{"id":"52c3c564-1648-4fd9-b4ab-f6bfa076148e","arxiv_id":"2412.08340","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A drift-diffusion model predicts that orbital torque in metallic bilayers scales with the product of orbital injection, orbit-to-spin conversion, and spin backflow efficiencies.","lead":"This paper builds a drift-diffusion model for spin and orbital currents in magnetic bilayers and shows that orbital torque is controlled by the balance between orbital current injection from a light metal and spin current backflow from the ferromagnet. It introduces orbit-spin and spin-orbit mixing conductances that connect this torque to orbital pumping through Onsager reciprocity.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central formula Eq. (17) depends on Eq. (7), which omits orbital precession around the crystal field; the paper's own citation of Refs. [32,33] leaves this simplification untested, so the compromise picture is conditional.","rationale":"I read the paper in good faith. The central claim is Eq. (17), the product formula for the orbital torque efficiency in the bulk conversion scenario. The derivation is not shown, but the formula is plausible and consistent with the SHE limit, as the reader notes. The most load-bearing assumption underpinning Eq. (17) is the orbital diffusion equation Eq. (7), which treats the orbital moment as a scalar relaxing quantity and deliberately drops the crystal-field precession that Refs. [32,33] predict. The authors flag this simplification, but they do not justify that it is negligible; indeed, saying it 'remained to be thoroughly discussed' indicates the issue is open. If precession matters, the simple product form of Eq. (17) and the associated design rules break down. This is exactly the reader's weakest assumption, so I agree. The concern could be settled by a concrete calculation with the precession term included. Since this is the same conditionality the reader already identified, the verdict remains CONDITIONAL; no adjustment is needed.","tokens_in":12646,"tokens_out":6254,"duration_ms":63399,"concrete_test":"Extend the model by replacing Eq. (7) with a two-component orbital drift-diffusion equation including crystal-field precession, e.g., ∂_z^2 μ_F_o = (1/λ_F_o^2) μ_F_o + (1/λ_L^2)(μ_F_o × n) with n a fixed crystal-field axis, and re-derive the transmission efficiency η_o for the same boundary conditions. Choose λ_L values from Refs. [32,33] (for instance 1–5 nm) and compare η_o(d_F) against Eq. (17). If the difference exceeds typical experimental uncertainty or changes the sign of the thickness trend, the simplification is load-bearing.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The derivation of the central efficiency formula, Eq. (17), relies on the orbital diffusion equation in the ferromagnet, Eq. (7), which is a scalar relaxation equation, ∂_z^2 μ_F_o = μ_F_o / λ_F_o^2. Refs. [32,33] predict that the orbital moment precesses around the local crystal field, so the correct equation should contain a term such as (1/λ_L^2)(μ_F_o × n), making the orbital chemical potential and current vectorial and oscillatory in space. The authors explicitly acknowledge neglecting this precession in Section II.A after Eq. (7). If the precession length λ_L is comparable to or smaller than λ_F_o, then the orbital current profile in F is no longer a simple exponential decay, the interfacial orbital chemical potential acquires components perpendicular to the injected polarization, and the orbital injection and spin backflow factors in Eq. (17) become coupled and complex. Consequently, the thickness dependence of the orbital torque, the interpretation of λ_F_o, and the design rule eG_F_o ≫ eG_N_o derived from Eq. (17) would all be modified. Since the compromise picture is the paper's central quantitative claim, this neglected precession is load-bearing, not a cosmetic detail.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript develops a drift-diffusion phenomenology for orbital torque in normal-metal/ferromagnet (N/F) bilayers, treating spin and orbital currents as coupled channels in the ferromagnet and at the interface. The central result is Eq. (17), which expresses the orbital-torque efficiency as a product of orbital-current injection from N, orbit-to-spin conversion in F, and spin-current backflow from F to N. The paper also discusses the thickness dependence of the torque, introduces orbit-spin and spin-orbit mixing conductances, and uses Onsager reciprocity to connect the torque to orbital pumping.","tokens_in":12939,"tokens_out":8675,"duration_ms":101584,"significance":"If correct, Eq. (17) provides a transparent design rule for orbital torque and extends the mixing-conductance framework from spin to orbital transport. The model is internally coherent, reproduces the known SHE limit in Eq. (16), and produces qualitative thickness dependences consistent with reported orbital-torque experiments. It also has the virtue of not fitting the torque target: the polarization parameters Pos and Pso are taken from the authors' microscopic calculations. However, the central quantitative claim rests on a scalar orbital diffusion equation whose validity is not established, and the derivation of the central efficiency formula is not shown.","major_comments":[{"comment":"The orbital diffusion equation in the ferromagnet is scalar, ∂_z^2 μ_F_o = μ_F_o / λ_F_o^2, while Refs. [32,33] predict that the orbital moment precesses around the local crystal field, which would make the orbital chemical potential and current vectorial and oscillatory in space. The text explicitly acknowledges this neglect, but the simplification is load-bearing because Eq. (17) and the thickness/design conclusions in Section IV.B follow from the exponential profile obtained from Eq. (7). If the orbital precession length is comparable to or smaller than λ_F_o, the orbital injection and spin-backflow factors in Eq. (17) become coupled tensors and the design rule ㄢ1G_F_o ≫ ㄢ1G_N_o is modified. Please provide an estimate of the precession length or extend the model to include this term.","section":"Section II.A, Eq. (7)"},{"comment":"The central efficiency formulas are introduced with 'we obtain' but no derivation is shown. Because Eq. (17) is the main quantitative claim of the paper, the derivation should be provided in an appendix, including the boundary conditions, the sign conventions for the spin backflow, and the approximations (thick F, weak SOC) under which the product structure is exact. The same applies to Eq. (18) for interfacial orbit-to-spin conversion.","section":"Section III.C, Eqs. (17)-(18)"},{"comment":"The orbital pumping response is fixed by invoking Onsager reciprocity, but the spin-orbit mixing conductance Gm_so is neither derived nor computed, and the reciprocal relation is stated without justification. Since orbital angular momentum is not a conserved quantity even in the absence of SOC, the Onsager relation between the torque and orbital pumping requires a derivation or at least an explicit statement of the assumptions under which it holds in this two-channel diffusive model.","section":"Section IV.C, Eq. (22)"}],"minor_comments":[{"comment":"The phrase 'compare with [Fig. 2(a,c)]' appears to be a typo and should be 'compare with panels (a,c)'.","section":"Fig. 2 caption"},{"comment":"The notation cosh^{-1}(d_N/λ) is used for the factor usually written as sech(d_N/λ); defining this once would avoid confusion.","section":"Section III.A, Eqs. (13)-(14)"},{"comment":"It would be helpful to state explicitly whether the plotted damping-like torque is the real part of the complex efficiency η or a magnitude, since the efficiencies are complex.","section":"Section IV.B, Fig. 3"},{"comment":"The statement that the factor 2 in the original spin-mixing conductance is absorbed into G_s^{↑↓} appears only in passing; it should be stated clearly where G_s^{↑↓} is first introduced.","section":"Section II.B, Eqs. (10)-(11)"},{"comment":"Reference [33] is cited as an arXiv preprint; if it has been published, the published reference should be provided.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for the journal and the phenomenological model is a useful contribution. The main concern is that the central compromise formula depends on a simplification that the authors themselves flag as potentially important; this is fixable by either extending the model to include orbital precession or by providing a quantitative justification for the scalar limit. The omitted derivation of Eq. (17) also needs to be supplied. I therefore recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: this paper gives the first drift-diffusion treatment of orbital torque that keeps orbital diffusion inside the ferromagnet, rather than assuming full absorption at the interface, and it gets a clean closed form for the efficiency, Eq. (17), as the product of orbital injection, orbit-to-spin conversion in F, and spin backflow to N. That compromise picture is the genuinely new result, and the extension to orbit-spin and spin-orbit mixing conductances plus Onsager-linked orbital pumping is a natural and useful step beyond the existing spin-only machinery.\n\nCredit where due: the model is internally coherent, reduces correctly to the SHE limit, and the predicted thickness trends match the qualitative behavior reported in the orbital-torque experiments. The authors are also transparent about the main modeling choice: Eq. (7) is a scalar relaxation equation for the orbital chemical potential, and they state in Section II.A that they neglect the orbital precession around the crystal field predicted in Refs. [32,33]. That is a real limitation, not a hidden one.\n\nThe soft spots are proportionate. First, the derivation of Eqs. (17) and (18) is not shown; the paper states them in the thick-F, weak-SOC limit and then moves on. Since the factorized efficiency is the load-bearing claim, that algebra should be in an appendix. Second, the omitted precession term is potentially more than a detail. If the precession length is not much larger than the orbital relaxation length, the orbital chemical potential becomes vectorial and oscillatory, the simple exponential profile in F breaks down, and the clean factorization in Eq. (17) becomes approximate. The paper acknowledges this but never says when the simplification is safe. I don't think it sinks the central physics point—the injection/backflow compromise is fairly robust—but the quantitative scalings with lambda_F_o and the eG_F_o >> eG_N_o design rule could be modified. Third, there is no direct quantitative comparison to experiments, which is acceptable for a phenomenology paper but worth noting.\n\nOverall, this is a solid, readable contribution for the orbital spintronics community. It deserves a serious referee, with the main requests being the missing derivation and an honest discussion of the validity regime of Eq. (7).\n\nRecommendation: send to peer review, asking for an appendix with the algebra and a paragraph on when the scalar orbital-diffusion approximation holds.","headline":"Clear closed forms for orbital torque as injection/backflow compromise, with a load-bearing simplification in Eq. (7) that needs an explicit validity check.","tokens_in":13480,"tokens_out":1843,"would_cite":true,"duration_ms":18852,"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":"This paper shows that orbital torque in normal-metal/ferromagnet bilayers is governed by a factorized efficiency combining orbital current injection, orbit-to-spin conversion, and spin current backflow, with a clear material design rule.","keywords":["orbital torque","orbital Hall effect","drift-diffusion model","spin-orbit conversion","mixing conductance","orbital pumping","spin current backflow","metallic bilayers"],"falsifier":"Measure the damping-like orbital torque as a function of ferromagnet thickness in a bilayer whose ferromagnet has a strong crystal-field splitting; if the torque-versus-thickness curve shows an oscillatory component or a sign change instead of the monotonic saturating form implied by $\\eta_o$ in Eq. (17), the dropped orbital-precession term in Eq. (7) is load-bearing and the model's central scaling fails.","tokens_in":12462,"feed_emoji":"🧲","tokens_out":10901,"duration_ms":94403,"temperature":0.7,"pith_summary":"This paper proposes a phenomenological drift-diffusion model for the orbital torque in normal-metal/ferromagnet bilayers. It claims that the torque is not set by the orbital current alone, but by a compromise between orbital current injection from the light-metal source into the ferromagnet and spin current backflow from the ferromagnet to the source. The central result is a compact efficiency formula, $\\eta_o = P_{os}\\left(1 + \\tilde{G}^{F}_{s}/\\tilde{G}^{N}_{s}\\right)^{-1}\\left(1 + \\tilde{G}^{N}_{o}/\\tilde{G}^{F}_{o}\\right)^{-1}$, which identifies the design rule: maximize the spin conductance of the source and the orbital conductance of the ferromagnet. The same framework yields orbital versions of the mixing conductance and an Onsager partner, orbital pumping. If correct, it gives device designers a transparent criterion for choosing materials for orbital-torque devices.","feed_headline":"Orbital torque is a compromise between injection and backflow","feed_subtitle":"Drift-diffusion model yields a design rule: let orbital current enter the magnet and spin current return to the source.","key_machinery":"The load-bearing object is the effective (spin or orbital) conductance $\\tilde{G}^{N(F)}_{s(o)} = \\sigma^{N(F)}_{s(o)}/\\tilde{\\lambda}^{N(F)}_{s(o)}$, which measures how well a metal absorbs a spin or orbital current. The argument is carried by the factorization of the orbital transmission efficiency, $\\eta_o = P_{os}\\left(1 + \\tilde{G}^{F}_{s}/\\tilde{G}^{N}_{s}\\right)^{-1}\\left(1 + \\tilde{G}^{N}_{o}/\\tilde{G}^{F}_{o}\\right)^{-1}$, where the first factor is orbital injection from N into F, the second is orbit-to-spin conversion inside F, and the third is spin current backflow from F to N. Around this formula the paper builds two-channel drift-diffusion equations that couple spin and orbital currents in F, interfacial boundary conditions with spin conductance and spin-orbit transfer conductances, and the resulting definitions of spin-mixing, orbit-spin-mixing, and spin-orbit-mixing conductances.","core_discovery":"The central claim is that the orbital torque in an N/F bilayer is not controlled by orbital injection alone but by a two-step compromise. The orbital Hall current generated in N must be injected into F, converted there into a spin current with efficiency $P_{os}$, and that spin current must then be transmitted back into N, where it is measured as a torque on the magnetization. Equation (17) factorizes this as $\\eta_o = P_{os}\\left(1 + \\tilde{G}^{F}_{s}/\\tilde{G}^{N}_{s}\\right)^{-1}\\left(1 + \\tilde{G}^{N}_{o}/\\tilde{G}^{F}_{o}\\right)^{-1}$, and the same structure survives when the conversion is moved to the N/F interface. The model predicts that the torque increases with F thickness and decreases with the orbital relaxation length in F, because long orbital diffusion in F allows backflow into N. It also defines orbit-spin- and spin-orbit-mixing conductances and uses Onsager reciprocity to predict an orbital pumping current alongside the usual spin pumping.","pith_inferences":["Not stated in the paper, the same injection/backflow compromise should reappear in interfacial orbital Rashba-Edelstein systems, where the source is an interface rather than a diffusive bulk layer; replacing the bulk conductances with interfacial orbital conductances would give a direct test of the factorization.","If the neglected orbital precession is real, the orbital current inside F should acquire an oscillating component; a fine thickness-series measurement of the orbital torque near the orbital relaxation length would reveal whether the monotonic Eq. (17) or an oscillating correction controls the signal.","The model separates bulk and interfacial orbit-to-spin conversion into two different efficiency formulas; inserting a heavy-metal spacer of variable thickness between the light-metal source and the magnet and checking which formula the torque follows would identify the dominant conversion site in real devices."],"forward_implications":["Design rule: to maximize orbital torque, choose the source metal with large spin conductance and small orbital conductance relative to the ferromagnet, i.e. $\\tilde{G}^{N}_{s} \\gg \\tilde{G}^{F}_{s}$ and $\\tilde{G}^{F}_{o} \\gg \\tilde{G}^{N}_{o}$.","A longer orbital relaxation length in the ferromagnet weakens the torque for a fixed thickness, because the orbital current can flow back into the source instead of being converted.","When orbit-to-spin conversion is interfacial rather than bulk, the same compromise is controlled by the interfacial conductances $G_{os}$, $G_{s}$, $G_{o}$ and the spin precession and dephasing in F, so inserting a strong-SOC interface layer changes the prefactor but not the physics.","The Onsager reciprocal of the orbital torque is orbital pumping: a precessing magnetization pumps an orbital current into the adjacent metal with strength set by the spin-orbit-mixing conductance $G^{m}_{so}$.","The perpendicular component of the orbital current, generated by successive orbit-to-spin and spin-to-orbit conversion, is much smaller than the in-plane component, which can serve as a fingerprint of the conversion chain."],"supporting_citations":[{"why":"Supplies the drift-diffusion plus spin-mixing-conductance framework that this paper extends from spin-only to coupled spin-orbital transport.","marker":"[25]"},{"why":"Provides the orbit-to-spin and spin-to-orbit polarization values used in Eqs. (3)-(4) and predicts the orbital precession that the model explicitly sets aside.","marker":"[33]"},{"why":"Demonstrates that orbital currents are not fully absorbed at the interface and propagate over long distances, which motivates solving for orbital diffusion inside F.","marker":"[19]"},{"why":"Gives the earlier drift-diffusion orbital-transport model that assumed full interfacial absorption, the assumption the present work corrects.","marker":"[31]"},{"why":"Defines the spin-mixing conductance, the concept this paper generalizes to orbit-spin and spin-orbit mixing conductances.","marker":"[38]"},{"why":"Supplies the Onsager reciprocity and response-matrix formalism used to connect orbital torque to orbital pumping.","marker":"[49]"},{"why":"Predicts precession of the orbital momentum around the local crystal field, the neglected effect identified as the model's weakest premise.","marker":"[32]"}],"fun_headline_variants":["Orbital torque hinges on injection and spin backflow","Orbital torque: a compromise between orbital and spin currents","Mixing conductances quantify orbital torque balance","Drift-diffusion reveals orbital torque's two-step nature","Orbital torque from injection–backflow compromise"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The model assumes that inside the ferromagnet the orbital density only relaxes without precessing around the local crystal field, so if that precession is significant the orbital current profile, the thickness dependence, and the claimed scaling with the orbital relaxation length would all change.","fun_headline_variants_meta":{"raw":{"variants":["Orbital torque hinges on injection and spin backflow","Orbital torque: a compromise between orbital and spin currents","Mixing conductances quantify orbital torque balance","Drift-diffusion reveals orbital torque's two-step nature","Orbital torque from injection–backflow compromise"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000309,"raw_usage":{"total_tokens":1753,"prompt_tokens":925,"completion_tokens":828,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":541,"completion_tokens_details":{"reasoning_tokens":749}},"tokens_in":541,"tokens_out":828,"duration_ms":12705,"temperature":1.0,"reasoning_tokens":749,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T17:55:15.052909+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the damping-like orbital torque as a function of ferromagnet thickness in a bilayer whose ferromagnet has a strong crystal-field splitting; if the torque-versus-thickness curve shows an oscillatory component or a sign change instead of the monotonic saturating form implied by $\\eta_o$ in Eq. (17), the dropped orbital-precession term in Eq. (7) is load-bearing and the model's central scaling fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the drift-diffusion plus spin-mixing-conductance framework that this paper extends from spin-only to coupled spin-orbital transport."},{"cited_title":"Orbital diffusion, polarization and swapping in centrosymmetric metals","cited_arxiv_id":"2310.04763","evidence_quote":"Provides the orbit-to-spin and spin-to-orbit polarization values used in Eqs. (3)-(4) and predicts the orbital precession that the model explicitly sets aside."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates that orbital currents are not fully absorbed at the interface and propagate over long distances, which motivates solving for orbital diffusion inside F."},{"cited_title":"Sala and P","cited_arxiv_id":null,"evidence_quote":"Gives the earlier drift-diffusion orbital-transport model that assumed full interfacial absorption, the assumption the present work corrects."},{"cited_title":"OBELIX” and by grant ANR-20-CE30-0022-01 “ORION","cited_arxiv_id":null,"evidence_quote":"Defines the spin-mixing conductance, the concept this paper generalizes to orbit-spin and spin-orbit mixing conductances."},{"cited_title":"Hayashi, D","cited_arxiv_id":null,"evidence_quote":"Supplies the Onsager reciprocity and response-matrix formalism used to connect orbital torque to orbital pumping."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Predicts precession of the orbital momentum around the local crystal field, the neglected effect identified as the model's weakest premise."}],"review_version":1}