{"id":"1ebb6e0c-7c90-4bdc-990c-5ee8d0336aed","arxiv_id":"1908.02680","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"First-principles calculations show that interfacial contributions to damping-like spin-orbit torque and magnetoresistance in Co/Pt and Co/Au are comparable to, or larger than, the bulk spin-Hall contribution.","lead":"This computational study calculates spin-orbit torque and magnetoresistance in cobalt/heavy-metal bilayers from first principles. It finds that interface effects, not just the bulk spin Hall effect, can dominate the measured signals, which matters for spintronic device design.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation (4), the spin-Hall magnetoconductance benchmark behind the >10x interfacial-excess claim, is dimensionally inconsistent as printed; the quantitative conclusion must be re-derived before it can stand.","rationale":"The paper is a serious computation with independent support: disorder-averaged NEGF calculations, explicit finite-size checks at w = 16 ML, and a complete vector-spherical-harmonic expansion of the torque. The reader's Eq. (1) concern is legitimate and is even flagged by the authors themselves. However, for the paper's strongest claim—the magnetoconductance Δyzg of order e²/h with a >10x excess over the spin-Hall prediction—the more load-bearing step is the benchmark formula Eq. (4). As printed it has the wrong dimensions, and the Fig. 2 comparison is complicated by the unexplained factor of 10. That does not refute the interfacial-origin conclusion, and a rough corrected estimate suggests the qualitative gap may remain, but the quantitative 'more than an order of magnitude' statement is not currently supported by the printed derivation. The appropriate response is to keep the paper CONDITIONAL on a corrected, dimensionally consistent derivation of Eq. (4) and a recomputed comparison; this does not change the reader's verdict, so I set verdict_should_be to UNCHANGED.","tokens_in":9226,"tokens_out":11572,"duration_ms":134128,"concrete_test":"Independently re-derive Eq. (4) from the spin-diffusion model of Refs. [12,41] under the stated 'thin magnetic layer' assumption, tracking all dimensions and boundary conditions. Then recompute Δyzg_SH(d_N) for Co/Pt using the Table I parameters with the corrected formula (for example, θ_SH² λ/ρ̄ tanh²(d_N/2λ) or the exact expression), without the arbitrary factor 10, and compare point-by-point with the calculated Δyzg in Fig. 2. If the corrected curve falls within the error bars, the >10x interfacial-excess claim fails; if it remains at least an order of magnitude below the calculated values over the plotted d_N range, the interfacial-origin conclusion is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The strongest claim is quantitative: the calculated Δyzg in Co/Pt is of order e²/h and exceeds the spin-Hall magnetoconductance prediction by more than an order of magnitude. The benchmark for that comparison is Eq. (4). As printed, Eq. (4) cannot be correct: Δyzg is defined in Eq. (3) as a reduced conductance LG/w with units of siemens, while the right-hand side θ_SH² ρ̄ tanh²(d_N/2l_sf) tanh(d_N/l_sf) has units of resistivity (Ω·m) for every thickness. No dimensionally consistent spin-diffusion derivation is supplied, and the cited Refs. [12,41] are not translated into the form used here. The dashed curve in Fig. 2 is additionally 'scaled by a factor of 10', without showing the unscaled benchmark. If the intended formula is the standard spin-diffusion SMR expression, for example θ_SH² (λ/ρ̄) tanh²(d_N/2λ) possibly modified for the metallic ferromagnet boundary condition, both its magnitude and its thickness dependence differ from Eq. (4). A rough estimate using the Table I Co/Pt parameters at d_N ≈ 12 ML gives a value near 0.1 e²/h, still below the computed conductance, so the qualitative interfacial-origin conclusion may survive; but the printed comparison does not establish the claimed order-of-magnitude excess. The central magnetoresistance claim therefore rests on an unverified and apparently misprinted formula.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports first-principles non-equilibrium Green's function (NEGF) calculations of spin-orbit torque (SOT) and magnetoresistance in Co/Pt and Co/Au bilayers with explicit Anderson disorder. The damping-like SOT is fitted to τ0 + τSH[1 - sech(dN/lsf)], yielding a thickness-independent interfacial part τ0 comparable to the spin-Hall part τSH. The transverse magnetoconductance Δyzg in Co/Pt is reported to be of order e²/h, exceeding the spin-Hall magnetoresistance prediction of Eq. (4) by more than an order of magnitude, which the authors attribute to an interfacial contribution. The paper also analyzes the field-like torque and the planar-Hall-like term, and proposes that the spin-Hall mechanism cannot account for the observed magnetoconductance.","tokens_in":9587,"tokens_out":6233,"duration_ms":63642,"significance":"If the central quantitative claim holds, the paper would provide first-principles evidence that interfacial transport processes dominate the magnetoresistance of Co/Pt bilayers, complementing similar conclusions for damping-like SOT. The methodological strengths include the use of NEGF with explicit disorder, the systematic expansion of SOT in vector spherical harmonics, and a genuine cross-observable check: θSH and lsf are fitted to SOT and then used to predict the magnetoresistance, not fitted to it. However, the dimensional inconsistency in Eq. (4) and the model dependence of the SOT decomposition currently prevent the central quantitative claim from being substantiated as written.","major_comments":[{"comment":"Equation (4) as printed is dimensionally inconsistent: Δyzg is defined in Eq. (3) as a reduced conductance with units of siemens, while the right-hand side θ_SH² ρ̄ tanh²(d_N/2l_sf) tanh(d_N/l_sf) has units of resistivity (Ω·m). The dashed line in Fig. 2 is additionally 'scaled by a factor of 10' without showing the unscaled curve, so the claim that the spin-Hall mechanism is too weak by more than an order of magnitude is not reproducible from the manuscript. The authors should re-derive the spin-diffusion SMR expression from the cited references, state any unit convention or missing length factor explicitly, and plot the unscaled benchmark.","section":"Magnetoresistance, Eq. (4)"},{"comment":"The decomposition of the damping-like torque into τ0 + τSH[1 - sech(dN/lsf)] assumes a strictly thickness-independent interfacial contribution. The paper itself warns that this 'assumes a geometrical interface between homogeneous bulk regions and ignores thickness-dependent perturbations and finite-size effects.' Because the central claim of a comparable interfacial contribution rests on the fitted τ0, the authors should provide a sensitivity analysis, for example by allowing τ0 to have a weak thickness dependence or by fitting only data in a range where the interface is expected to be converged. Without such a test, the separation of τ0 and τSH in Table I is model-dependent.","section":"Thickness dependence of SOT, Eq. (1)"},{"comment":"Even after fixing the units in Eq. (4), the benchmark uses the same θSH and lsf values that were obtained from the model-dependent SOT decomposition of Eq. (1). The comparison in Fig. 2 therefore inherits the assumptions of that decomposition. The manuscript should acknowledge this coupling explicitly and, if possible, show how the benchmark would change if the SOT fit parameters are varied within their uncertainty.","section":"Magnetoresistance, comparison with computed Δyzg"}],"minor_comments":[{"comment":"The abstract and conclusions state that Δyzg is 'of the order of a conductance quantum per interfacial atom,' but the body text (Magnetoresistance section) states it is 'of the order of one conductance quantum' without a per-atom normalization. Please clarify which quantity is meant and define the normalization.","section":"Abstract and Conclusions"},{"comment":"The dashed line is described as 'scaled by a factor of 10' but the unscaled prediction is not shown. Please include the unscaled curve (or an inset) so the reader can directly compare the computed points with the theoretical benchmark.","section":"Fig. 2"},{"comment":"There is a typo: 'τ0 andτSH' should have a space after 'and'.","section":"Thickness dependence of SOT, first paragraph"},{"comment":"The notation '∆gyz/g' in the paragraph on the growth at small thicknesses is inconsistent with the earlier definition '∆µνg/g'; please unify the notation.","section":"Magnetoresistance, last paragraph"},{"comment":"The quantity 'M/A' is used in the definition of θSH but is not explicitly defined in the main text; please define it as the total magnetic moment per unit area and specify its units.","section":"Eq. (2)"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a timely and important question, and the NEGF methodology is of high quality. The main issue is the dimensional inconsistency in Eq. (4), which undermines the quantitative benchmark behind the headline claim. With a corrected derivation and a transparent unscaled comparison, plus a sensitivity analysis of the SOT decomposition, the paper could be made publishable. I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth knowing: this paper has two real results and one botched comparison. The SOT decomposition is genuinely new and mostly solid; the magnetoresistance benchmark is not reproducible as printed.\n\nWhat is new: thickness-resolved NEGF-LMTO calculations of damping-like torque in Co/Pt and Co/Au with explicit Anderson disorder, decomposed using a vector-spherical-harmonics expansion. The clean result is that tau0, the zero-thickness intercept, is comparable to tauSH, the spin-Hall coefficient, in both systems, and larger in Co/Pt. The cross-check idea is good: they fit theta_SH and l_sf to the SOT thickness data, then use those parameters to predict the spin-Hall magnetoconductance without fitting to the MR data. That is not circular.\n\nMain problem: Eq. (4) is dimensionally inconsistent. Delta_yz g is defined as LG/w, which has units of siemens. The right-hand side is theta_SH^2 rho_bar times dimensionless tanh factors, which has units of resistivity. So the dashed curve in Fig. 2, “scaled by a factor of 10,” cannot be reproduced without some unstated conversion. The authors cite Refs. [12,41] but do not translate those results into their units. This matters because the claim that interfacial magnetoconductance exceeds the spin-Hall prediction by more than an order of magnitude is quantitative. The qualitative conclusion may survive—a rough estimate from their own parameters at d_N = 12 ML gives something near 0.1 e^2/h, below the computed conductance—but the paper as printed does not establish the stated excess. They need to re-derive the benchmark with correct units, show the unscaled curve, and give error bars on the SOT fit parameters.\n\nSecond soft spot: the tau0/tauSH split depends on the assumed functional form Eq. (1), which presumes a sharp interface and a thickness-independent interfacial torque. They acknowledge this themselves. The thickness-independent intercept is suggestive, but it is not a unique physical decomposition. I would treat it as a model-dependent estimate, not a direct measurement of a well-defined quantity.\n\nCitation pattern is fine; they extend their own Ref. [26] and cite the relevant prior experimental and theoretical work. No invented entities, no data overclaim beyond the MR issue.\n\nWho this is for: spintronics theorists and experimentalists working on SOT/SMR in Co/Pt and Co/Au. It deserves a serious referee. Send it to peer review with the expectation of heavy revision: Eq. (4) must be fixed and the quantitative MR claim re-derived. If that works out, the paper is a solid contribution. If not, the SOT decomposition alone is still worth publishing.","headline":"Serious first-principles SOT decomposition with a solid interfacial-torque result, but the magnetoresistance benchmark is dimensionally inconsistent as printed and the order-of-magnitude claim needs a re-derivation.","tokens_in":10107,"tokens_out":3670,"would_cite":true,"duration_ms":40230,"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 claims that in ferromagnet/heavy-metal bilayers, both damping-like spin-orbit torque and the transverse magnetoconductance usually attributed to spin-Hall magnetoresistance carry comparably large interfacial contributions that…","keywords":["spin-orbit torque","spin-Hall magnetoresistance","interfacial transport","non-equilibrium Green's function","ferromagnet/heavy-metal bilayer","damping-like torque","vector spherical harmonics","magnetoconductance"],"falsifier":"Compute $\\Delta_{yz}g$ for Co/Pt bilayers with the spin-orbit coupling in the Pt bulk turned off while leaving the interface unchanged. The paper's interfacial mechanism predicts $\\Delta_{yz}g$ remains of order $e^2/h$; if it collapses to the spin-Hall prediction, the central claim is wrong.","tokens_in":9041,"feed_emoji":"🧲","tokens_out":9127,"duration_ms":94265,"temperature":0.7,"pith_summary":"This paper asks where spin-orbit torque and magnetoresistance actually arise in ferromagnet/heavy-metal bilayers such as Co/Pt and Co/Au. Using first-principles transport calculations with disorder, it finds that the damping-like torque decomposes into a bulk spin-Hall part that grows with heavy-metal thickness and a comparable interfacial part that does not. It further finds that the transverse magnetoconductance in Co/Pt is about one conductance quantum, more than an order of magnitude larger than the spin-Hall magnetoresistance model predicts. The conclusion is that both spin-orbit torque and magnetoresistance carry large interfacial contributions unrelated to the bulk spin-Hall effect, so measurements should not be interpreted solely through the spin-Hall mechanism.","feed_headline":"Interfacial effects dwarf spin-Hall magnetoresistance in Co/Pt","feed_subtitle":"First-principles calculations show spin-Hall theory alone cannot explain Co/Pt's transverse magnetoresistance.","key_machinery":"The argument rests on three tools: a first-principles non-equilibrium Green's function calculation with Anderson disorder that treats the whole bilayer quantum-mechanically; an expansion of the torquance tensor, the tensor mapping electric field to torque, in vector spherical harmonics, an orthonormal basis that cleanly separates damping-like from field-like torque terms; and a thickness-dependence analysis that fits the damping-like torque to $\\tau_0+\\tau_{\\rm SH}[1-\\mathrm{sech}(d_N/l_{\\rm sf})]$ and compares the magnetoconductance with the spin-Hall SMR formula. The vector-harmonic expansion is what lets the paper identify which torque terms are damping-like and which are field-like, and the thickness fit is what separates the bulk spin-Hall piece from the interfacial piece.","core_discovery":"The central claim is that in Co/Pt and Co/Au bilayers, the damping-like spin-orbit torque is the sum of a bulk spin-Hall contribution that grows with heavy-metal thickness and a thickness-independent interfacial contribution of comparable size. In Co/Pt, the fitted interfacial constant is about 110 ns/m versus about 97 ns/m for the bulk spin-Hall part, and in Co/Au the interfacial part is also substantial. The paper additionally finds that the transverse magnetoconductance $\\Delta_{yz}g$ in Co/Pt is of order $e^2/h$, which exceeds the spin-Hall magnetoresistance formula $\\Delta_{yz}g^{\\rm SH}=\\theta_{\\rm SH}^2\\bar\\rho\\tanh^2(d_N/2l_{\\rm sf})\\tanh(d_N/l_{\\rm sf})$ by more than an order of magnitude. The paper concludes that this magnetoconductance, like the damping-like torque, likely has an interfacial origin rather than a spin-Hall origin.","pith_inferences":["If this interfacial picture is right, spin-Hall angles extracted from torque experiments on nanometer-thick bilayers are not bulk material parameters; they depend on interface quality and disorder, so comparing samples requires controlling the interface as well as the metal.","A direct test would be to insert a single monolayer spacer at the Co/Pt interface or to vary interfacial roughness while keeping the Pt bulk unchanged; the interfacial torque and the large $\\Delta_{yz}g$ should change substantially if they are truly interfacial.","A natural computational extension is to repeat the thickness series with spin-orbit coupling scaled separately in the bulk and at the interface, mapping where the torque and magnetoresistance are generated and turning the fitted separation into a microscopic assignment.","The same vector-spherical-harmonic expansion could be applied to other bilayer combinations, including systems with weak bulk spin-Hall effect, potentially revealing interfacial spin-orbit torque in materials previously classified as spin-Hall dominated."],"forward_implications":["For Co/Pt at typical disorder, the interfacial damping-like torque constant is about 110 ns/m, comparable to the bulk spin-Hall value near 97 ns/m, so neglecting the interface would overestimate the bulk spin-Hall torque.","The spin-Hall magnetoresistance formula underpredicts the calculated $\\Delta_{yz}g\\sim e^2/h$ by more than an order of magnitude, so measured SMR in metallic bilayers cannot be taken as direct evidence for the spin-Hall effect.","Since the angular functions of SMR, AMR, and interfacial anomalous magnetoresistance are linearly dependent, angular scans alone can determine only two independent parameters; separating the mechanisms requires thickness dependence or interface control.","If the interfacial contribution is included, the effective spin-Hall angle inferred for Co/Pt rises from about $\\theta_{\\rm SH}\\approx 0.02$ to about $0.06$, matching typical experimental values.","The linear growth of SMR at small thicknesses, often cited in favor of spin-Hall theory, may instead indicate the thickness at which a continuous metal film forms, because the interface properties change over a few monolayers."],"supporting_citations":[{"why":"Supplies the ab initio non-equilibrium Green's function methodology and the earlier calculation that spin-orbit coupling on Co does not affect damping-like SOT, which the paper extends.","marker":"[26]"},{"why":"Provides the phenomenological spin-Hall magnetoresistance theory whose angular and thickness dependence the paper tests against ab initio results.","marker":"[12]"},{"why":"Introduces the conventional spin-Hall torque interpretation and gives the experimental spin-Hall angle value used for comparison.","marker":"[4]"},{"why":"Gives the theory of damping-like torque generated by interfacial scattering, which the paper invokes to explain the thickness-independent term.","marker":"[8]"},{"why":"Reports metallic-bilayer SMR measurements with thickness dependence that the paper argues do not uniquely imply the spin-Hall mechanism.","marker":"[41]"},{"why":"Provides the Co/Pt magnetoresistance experiment previously attributed to spin Hall, which the paper's interfacial interpretation questions.","marker":"[42]"},{"why":"Supplies the Berry-phase spin Hall conductivity used to benchmark the bulk spin-Hall part of the torque.","marker":"[39]"}],"fun_headline_variants":["Interfacial effects rival spin-Hall in Co/Pt bilayers","Spin-Hall theory falls short for Co/Pt transport and torque","Co/Pt interfacial contributions match spin-Hall in torque and MR","Co/Pt spin-orbit torque: interfaces matter as much as bulk","Spin-Hall model cannot explain Co/Pt magnetoresistance"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The fit splits the torque into a bulk spin-Hall term with thickness dependence $1-\\mathrm{sech}(d_N/l_{\\rm sf})$ and a thickness-independent interfacial term; if the interfacial contribution itself changes with thickness, the split is not a physical decomposition.","fun_headline_variants_meta":{"raw":{"variants":["Interfacial effects rival spin-Hall in Co/Pt bilayers","Spin-Hall theory falls short for Co/Pt transport and torque","Co/Pt interfacial contributions match spin-Hall in torque and MR","Co/Pt spin-orbit torque: interfaces matter as much as bulk","Spin-Hall model cannot explain Co/Pt magnetoresistance"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001108,"raw_usage":{"total_tokens":4598,"prompt_tokens":905,"completion_tokens":3693,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":521,"completion_tokens_details":{"reasoning_tokens":3605}},"tokens_in":521,"tokens_out":3693,"duration_ms":29706,"temperature":1.0,"reasoning_tokens":3605,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:41:02.409289+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute $\\Delta_{yz}g$ for Co/Pt bilayers with the spin-orbit coupling in the Pt bulk turned off while leaving the interface unchanged. The paper's interfacial mechanism predicts $\\Delta_{yz}g$ remains of order $e^2/h$; if it collapses to the spin-Hall prediction, the central claim is wrong.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the ab initio non-equilibrium Green's function methodology and the earlier calculation that spin-orbit coupling on Co does not affect damping-like SOT, which the paper extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the theory of damping-like torque generated by interfacial scattering, which the paper invokes to explain the thickness-independent term."},{"cited_title":"Kawaguchi, D","cited_arxiv_id":null,"evidence_quote":"Provides the Co/Pt magnetoresistance experiment previously attributed to spin Hall, which the paper's interfacial interpretation questions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Berry-phase spin Hall conductivity used to benchmark the bulk spin-Hall part of the torque."}],"review_version":1}