{"id":"acaef03b-ef71-406b-856f-f6e24890f5e3","arxiv_id":"2607.27306","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Native oxidation of Cr in NiFe/Cr bilayers creates an interfacial orbital-current source yielding a giant damping-like torque efficiency of 3.9 × 10^6 Ω^-1 m^-1 and field-free switching at 1.58 × 10^11 A/m^2.","lead":"A nickel-iron/chromium bilayer left to oxidize in air produces an unusually strong current-driven torque on its magnetization, reportedly outperforming platinum and tantalum without any heavy-metal layer. The authors attribute the effect to the native chromium-oxide interface, which acts as an orbital-current generator and enables field-free switching.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central interfacial-dominance decomposition rests on the ad hoc tanh activation function in Eq. (4); without a microscopic derivation, the reported ξ_I/ξ_B ≈ 13 and λ ≈ 4 nm are not independently established.","rationale":"The reader's weakest_assumption correctly identifies Eq. (4)'s tanh activation function as the linchpin of the quantitative decomposition. The paper's central claim is not merely that oxidation matters—that part is strongly supported by the Ta-cap and Cu-spacer controls and by the OUMR correlation—but that the interfacial OREE channel is ~13× stronger than the bulk OHE with λ≈4 nm. This quantitative hierarchy is produced by a four-parameter fit (ξ_B, ξ_I, λ, plus the functional form of g) to a single thickness series. The functional form of g is chosen after the fact to produce the desired non-monotonic shape. A different physically motivated g could easily shift the bulk/interfacial ratio while still fitting the data. Therefore the load-bearing concern is not about fraud or overreach but about model identifiability: the data do not constrain the model enough to support the specific quantitative decomposition. The reader's recommendation of CONDITIONAL is appropriate, and no verdict adjustment is needed; the authors should provide a microscopic derivation of g or an independent test of the decomposition. I agree with the reader's emphasis, and the suggested refitting with the measured oxygen profile is a concrete, feasible check that would either validate or falsify the extracted interfacial dominance.","tokens_in":543,"tokens_out":2486,"duration_ms":64750,"concrete_test":"Refit the ξ_DL vs Cr* thickness data using an activation function derived from the measured XPS oxygen concentration profile instead of tanh(d/λ). For example, set g(d)=∫_0^d c(x) q(d−x) dx, where c(x) is the measured oxygen concentration and q is a local response function (e.g., the orbital current generated by oxidized Cr at position x, weighted by transmission to the interface). If the resulting fitted ξ_I/ξ_B shifts by more than 30% from the current value, or if the non-monotonic peak can be reproduced with a much smaller interfacial term, then the reported 13-fold interfacial dominance is an artifact of the chosen g(t).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's key quantitative claim—that the Cr/CrOx interface generates orbital currents more than an order of magnitude stronger than the bulk OHE—is obtained solely by fitting Eq. (4). That equation multiplies the interfacial source by g(t)=tanh(d/λ), a function introduced by hand as 'diffusion-alike' without microscopic justification. The authors correctly show that a constant two-channel model (Eq. 3) is monotonic, but they then replace ξ_I with ξ_I tanh(d/λ) to force an interior maximum. This activation form is not derived from the oxidation chemistry, the boundary conditions, or the drift-diffusion equations; it is one of infinitely many smooth increasing functions that would produce a peak when multiplied by sech(d/λ). Consequently, the extracted parameters—especially ξ_I ≈ 3.7×10^6 Ω⁻¹m⁻¹, ξ_B ≈ 0.29×10^6 Ω⁻¹m⁻¹, and λ ≈ 4.1 nm—are not uniquely determined by the thickness data. The fit is consistent with the data but does not test the mechanism. In addition, room-temperature antiferromagnetic order in Cr (bulk TN ≈ 311 K) is never addressed; if even a fraction of the Cr layers is antiferromagnetically ordered, it could contribute its own spin or orbital transport channels and complicate the torque analysis. Together these issues mean the headline interfacial-vs-bulk decomposition is currently unsupported, even though the control experiments strongly support an oxide-interface role.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports that naturally oxidized NiFe/Cr bilayers act as a self-contained orbital-current source, with a damping-like torque efficiency of (3.9 ± 0.7) × 10^6 Ω⁻¹m⁻¹ at Cr* = 8 nm, non-monotonic thickness dependence, and field-free magnetization switching at 1.58 × 10^11 A/m². First-principles calculations predict an approximately threefold enhancement of the orbital Hall conductivity upon surface oxygenation. A drift-diffusion model with a bulk orbital Hall effect and an oxidation-activated interfacial orbital Rashba–Edelstein source is used to decompose the torque into bulk and interfacial contributions. Two control experiments (Ta capping and Cu spacer) and OUMR correlations are presented to support the oxide-interface mechanism. The authors conclude that the Cr/CrOx interface dominates over the bulk orbital Hall channel by roughly an order of magnitude, with an orbital transport length of ≈4 nm.","tokens_in":18410,"tokens_out":7573,"duration_ms":62704,"significance":"If the conclusions are correct, the work is significant: it would establish native oxidation of a 3d metal as a scalable, heavy-metal-free orbital-torque source with record efficiency and field-free switching. The paper has genuine strengths: two control experiments separating source and conversion, an OUMR–torque correlation across the thickness series, and first-principles support for oxygen-enhanced orbital Hall conductivity. However, the quantitative interfacial-dominance claim rests on an ad hoc activation function, and the reported fit parameters are numerically inconsistent with the headline data. These issues are load-bearing for the central claim and must be resolved before the quantitative message can be accepted.","major_comments":[{"comment":"The quoted fit parameters are numerically impossible. Since tanh(x)sech(x) ≤ 0.5 for all x, the maximum of ξ_DL^E(t) = ξ_B[1 − sech(d/λ)] + ξ_I tanh(d/λ)sech(d/λ) is at most ξ_B + 0.5 ξ_I. With ξ_B = 0.29×10^6 and ξ_I = 3.7×10^6 Ω⁻¹m⁻¹, this upper bound is ≈2.14×10^6 Ω⁻¹m⁻¹, yet the data peak is (3.9 ± 0.7)×10^6 Ω⁻¹m⁻¹. Thus Eq. (4) with the stated parameters cannot reproduce the reported peak. The values of ξ_I, ξ_B, and λ — and the derived ξ_I/ξ_B ≈ 13 — need to be re-evaluated and the fit re-presented.","section":"§2 around Eq. (4)"},{"comment":"The activation function is introduced ad hoc as 'diffusion-alike' but is not derived from the oxidation kinetics, boundary conditions, or drift-diffusion equations. Because this functional form is chosen specifically to produce an interior maximum, the extracted interfacial dominance is not an independent test of the mechanism. The authors should either derive g(d) from a microscopic model (e.g., the XPS oxidation profile) or present model-selection/uncertainty analysis against other source-growth functions (step, linear, error-function). The same concern applies to the FL-based λ: Eq. (2) assumes a single bulk source, whereas the authors' own model allows an interfacial contribution to the FL torque.","section":"§2, Eq. (2) and Eq. (4)"},{"comment":"Bulk Cr is an antiferromagnet below ≈311 K, yet the manuscript nowhere discusses whether the Cr layers in these polycrystalline films are magnetically ordered at room temperature, nor how possible AFM order or proximity effects at the NiFe/Cr interface would affect the harmonic-Hall analysis, the drift-diffusion model, or the DFT calculation of the orbital Hall conductivity. The authors should address this with measurements (e.g., exchange-bias field, M(T), or a thin-film TN suppression argument) or explicitly justify treating Cr as nonmagnetic in the model.","section":"§2 and DFT sections"}],"minor_comments":[{"comment":"Ref. [34] has a malformed DOI ('10.1103/qgdy-k39l') and Ref. [48] lacks volume/page details; both need correction.","section":"References"},{"comment":"'NiFe\\Cr' in the abstract should be 'NiFe/Cr' for consistency.","section":"Abstract"},{"comment":"Equation (2) is presented without a prefactor; define the full expression or state that it represents the functional form only.","section":"Eq. (2)"},{"comment":"The phrase 'At sufficiently large magnetic fields which suppresses the magnon contribution' is grammatically awkward; rephrase.","section":"OUMR section"},{"comment":"The field-free switching is attributed to an in-plane anisotropy tilt, but the origin of the tilt is not explained; a sentence or reference clarifying this would be helpful.","section":"Switching section"}],"recommendation":"major_revision","confidential_remarks":"The numerical inconsistency in the reported fit parameters (Eq. (4) max bound vs. data peak) is the most serious issue; it must be corrected before the quantitative claim can be evaluated. If the model is retained, the authors should also address the ad hoc nature of the tanh activation and the Cr antiferromagnetism. The experimental controls and OUMR correlation are valuable and strongly support an oxide-related mechanism, so this is not a reject if the quantitative analysis is fixed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe headline: this is a serious experimental paper with a clever dual-channel idea and excellent controls, but the quantitative claim that the interfacial Cr/CrOx source is about 13 times stronger than the bulk Cr orbital Hall effect rests on a hand-picked tanh activation function in Eq. (4). That decomposition is plausible, not proven.\n\nWhat is actually new: NiFe/Cr with native oxide as a self-contained orbital-torque source, no heavy-metal converter, and field-free switching at 1.58e11 A/m2. The Ta-cap and Cu-spacer controls are strong and show you need both the oxide interface and the direct NiFe/Cr contact. The OUMR correlation across thickness is a nice independent probe, and the first-principles trend of enhanced orbital Hall conductivity upon oxygenation is consistent. These are real contributions, and the paper cites the prior Cu/CuOx and Cr bulk-OHE literature fairly.\n\nSoft spots: Eq. (4) is the load-bearing assumption. The paper honestly shows that two constant sources (Eq. 3) cannot produce the observed peak, then fixes it by making the interfacial source \"activate\" as tanh(d/λ). That is an assumption, not a derivation from the diffusion equation, oxidation chemistry, or boundary conditions. Any sigmoidal-like activation would produce an interior maximum, so ξ_I, ξ_B, and λ are not uniquely pinned by this fit. If the activation form is wrong, the 13:1 interface-to-bulk ratio and the 4 nm transport length go with it. The authors should either derive g(d) from a microscopic model or test it with independent control over the oxidation level. Also, chromium is antiferromagnetic below ~311 K, and the paper runs at room temperature without mentioning it. It may be irrelevant, but it should be addressed. Data and DFT inputs are only available \"upon request,\" which limits independent checking.\n\nOverall, the experimental core — torque efficiency, controls, switching — looks credible and is the kind of result the orbitronics field needs. The mechanism model is an interesting hypothesis hanging off a good dataset. That is a reasonable state for a paper, but the abstract overstates it as \"revealing\" rather than \"consistent with.\"\n\nWho this is for: anyone working on orbital torques, oxide interfaces, or spin-orbit-torque switching. I would send it to peer review; a good referee can push the authors to harden the model and release data. I would cite the experimental findings while staying cautious about citing the decomposition.\n\nMy recommendation: accept for review, with heavy revision expected on the modeling side.","headline":"Serious experimental claim with strong controls, but the interface-dominance decomposition leans on an ad hoc tanh activation that is assumed, not derived.","tokens_in":18984,"tokens_out":3029,"would_cite":true,"duration_ms":27124,"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":"The paper claims that the native oxide that spontaneously forms on chromium turns a simple NiFe/Cr bilayer into a dual-channel orbital-current source, with a damping-like torque efficiency of 3.9 × 10^6 Ω⁻¹ m⁻¹ and field-free magnetization","keywords":["orbital Hall effect","orbital Rashba-Edelstein effect","chromium oxide","native oxidation","spin-orbit torque","field-free magnetization switching","orbital current","NiFe/Cr bilayers"],"falsifier":"Prepare a series of NiFe/Cr devices in which oxidation is precisely controlled—ranging from zero (in-situ capping, no air exposure) to several nanometres of deliberately grown Cr2O3—and measure the damping-like torque efficiency versus total Cr thickness. If the large torque and its non-monotonic peak are absent when no oxide is present, or if the peak thickness fails to move by the oxide thickness as d = t − t_ox changes, the oxidation-gated interfacial-source mechanism is falsified.","tokens_in":17919,"feed_emoji":"🧲","tokens_out":5977,"duration_ms":50369,"temperature":0.7,"pith_summary":"This paper aims to show that the oxide that naturally forms on chromium, usually considered a parasitic by-product, is actually the engine of an unusually efficient orbital-torque device. In a NiFe/Cr bilayer left to oxidize in air, the Cr/CrOx interface generates orbital current (a flow of orbital angular momentum rather than spin) that propagates through metallic Cr and is converted into spin torque in NiFe. The claimed damping-like torque efficiency reaches 3.9 × 10^6 Ω⁻¹ m⁻¹, about an order of magnitude above Pt and two above Ta, with no heavy-metal conversion layer. The same stack achieves deterministic field-free switching at 1.58 × 10^11 A/m². If correct, this makes native oxidation a scalable design tool rather than a defect.","feed_headline":"Chromium's native oxide delivers torque beyond Pt and Ta","feed_subtitle":"A simple NiFe/Cr bilayer left in air reaches high torque efficiency and switches magnetization with no external field.","key_machinery":"The load-bearing object is the Cr/CrOx interface formed by self-limiting native oxidation (about 3 nm of graded Cr2O3 on top of metallic Cr). The mechanism is the interfacial orbital Rashba–Edelstein effect: the inversion-symmetry-breaking oxide interface converts charge current into orbital angular-momentum current, which then travels through the metallic Cr channel and is transferred into NiFe. The fitting model introduces an 'oxidation-gated' source term, multiplying the interfacial efficiency by g(t) = tanh(d/λ), where d is the metallic Cr thickness minus the oxide thickness and λ ≈ 4 nm is the orbital transport length; the product of this rising activation with the decaying transmission","core_discovery":"The central claim is that a naturally oxidized NiFe/Cr heterostructure acts as a self-contained dual-channel orbital-current source. First-principles calculations indicate that oxygenating the Cr surface nearly triples the orbital Hall conductivity, through Cr 3d–O 2p hybridization. Harmonic Hall measurements find that the damping-like torque efficiency grows with Cr thickness to a clear maximum near 8 nm and then falls, a non-monotonic shape that a constant bulk plus constant interfacial source cannot produce. The paper accounts for it with a drift-diffusion model in which the interfacial orbital Rashba–Edelstein source is 'activated' by oxidation through a tanh(d/λ) factor, yielding an int","pith_inferences":["A strong consequence the paper leaves implicit: the peak torque thickness should shift by roughly the oxide thickness if oxidation conditions are changed, because the metallic channel thickness d = t − t_ox sets the transmission; this can be checked with controlled plasma oxidation.","If the oxidation-gated description is right, the interfacial torque could be modulated in operando by oxygen migration under an electric field or current, opening a route to electrically tunable orbital torque.","The close correlation between torque and orbital unidirectional magnetoresistance suggests a cheap all-electrical screening protocol for other candidate oxides: measure longitudinal second-harmonic resistance first, then reserve full harmonic-Hall torque analysis for the strongest candidates."],"forward_implications":["If the interfacial source is real, suppressing oxidation (for example by capping) should kill the large torque, as the authors demonstrate; this turns intentional oxidation into a control knob rather than an uncontrollable defect.","Devices based on this mechanism need no heavy-metal layer for orbital-to-spin conversion, removing the Pt/Ta/W overhead that usually accompanies orbital-torque devices.","Field-free switching at 1.58 × 10^11 A/m² follows from the large torque combined with a built-in anisotropy tilt created by the same stack, so one material system provides generation, conversion, and switching.","Orbital unidirectional magnetoresistance tracks the torque across the thickness series, giving a separate transport signature that can identify orbital accumulation in other candidate materials.","Other light metals with self-limiting native oxides could be screened for the same behavior, expanding the material set for low-power orbitronics."],"fun_headline_variants":["Rusty chromium layer delivers torque beyond heavy metals","Oxidized Cr surface triples orbital torque, switches magnets","Air-exposed NiFe/Cr bilayer beats Pt and Ta for spin torque","Native chromium oxide turns simple bilayer into torque giant","CrOx interface drives field-free switching at record efficiency"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the interfacial orbital-current source switches on with the ad hoc activation function g(t) = tanh(d/λ); because that functional form is assumed rather than derived, an incorrect source-thickness dependence would collapse the central division into a dominant interfacial and a minor bulk channel.","fun_headline_variants_meta":{"raw":{"variants":["Rusty chromium layer delivers torque beyond heavy metals","Oxidized Cr surface triples orbital torque, switches magnets","Air-exposed NiFe/Cr bilayer beats Pt and Ta for spin torque","Native chromium oxide turns simple bilayer into torque giant","CrOx interface drives field-free switching at record efficiency"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000234,"raw_usage":{"total_tokens":1371,"prompt_tokens":823,"completion_tokens":548,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":567,"completion_tokens_details":{"reasoning_tokens":466}},"tokens_in":567,"tokens_out":548,"duration_ms":6369,"temperature":1.0,"reasoning_tokens":466,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T09:49:04.364463+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Prepare a series of NiFe/Cr devices in which oxidation is precisely controlled—ranging from zero (in-situ capping, no air exposure) to several nanometres of deliberately grown Cr2O3—and measure the damping-like torque efficiency versus total Cr thickness. If the large torque and its non-monotonic peak are absent when no oxide is present, or if the peak thickness fails to move by the oxide thickness as d = t − t_ox changes, the oxidation-gated interfacial-source mechanism is falsified.","supporting_citations":[],"review_version":1}