{"id":"7bd9bd8f-e2aa-471a-8a29-e1ee2246efb2","arxiv_id":"2501.00737","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"First-principles calculations predict out-of-plane spin currents in Fe, Co, Ni and alloys, with a non-relativistic spin Hall angle near 0.25 in MnAl(101) enabling field-free switching.","lead":"Computer simulations show that ordinary ferromagnetic metals like iron, cobalt, nickel, and the alloy MnAl can create spin currents with out-of-plane polarization, which could switch magnetic memory without an external magnetic field. The standout prediction is a non-relativistic spin Hall angle near 0.25 in MnAl(101), far larger than the 0.01 to 0.02 angles found in iron, cobalt, and nickel.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"MnAl(101) giant spin Hall result relies on bulk tensor rotation and ideal interface; thin-film/interface effects could invalidate the switching claim.","rationale":"The reader's weakest assumption correctly identifies the bulk-to-film transfer and perfect interface transparency as the pivotal conditions. I agree: the entire MnAl(101) result is derived from a coordinate rotation of the bulk (001) conductivity tensor (Eq. 3), and the switching analysis adopts the resulting spin Hall angle with no interface correction. This is the least secure step in the argument; if it fails, the headline numbers do not apply to a real device. However, this is a standard type of approximation and is addressable by a slab/interface calculation, so the appropriate verdict remains CONDITIONAL rather than REJECT. No additional internal inconsistency was found; the tensor transformation itself is algebraically correct, and the resistivity validation for Fe/Co/Ni gives some confidence in the methodology.","tokens_in":12193,"tokens_out":10187,"duration_ms":100630,"concrete_test":"Perform a first-principles transport calculation for an explicit MnAl(101)/CoFeB heterostructure (or at minimum a free-standing 10–20 layer MnAl(101) slab) using the same Kubo-Bastin KKR-GF formalism, and compute the effective spin torque efficiency including interface transmission. If the device-level θ_eff differs from the bulk-rotated θ_D by more than ~30%, then Eqs. (2)–(4) do not justify the claim of deterministic low-current field-free switching.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central quantitative claim—σ_SH up to 10^4 (ℏ/2e)(Ω·cm)−1 and θ≈0.25 for L10-MnAl(101)—is obtained by rotating the bulk (001) spin conductivity tensor using Eqs. (2)–(4). This assumes that the bulk conductivity tensor is a size-independent property and that a (101) film can be described by the same tensor with the coordinate axes reoriented. In a thin film, quantum confinement along the film normal changes the band structure and Fermi surface; surface states and reduced symmetry at the interface can also alter the anisotropic longitudinal spin conductivities σ_xx^z and σ_zz^z that enter Eq. (4). Moreover, the LLG switching calculation in Sec. D uses the bulk spin Hall angle θ_D and assumes perfect interface transparency (J_sw = (2e/ℏ)τ_DL/(t_z θ_D)), neglecting interfacial spin memory loss and field-like torques. If the real spin transmission is lower or the anisotropy is modified, the switching current density J_sw could increase by an order of magnitude, undermining the practical usefulness claim. The bulk calculation alone does not establish the device-level performance.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports first-principles KKR-GF calculations of the spin Hall effect in 3d ferromagnetic metals and alloys (Fe, Co, Ni, FeCo, NiFe) and in L10-MnAl. For Fe, Co, Ni and their alloys, the authors compute SOC-driven spin Hall conductivities with out-of-plane spin polarization on the order of 1000 (ℏ/2e)(Ω cm)^−1 and spin Hall angles of 0.01–0.02 at room temperature. For L10-MnAl, they argue that a non-relativistic spin Hall effect arises from anisotropic longitudinal spin conductivity; by rotating the bulk (001) spin conductivity tensor to the (101) orientation, they predict spin Hall conductivities around 3×10^4 (ℏ/2e)(Ω cm)^−1 and a spin Hall angle around 0.25 at 300 K. An LLG model is used to show that these ferromagnetic spin sources can switch perpendicular magnetization field-free, with MnAl(101) giving the lowest switching current density and power density.","tokens_in":12360,"tokens_out":4048,"duration_ms":41693,"significance":"If the results hold, the paper provides a systematic first-principles benchmark of spin Hall properties in 3d ferromagnets and identifies a promising non-relativistic spin Hall mechanism in anisotropic ferromagnets for field-free SOT switching. The computational setup (KKR-GF, Kubo-Bastin, CPA, dense k-meshes) is standard and the calculated resistivities agree well with experimental values, lending credibility to the transport calculations for Fe, Co, Ni, and their alloys. The identification of two distinct spin Hall mechanisms and the explicit symmetry analysis are useful contributions. However, the central quantitative claim for L10-MnAl(101) rests on a bulk tensor rotation and an idealized LLG switching model; these assumptions are not validated against thin-film or interface effects.","major_comments":[{"comment":"The giant non-relativistic spin Hall conductivities for MnAl(101) are obtained by rotating the bulk (001) spin conductivity tensor using Eqs. (2)–(4). This assumes that the bulk longitudinal spin conductivities σ_xx^z and σ_zz^z remain unchanged in a thin-film (101) geometry, which neglects quantum confinement, surface states, and interface-induced symmetry lowering. Since the predicted spin Hall angle θ≈0.25 and the associated switching efficiency directly scale with the anisotropy σ_xx^z − σ_zz^z, the device-level claims are not yet supported. The authors should either perform thin-film or supercell transport calculations that include the (101) surface, or clearly restate the result as a bulk-rotation estimate and temper the switching-efficiency conclusions accordingly.","section":"Section C, Eqs. (2)–(4) and Fig. 5"},{"comment":"The switching current density is evaluated as Jsw = (2e/ℏ)τ_DL/(t_z θ_D), using bulk spin Hall angles and implicitly assuming perfect interface transparency and zero spin memory loss. For a ferromagnetic spin source in contact with a second ferromagnetic layer, interfacial spin-dependent reflection, spin memory loss, and the reduced effective spin current transmitted into the free layer can substantially increase the required current density. The quantitative Jsw and power-density comparisons in Fig. 6(e) are therefore optimistic upper bounds. The authors should state this limitation explicitly and, if possible, incorporate interface parameters or compare with experimental SOT efficiencies in similar ferromagnetic bilayers.","section":"Section D, Jsw formula and Fig. 6"},{"comment":"The scaling relations σ_zx^z ∝ M(T)σ_xx and σ_zx^y ∝ a σ_xx^2 are described as \"universal\" for ferromagnetic materials, but they are presented as empirical fits to the present calculations without a derivation. If this universality is intended to support predictions for other materials, the authors should provide a more formal argument (e.g., based on the Boltzmann equation or the skew-scattering side-jump decomposition) or soften the claim to an observed trend within the studied alloy set.","section":"Section B, Fig. 3 and accompanying text"}],"minor_comments":[{"comment":"The spin Hall angle for MnAl(101) is quoted as \"around 0.25\" in the abstract and Section D, but the values θ_x=0.25 and θ_z=0.19 in Fig. 5(d) imply a magnitude θ_D=sqrt(0.25^2+0.19^2)≈0.31. The inconsistency should be resolved.","section":"Abstract and main text"},{"comment":"The rotation matrix in Eq. (2) is garbled in the manuscript text; it should be typeset clearly so that the transformation R and the sign conventions in Eq. (4) can be verified.","section":"Equation (2)"},{"comment":"The tensor entries in Table I are difficult to parse because of formatting, particularly the distinction between Todd and Teven components. A cleaner presentation with explicit 3×3 matrices would improve readability.","section":"Table I"},{"comment":"The LLG equation (5) is written with a torque term (m×s) for the field-like torque and (m×(m×s)) for the damping-like torque, but the derivation of the effective field B_M from K_eff is stated without the standard factor of 2; please check the convention against the cited literature.","section":"Section D"},{"comment":"The acronym \"ABSTRCT\" appears in the abstract; this typo should be corrected.","section":"Introduction"}],"recommendation":"major_revision","confidential_remarks":"The paper's main quantitative claim for MnAl(101) is vulnerable to thin-film and interface effects that are not modeled. The authors should be asked to either provide explicit thin-film transport calculations or substantially soften the device-level statements. The bulk results for Fe, Co, Ni, and alloys appear sound and are a useful dataset. The citation practice is acceptable, though several of the authors' own papers are cited for closely related results."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The main news here is a set of new ab initio numbers: temperature- and composition-dependent spin Hall conductivities for Fe, Co, Ni, FexCo100-x, and Ni100-xFex, plus a non-relativistic spin Hall effect in L10-MnAl(101) with a predicted spin Hall angle around 0.25 at 300 K. The computational core is standard but well executed: KKR-GF with Kubo-Bastin, CPA for disorder, 10^7 k-points, and the resistivity matches experiment across a wide temperature range. That agreement is a real anchor, and it makes the SOC-driven spin Hall results credible. The scaling of the Todd and Teven spin Hall conductivities with longitudinal conductivity is a nice organizing observation, even if those scalings are fits to the computed outputs rather than independent predictions. No circularity problem: the central conductivities are calculated, not fitted to target values.\n\nThe soft spots are where the paper reaches beyond what it actually computed. The MnAl(101) giant spin Hall effect comes from rotating the bulk (001) longitudinal spin-conductivity tensor using Eqs. (2)–(4). That assumes the bulk tensor survives unchanged in a thin film, with no quantum confinement, surface states, or interface reconstruction. The LLG switching analysis then assumes perfect interface transparency and ignores field-like torques and spin memory loss. Those are serious caveats for any device claim: if spin transmission is lower or the transport anisotropy is modified, Jsw could rise substantially. The switching demonstration is a model calculation, not a device demonstration, and the paper would be improved by stating that plainly. Also, no code, input files, or data are shared, which limits reproducibility for a calculation-heavy paper.\n\nThose issues are addressable rather than fatal. The bulk calculations stand on their own, and the MnAl prediction is a well-defined falsifiable target for experiment. The paper is not overselling its novelty—the two mechanisms, magnetic spin Hall effect and non-relativistic anisotropic transport, are already known—but the material-specific results are new. It deserves a serious referee, especially one who can probe the tensor rotation and the interface assumptions. I would send it to review with a request to moderate the device claims and ideally release inputs.\n\nWho is this for? People working on spin-orbit torque materials and field-free switching. It is a useful reference set for 3d ferromagnet spin Hall conductivities, and the MnAl prediction is worth knowing even if it ultimately gets corrected by film effects.","headline":"A solid, useful first-principles study of spin Hall effects in 3d ferromagnets, with a big MnAl(101) prediction that rests on a bulk tensor rotation and an idealized interface; worth serious refereeing but the device-level claims need toning down.","tokens_in":12957,"tokens_out":960,"would_cite":true,"duration_ms":11703,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["72.25.Ba","75.76.+j"],"model":"deepseek-v4-flash","headline":"The paper argues that ferromagnetic metals can replace heavy metals as spin current sources for spin-orbit torque switching, with an out-of-plane spin polarization that removes the need for an external magnetic field.","keywords":["spin Hall effect","spin-orbit torque","field-free switching","perpendicular magnetization","L10-MnAl","ferromagnetic metals","first-principles transport","Kubo-Bastin formalism"],"falsifier":"A spin-torque ferromagnetic resonance or harmonic Hall measurement on an epitaxial L10-MnAl(101) bilayer: if the measured damping-like SOT efficiency comes out far below 0.25, or if the torque does not follow the predicted $\\sin\\phi\\cos\\phi$ variation across MnAl films grown on different plane orientations, the bulk-rotation picture is wrong. A slab-geometry first-principles transport calculation that includes surface states and interface disorder would give the same verdict.","tokens_in":11965,"feed_emoji":"🧲","tokens_out":7545,"duration_ms":65580,"temperature":0.7,"pith_summary":"The paper tries to establish that ferromagnetic metals can replace heavy metals as spin current sources for spin-orbit torque switching of perpendicular magnetization. It computes two mechanisms from first principles: the spin-orbit-coupling-driven spin Hall effect in Fe, Co, Ni, and their alloys, which yields spin Hall conductivities around $1000\\,(\\hbar/2e)(\\Omega\\,\\text{cm})^{-1}$ at room temperature but small spin Hall angles of 0.01–0.02, and a non-relativistic spin Hall effect in L10-MnAl(101) that reaches conductivities of order $10^4$ with a spin Hall angle near 0.25. The reason this matters is that the out-of-plane spin polarization produced by a ferromagnetic source breaks the mirror symmetry that forces heavy-metal devices to require an external in-plane field, so perpendicular magnetization can be switched deterministically with no applied field. The authors demonstrate such switching by solving the Landau–Lifshitz–Gilbert equation with the calculated torques, identifying L10-MnAl(101) as the most energy-efficient of the studied sources.","feed_headline":"Spin Hall angle of 0.25 makes MnAl switch magnets field-free","feed_subtitle":"First-principles transport shows a rotated ferromagnet can beat heavy metals for spin-orbit-torque switching.","key_machinery":"The load-bearing object is the third-rank spin conductivity tensor $\\sigma^s_{\\mu\\nu}$ together with its symmetry constraints. For cubic ferromagnets the argument splits the tensor into time-reversal odd (Todd) and time-reversal even (Teven) parts and derives scaling laws, $\\sigma^z_{zx} \\propto M(T)\\sigma_{xx}$ and $\\sigma^y_{zx} \\propto \\sigma_{xx}^2$, that connect the magnetic spin Hall effect to magnetization and disorder. For MnAl the mechanism is a tensor rotation: the (001) plane's longitudinal anisotropic spin conductivity, which is nonzero even without spin-orbit coupling, is rotated to the (101) plane through Eqs. (2)–(4), turning the anisotropy $\\tilde{\\sigma}^z_{zz} - \\tilde{\\sigma}^z_{xx}$ into off-diagonal spin Hall terms proportional to $\\sin\\phi\\cos\\phi$. The paper then feeds the resulting conductivities into the Landau–Lifshitz–Gilbert equation with a damping-like torque that contains an out-of-plane spin component, and solves for the magnetization dynamics to demonstrate switching.","core_discovery":"On the paper's own terms, the discovery is that ferromagnets produce usable out-of-plane (Sz) spin currents by two routes. In cubic 3d metals, magnetization parallel to the current breaks the mirror planes and generates a magnetic, spin-orbit-driven spin Hall component $\\sigma^z_{zx}$ of order $1000\\,(\\hbar/2e)(\\Omega\\,\\text{cm})^{-1}$ at 300 K, scaling linearly with magnetization and longitudinal conductivity, yet the accompanying spin Hall angle is only about 0.01–0.02 because of the high longitudinal conductance. In tetragonal L10-MnAl, the longitudinal spin conductivity is strongly anisotropic, with $\\tilde{\\sigma}^z_{zz} \\approx 1.2\\times10^5$ versus $\\tilde{\\sigma}^z_{xx} \\approx 3.3\\times10^4\\,(\\hbar/2e)(\\Omega\\,\\text{cm})^{-1}$, and rotating the crystal plane from (001) to (101) converts this anisotropy into off-diagonal spin Hall conductivities of about $3\\times10^4$ with a spin Hall angle near 0.25 that persists up to 400 K. With these values, the LLG equation gives deterministic, field-free switching of perpendicular magnetization, with L10-MnAl(101) requiring the lowest switching current density and threshold power density among the materials compared.","pith_inferences":["The $\\sin\\phi\\cos\\phi$ rotation law of Eq. (4) is a design rule the paper applies only to the (101) plane: the same recipe should work for any ferromagnet with strong longitudinal spin-conductivity anisotropy, so screening tetragonal magnets such as the Mn-Ga, Fe-Pt, and Mn-Bi families for large $\\tilde{\\sigma}^z_{zz} - \\tilde{\\sigma}^z_{xx}$ could reveal materials with comparable or larger effect","The 0.25 spin Hall angle assumes perfect interface transparency, so real MnAl(101)/ferromagnet stacks will lose efficiency to interface spin-mixing conductance and disorder; the paper's number is an upper bound to test against, not a device guarantee.","An explicit experimental probe of the mechanism is the growth-angle dependence: measuring the SOT efficiency on MnAl films cut at several vicinal angles should follow $\\sin\\phi\\cos\\phi$, which would separate the non-relativistic contribution from conventional spin-orbit-driven torque."],"forward_implications":["Fe, Co, Ni, and their FexCo1−x and NixFe1−x alloys can switch perpendicular magnetization without an external field, at switching current densities near $10^8\\,\\text{A}/\\text{cm}^2$, comparable to conventional heavy-metal systems but without the field requirement.","L10-MnAl(101), with a spin Hall angle near 0.25, brings the switching current down to the order of $10^7\\,\\text{A}/\\text{cm}^2$ and shows the lowest threshold power density among the studied sources, placing it ahead of the antiferromagnet and alloy references.","The MnAl spin Hall angle stays close to 0.25 from low temperature up to 400 K, so the non-relativistic mechanism is not degraded by thermal disorder the way the spin-orbit-driven angles are.","Heavy-element ferromagnets such as Fe50Pt50 sit between the two extremes, with a spin Hall angle of about 0.1 and a low power density, suggesting alloying 3d magnets with 5d elements as a tuning knob.","The linear and quadratic scaling laws for the Todd and Teven spin Hall conductivities give a predictive rule for how these torques behave with temperature and composition across alloy families."],"supporting_citations":[{"why":"Predicted the non-relativistic spin Hall mechanism in magnetic metals with anisotropic transport spin polarization, the mechanism the paper applies to L10-MnAl.","marker":"[29]"},{"why":"Prior first-principles treatment of the magnetic spin Hall effect in ferromagnets that supplies the Fe50Pt50 benchmark (spin Hall angle \\approx 0.1, \\xi = 0.66) and the Todd/Teven decomposition.","marker":"[24]"},{"why":"Provides the Kubo-Bastin linear response formalism used to compute all spin and longitudinal conductivities.","marker":"[34]"},{"why":"Supplies the multiple-scattering Korringa-Kohn-Rostoker Green's function method underlying the first-principles transport calculations.","marker":"[30]"},{"why":"Gives the treatment of phonon and spin-disorder scattering within the coherent potential approximation used for the temperature dependence.","marker":"[35]"},{"why":"Supplies the Landau-Lifshitz-Gilbert switching framework and the relations among switching current density, spin Hall angle, and power density used in the switching analysis.","marker":"[12]"},{"why":"Provides the Mn3Pt spin Hall conductivity value of $275\\,(\\hbar/2e)(\\Omega\\,\\text{cm})^{-1}$ used as a comparison baseline for the 3d ferromagnet results.","marker":"[45]"}],"fun_headline_variants":["MnAl spin Hall angle 0.25 enables field-free switching","No field needed: MnAl's spin Hall effect switches magnets","Ferromagnet MnAl: spin Hall angle 0.25 for zero-field switching","Spin Hall effect in ferromagnets: field-free switching without heavy metals","L10-MnAl's 0.25 spin Hall angle switches magnets without a field"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Everything hinges on the assumption that the bulk MnAl(001) conductivity tensor, rotated to the (101) plane, describes what a real thin-film device does, together with a perfectly transparent interface in the switching analysis; if surface states, quantum confinement, or interface scattering alter the transport anisotropy or reduce spin transmission, the 0.25 spin Hall angle and the switching efficiency will drop.","fun_headline_variants_meta":{"raw":{"variants":["MnAl spin Hall angle 0.25 enables field-free switching","No field needed: MnAl's spin Hall effect switches magnets","Ferromagnet MnAl: spin Hall angle 0.25 for zero-field switching","Spin Hall effect in ferromagnets: field-free switching without heavy metals","L10-MnAl's 0.25 spin Hall angle switches magnets without a field"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.002108,"raw_usage":{"total_tokens":8294,"prompt_tokens":1156,"completion_tokens":7138,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":772,"completion_tokens_details":{"reasoning_tokens":7037}},"tokens_in":772,"tokens_out":7138,"duration_ms":44438,"temperature":1.0,"reasoning_tokens":7037,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:44:04.226479+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A spin-torque ferromagnetic resonance or harmonic Hall measurement on an epitaxial L10-MnAl(101) bilayer: if the measured damping-like SOT efficiency comes out far below 0.25, or if the torque does not follow the predicted $\\sin\\phi\\cos\\phi$ variation across MnAl films grown on different plane orientations, the bulk-rotation picture is wrong. A slab-geometry first-principles transport calculation that includes surface states and interface disorder would give the same verdict.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Predicted the non-relativistic spin Hall mechanism in magnetic metals with anisotropic transport spin polarization, the mechanism the paper applies to L10-MnAl."},{"cited_title":"Zheng, M","cited_arxiv_id":null,"evidence_quote":"Prior first-principles treatment of the magnetic spin Hall effect in ferromagnets that supplies the Fe50Pt50 benchmark (spin Hall angle \\approx 0.1, \\xi = 0.66) and the Todd/Teven decomposition."},{"cited_title":"Ködderitzsch, K","cited_arxiv_id":null,"evidence_quote":"Provides the Kubo-Bastin linear response formalism used to compute all spin and longitudinal conductivities."},{"cited_title":"Ebert, D","cited_arxiv_id":null,"evidence_quote":"Supplies the multiple-scattering Korringa-Kohn-Rostoker Green's function method underlying the first-principles transport calculations."},{"cited_title":"Chadova, S","cited_arxiv_id":null,"evidence_quote":"Gives the treatment of phonon and spin-disorder scattering within the coherent potential approximation used for the temperature dependence."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Mn3Pt spin Hall conductivity value of $275\\,(\\hbar/2e)(\\Omega\\,\\text{cm})^{-1}$ used as a comparison baseline for the 3d ferromagnet results."}],"review_version":1}