Pt and Ta, indicating standard SOT physics in this system."],"fun_headline_variants":["Sputtered Mn4N flips spins with current","Rare-earth-free Mn4N: current-driven spin switching","Epitaxial Mn4N: perpendicular spins, SOT switching","Current writes magnetization in sputtered Mn4N"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the DFT-computed strain states for idealized, perfectly tetragonally strained bulk Mn4N represent the anisotropy of the real films, even though STEM shows the 30-nm films are largely relaxed (a≈c≈3.8 Å) and the STO films are polycrystalline.","fun_headline_variants_meta":{"raw":{"variants":["Sputtered Mn4N flips spins with current","Rare-earth-free Mn4N: current-driven spin switching","Epitaxial Mn4N: perpendicular spins, SOT switching","Current writes magnetization in sputtered Mn4N"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001522,"raw_usage":{"total_tokens":5961,"prompt_tokens":798,"completion_tokens":5163,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":542,"completion_tokens_details":{"reasoning_tokens":5095}},"tokens_in":542,"tokens_out":5163,"duration_ms":33717,"temperature":1.0,"reasoning_tokens":5095,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T07:36:07.745484+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Grow a series of Mn4N films on MgO with systematically varied in-plane strain (for example, by thickness or buffer layers) while keeping the Pt interface and microstructure fixed, and measure the perpendicular anisotropy field; if BK does not increase with increasing in-plane strain, or if a fully relaxed film still shows BK≈700 mT, the strain-driven anisotropy trend would fail.","supporting_citations":[],"review_version":1}