{"id":"c2b0976d-a2a5-4919-90f7-8505996cc6ca","arxiv_id":"2412.02491","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Current along WTe2's a-axis induces orbital magnetization that switches Fe3GeTe2 without a field, while adding a b-axis pulse separately controls spin-orbit torque.","lead":"An electric current in the Weyl semimetal WTe2 creates an out-of-plane magnetization via a Berry curvature dipole, which can switch an adjacent ferromagnetic layer without any external magnetic field. The work also demonstrates a two-current scheme to independently control this orbital torque and the usual spin-orbit torque, with potential use in magnetoresistive memory.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The orbital-vs-spin torque attribution remains insecure: the spin Hall channel is not computed, yet could supply the measured out-of-plane torque.","rationale":"The paper is well organized and presents mutually consistent experiments, calculations, and simulations: field-free switching, the cosine angle dependence of the detected out-of-plane signal, the two-current independent control, and the AHE loop shifts are all strong controls. The FGT-probe concern raised by the reader is addressable with the existing angle dependence and two-Au control, so I do not see it as the most load-bearing element. The decisive issue is the attribution of the out-of-plane torque to the orbital Edelstein effect rather than to unconventional spin-orbit torque. WTe2 is a known source of large spin Hall effects, and prior work (e.g., Ref. 38) has already attributed field-free switching in WTe2/ferromagnet bilayers to out-of-plane spin-orbit torque. The paper's only quantitative separation between orbital and spin is the Fermi-surface Edelstein comparison, which does not capture the Fermi-sea spin Hall contribution. Since the measured out-of-plane torque efficiency is small, a quantitative spin Hall torque estimate could change the conclusion. I therefore keep the CONDITIONAL verdict but emphasize this specific omission as the primary correctness risk, solvable by a concrete computational check.","tokens_in":21219,"tokens_out":16772,"duration_ms":181522,"concrete_test":"Recompute the out-of-plane torque efficiency expected from the full spin Hall conductivity tensor of the same five-layer WTe2 slab used in Note 7 (e.g., via Wannier-interpolated Kubo formula for sigma_{xy}^z and sigma_{yx}^z), using the same Fermi-level estimation procedure and an assumed spin-mixing conductance at the WTe2/FGT interface. Compare this spin-Hall-derived Xi_DL^z with the measured 0.024 at 90 K. If the spin Hall channel alone reaches or exceeds 0.024, the orbital Edelstein effect is not required to explain the field-free switching; if it is much smaller, the orbital attribution is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that the orbital Edelstein effect (from the Berry curvature dipole) dominates the out-of-plane antidamping-like torque enabling field-free switching of FGT. The key quantitative evidence is the first-principles comparison in Fig. 3(e) showing M_z^orb ~100x larger than M_z^spin at the estimated Fermi level (63 ± 5 meV at 100 K). This comparison, however, uses only the Fermi-surface spin Edelstein coefficient (Eq. S5); it does not include the spin Hall effect, which in WTe2 is known to be large and, for a current along the a axis, can produce an out-of-plane spin current that exerts an antidamping torque on FGT. The measured out-of-plane torque efficiency Xi_DL^z = 0.024 (Note 11) is modest, so if the spin Hall channel alone can account for 0.024, the orbital mechanism is not necessary to explain the switching. Moreover, the Fermi level is inferred from a separate few-layer WTe2 device at 100 K, not from the actual heterostructure, leaving both sign and magnitude of the orbital Edelstein coefficient uncertain. Thus the core attribution—orbital torque, not spin-orbit torque, is dominant—is not yet securely established because the spin Hall channel is not quantitatively assessed.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports current-induced out-of-plane magnetization in few-layer WTe2, detected through a Fe3GeTe2 (FGT) magnetic probe electrode, and attributes it to the orbital Edelstein effect arising from the Berry curvature dipole (BCD). Based on first-principles calculations, the authors argue that the out-of-plane orbital magnetization is about two orders of magnitude larger than the spin magnetization at the estimated Fermi level, and that this orbital magnetization produces an out-of-plane antidamping-like torque enabling field-free perpendicular magnetization switching of an adjacent FGT layer. They further demonstrate a two-current scheme (DC along the a axis, pulses along the b axis) that separately controls orbital and spin-orbit torques, supported by anomalous Hall loop-shift measurements and micromagnetic simulations. The manuscript also includes an all-van-der-Waals magnetoresistive memory demonstration.","tokens_in":21399,"tokens_out":2332,"duration_ms":26767,"significance":"If the central attribution holds, the work would establish the Berry-curvature-dipole-induced orbital Edelstein effect as a functional torque source for field-free perpendicular switching, complementing the well-studied spin-orbit torque channel in low-symmetry semimetals. The paper combines several independent pieces of evidence: nonlinear Hall effect characterization, magnetization detection with angle and current dependence, first-principles orbital and spin Edelstein calculations, micromagnetic modeling, and reproducible switching in multiple devices. The two-torque independent-control scheme is an appealing and potentially useful device concept. The experimental data are presented in detail, including control measurements with nonmagnetic electrodes and heating-effect discussions. However, the quantitative attribution of the observed torque to the orbital rather than the spin Hall channel, and the interpretation of the probe-electrode signal, require additional support before the central claim is fully secured.","major_comments":[{"comment":"The magnetization-detection measurement uses a FGT electrode as the magnetic probe, and the hysteresis in V_m^omega is interpreted as the relative alignment of a current-induced out-of-plane WTe2 magnetization with MFGT. The control measurement with two Au electrodes (Fig. S5(a)) rules out a signal originating from proximity-induced magnetism in WTe2, but it does not rule out a spin-dependent magnetoresistance or contact-related effect at the FGT/h-BN/WTe2 interface, which would also produce a hysteresis loop whose polarity reverses with the driving-current direction. An additional control that reverses the magnetization of the FGT probe without changing the WTe2 current direction, or a measurement that directly correlates the sign of V_m^omega with the independently known switching direction in the same device, would strengthen the interpretation.","section":"Fig. 2 and Supplemental Note 6"},{"comment":"The claim that the orbital Edelstein effect dominates the out-of-plane antidamping-like torque relies on the comparison in Fig. 3(e), where the computed out-of-plane orbital magnetization exceeds the spin magnetization by about two orders of magnitude. However, the spin contribution is calculated only from the Fermi-surface spin Edelstein coefficient (Eq. S5); it does not include the spin Hall effect, which in WTe2 is known to produce sizable out-of-plane spin currents and corresponding out-of-plane spin-orbit torques. The measured out-of-plane torque efficiency is modest (xi_DL^z = 0.024, Note 11), so the spin Hall channel alone could plausibly account for the observed switching. The authors should either compute the spin Hall contribution to the out-of-plane torque in the same first-principles framework or provide an experimental bound that places the spin Hall contribution below the measured efficiency.","section":"Fig. 3(e) and Supplemental Note 7 (Eq. S5)"},{"comment":"The Fermi level mu_F = 63 ± 5 meV is extracted from carrier-density measurements on a separate few-layer WTe2 device at 100 K, and this value is then used to compute the orbital and spin magnetization in Fig. 3(e) for the WTe2/FGT heterostructure. The actual heterostructure may have a different doping due to the adjacent FGT and h-BN layers, and the measurement temperature of the switching experiments is 90 K rather than 100 K. Since the ratio M_z^orb/M_z^spin is strongly mu-dependent (Fig. 3(e) shows the spin magnetization changing sign in the shaded region), the quantitative dominance claim would benefit from a sensitivity analysis with respect to mu_F and from an estimate of the Fermi-level uncertainty in the heterostructure itself.","section":"Supplemental Note 8 and Fig. 3(e)"},{"comment":"The micromagnetic simulations use a canting angle theta to represent the ratio of out-of-plane to in-plane torque components, with theta = -1.09 deg assigned to the case of I_p along b with I_DC = +0.4 mA along a. This mapping is derived from torque efficiencies, but the simulations do not independently implement the orbital torque mechanism; they simply parameterize an out-of-plane antidamping torque. The simulations therefore illustrate that an out-of-plane antidamping torque can produce the observed partial and full switching, but they do not by themselves discriminate between an orbital and a spin-Hall origin. This should be stated more explicitly, and the simulation parameters (e.g., Ms = 1.7e5 A/m, alpha = 0.02) should be justified against the FGT films used in the experiments, whose reported Ms is lower (16 emu/cm^3 = 1.6e4 A/m in Note 11).","section":"Supplemental Note 13"}],"minor_comments":[{"comment":"The phrase 'characterized by a Berry curvature dipole' in the abstract is slightly ambiguous; the paper actually demonstrates the BCD via the nonlinear Hall effect and calculates it from first principles, so a wording such as 'in which the Berry curvature dipole is present and characterized' would be clearer.","section":"Abstract and Introduction"},{"comment":"The raw V_m^omega traces show a large background that is subtracted in Fig. 2(c), but the subtraction procedure is only described in a sentence. Please specify how the background is defined (e.g., linear fit outside the hysteresis region) and show at least one raw trace for a-axis and b-axis cases.","section":"Fig. 2(c,d) and Fig. S5(b)"},{"comment":"The symbol mu_b is introduced as 'the Bohr magneton' but the more common notation is mu_B; if the paper uses mu_b consistently, please note the convention. Also, in the same sentence 'reaches about 16 mu_b' should specify whether this is per atom, per unit cell, or per k-point; the figure caption states units of mu_b/nm^2 in Fig. 3(e), so consistency is needed.","section":"Main text, third paragraph after Eq. (1)"},{"comment":"Equation (S1) contains a notation mix: alpha_xz is first written as an integral over m_nk^z v_nk^x d_epsilon f, and then the same symbol alpha_xz is used for the total coefficient after the -mu D_xz + B_xz decomposition. Please distinguish the bare orbital moment contribution from the final alpha_xz and define all symbols (such as b_nk) in the main text before using them.","section":"Supplemental Note 7, Eq. (S1)"},{"comment":"The partial switching of about 65% is described in the text, but the figure panels do not show the full R_xy versus I_p range; adding a dashed line at the expected full-switching level would make the 'partial' claim immediately visible.","section":"Main text, Fig. 4(c,d)"}],"recommendation":"major_revision","confidential_remarks":"The paper presents a novel and potentially important mechanism, and the experimental dataset is substantial. The main issue is that the central attribution to the orbital Edelstein effect over the spin Hall effect is not yet quantitatively closed: the spin Hall channel is not computed or bounded, the FGT-probe detection needs a control that flips the probe magnetization independently, and the Fermi-level estimate carries device-to-device uncertainty. These are fixable with additional calculations and targeted experiments, so major revision rather than rejection is appropriate. The manuscript would also benefit from stating more explicitly which parts of the evidence are consistent with but not unique to the orbital mechanism."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Compared to the earlier WTe2/FGT field-free switching reports from this group, the paper adds two solid things: a direct detection of a current-induced out-of-plane magnetization in WTe2 whose angular dependence matches the Berry curvature dipole expectation, and a two-current scheme that independently controls an out-of-plane orbital torque (DC along a) and an in-plane SOT (pulse along b). The detection experiment is carefully controlled—two-Au-electrode reference, sign reversal with current direction, linear current dependence—and the first-principles calculation of the orbital Edelstein coefficient on a five-layer slab is credible and internally consistent with the measured NLHE. The multilevel memory states and the MTJ demonstration are useful. These genuinely extend the prior work.\n\nThe main soft spot is the quantitative attribution. The comparison in Fig. 3(e) uses only the Fermi-surface spin Edelstein coefficient; the spin Hall effect in WTe2, which is known to be large, is not computed. For current along the a axis, spin Hall can generate out-of-plane spins and an antidamping-like torque. The measured out-of-plane torque efficiency is 0.024, modest enough that the spin Hall channel alone might account for the switching. Without a spin Hall estimate, the title-level claim that the orbital torque is the dominant mechanism is not secured. This is not a fatal flaw—the data still show a BCD-correlated out-of-plane magnetization—but it is a load-bearing omission for the interpretation. The Fermi level used in the calculation is inferred from a separate WTe2 device at 100 K, and the orbital Edelstein coefficient is sensitive to μ, so the 100x ratio in Fig. 3(e) has real uncertainty. The FGT-probe detection, while well controlled, still rests on an interface magnetoresistance interpretation that is not independently calibrated. These are the three things I'd want a referee to push on.\n\nNone of this kills the paper. The symmetry arguments, the two-torque separation, and the device demonstrations are strong enough to warrant serious engagement. The authors are clearly in control of the experiments and the theory; the weakest pieces are omissions, not errors.\n\nSend it to peer review. Request a spin Hall calculation or a reasoned argument for its suppression, and a discussion of the Fermi-level sensitivity. If the authors can close the spin Hall gap, this becomes a very useful paper for the orbital torque community.","headline":"Solid two-torque control scheme and BCD-correlated magnetization detection; the orbital-dominance claim needs a spin Hall calculation to hold.","tokens_in":22002,"tokens_out":4146,"would_cite":true,"duration_ms":44120,"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":"A current along WTe2's a-axis induces an out-of-plane orbital magnetization that enables field-free perpendicular magnetization switching of an adjacent Fe3GeTe2 layer.","keywords":["orbital Edelstein effect","Berry curvature dipole","field-free switching","WTe2","spin-orbit torque","perpendicular magnetization","Fe3GeTe2","van der Waals heterostructure"],"falsifier":"If an independent probe of the WTe2 magnetization, such as magnetic circular dichroism or a scanning NV magnetometer, found no out-of-plane magnetization when a current is driven along the a axis, the orbital Edelstein torque explanation would be falsified.","tokens_in":1755,"feed_emoji":"🧲","tokens_out":6820,"duration_ms":95713,"temperature":0.7,"pith_summary":"This paper reports that a current along the a axis of the Weyl semimetal WTe2 induces an out-of-plane orbital magnetization in WTe2 itself, arising from the Berry curvature dipole of its band structure. The resulting orbital torque, rather than conventional spin-orbit torque, is claimed to be what enables deterministic switching of an adjacent perpendicularly magnetized Fe3GeTe2 layer with no applied magnetic field. The authors further show that a small direct current along the a axis and a pulsed current along the b axis independently control orbital and spin-orbit torques, allowing full or partial switching and a route to multilevel memory. First-principles calculations support the orbital Edelstein origin, with the computed out-of-plane orbital magnetization two orders of magnitude larger than the spin magnetization.","feed_headline":"Orbital torque, not spin, switches a magnet without a field","feed_subtitle":"A tiny current along WTe2's a-axis creates an out-of-plane orbital magnetization that toggles Fe3GeTe2.","key_machinery":"The central object is the Berry curvature dipole $D_{xz}$ of low-symmetry few-layer WTe2, together with the orbital magnetic moment texture $m^z_{\\mathrm{orb}}$ that mirrors the antisymmetric Berry curvature $\\Omega_z(k_x,k_y) = -\\Omega_z(-k_x,k_y)$. An electric field along the a axis drives a nonequilibrium out-of-plane orbital magnetization (orbital Edelstein effect) through the coefficient $\\alpha_{xz} \\approx -\\mu D_{xz}$, which produces an out-of-plane antidamping-like torque on the adjacent Fe3GeTe2 layer. The two-current scheme (a-axis DC for orbital torque, b-axis pulsed current for spin-orbit torque) allows independent control of the two torque components.","core_discovery":"The core claim is that the orbital Edelstein effect, driven by the Berry curvature dipole of few-layer Td-WTe2, produces a current-induced out-of-plane magnetization that exerts an out-of-plane antidamping-like torque on an adjacent Fe3GeTe2 layer, thereby enabling field-free perpendicular magnetization switching. Experiments detect the current-induced out-of-plane magnetization using a Fe3GeTe2 magnetic probe electrode separated by a h-BN tunnel barrier; the signal is linear in current, follows a cosine angle dependence with maximum along the a axis, and vanishes along the b axis. First-principles calculations attribute the dominant part of the orbital Edelstein coefficient to the Berry curvature dipole term $-\\mu D_{xz}$ at the experimental Fermi level, and the calculated orbital magnetization is about two orders of magnitude larger than the spin magnetization. When a small DC current along the a axis is combined with a pulsed current along the b axis, the orbital and spin-orbit torques act jointly: the sign of the DC current sets the final magnetization direction, and larger DC currents yield more complete switching, consistent with micromagnetic simulations.","pith_inferences":["The Berry curvature dipole mechanism should be transferable to other low-symmetry type-II Weyl semimetals such as TaIrTe4, and could be tunable by strain, twist, or electrostatic gating.","The magnetic-probe detection scheme could be adapted to quantify current-induced orbital magnetization in other materials, but its reliance on the ferromagnetic probe means a nonmagnetic tunnel probe would provide a cleaner control experiment.","The demonstrated partial switching states suggest a natural route to analog synaptic devices, although device-to-device reproducibility of the partial states remains to be tested.","The dominance of orbital over spin magnetization is established by comparing computed orbital and spin Edelstein coefficients; a direct experimental separation of orbital and spin torque contributions, for example through thickness- or angle-dependent torque measurements, would further solidify the claim."],"forward_implications":["Field-free perpendicular magnetization switching can be achieved without symmetry-breaking layers or external magnetic fields, using only the intrinsic Berry curvature dipole of a low-symmetry Weyl semimetal.","The orbital torque, rather than spin-orbit torque, is the dominant mechanism for currents along the a axis, which changes how such devices should be engineered.","Independent control of orbital and spin-orbit torques enables full, partial, and multilevel magnetization states, offering a path to multilevel nonvolatile memory and neuromorphic computing.","An all-van-der-Waals magnetic tunnel junction can be written electrically through the WTe2 layer, with a tunneling magnetoresistance of about 10 percent at low temperature.","The switching polarity is set by the sign of the a-axis current, not by Joule heating, since equal currents with opposite signs produce opposite final magnetization states."],"supporting_citations":[{"why":"Establishes the Berry curvature dipole as the intrinsic origin of the nonlinear Hall effect, which the paper uses to identify the orbital Edelstein contribution.","marker":"[22]"},{"why":"Demonstrates the nonlinear Hall effect in few-layer WTe2, confirming the presence of a Berry curvature dipole in the material used here.","marker":"[23]"},{"why":"Provides the orbital Edelstein effect theory that connects the Berry curvature dipole to current-induced orbital magnetization.","marker":"[28]"},{"why":"Gives the formalism for current-induced orbital magnetization in non-centrosymmetric systems, used to compute the orbital Edelstein coefficients.","marker":"[29]"},{"why":"Shows field-free perpendicular magnetization switching by orbit-transfer torque, providing a precedent for the orbital torque mechanism.","marker":"[34]"},{"why":"Demonstrates control of spin-orbit torques through crystal symmetry in WTe2/ferromagnet bilayers, giving the basis for the b-axis spin-orbit torque.","marker":"[36]"},{"why":"Shows all-electric magnetization switching in WTe2/ferromagnet heterostructures, supporting the possibility of current-driven switching in this system.","marker":"[37]"},{"why":"Derives the orbital magnetoelectric susceptibility including the Berry curvature dipole and effective magnetic field contributions, which the paper uses for its $\\alpha_{xz}$ calculation.","marker":"[47]"},{"why":"Provides the Mumax3 micromagnetic simulation framework used to model the collaborative torque dynamics.","marker":"[51]"}],"fun_headline_variants":["Orbital torque, not spin, switches magnet at zero field","Berry curvature dipole drives field-free magnetization switching","Orbital Edelstein effect enables field-free switching","Switching a magnet without a field using orbital torque","WTe2 orbital torque flips Fe3GeTe2 with no field"],"cache_read_input_tokens":24064,"weakest_assumption_plain":"The interpretation that the voltage hysteresis measured between the Fe3GeTe2 probe electrode and a gold reference electrode reflects the relative alignment of a current-induced orbital magnetization in WTe2 with the Fe3GeTe2 magnetization, rather than a spin-dependent magnetoresistance of the Fe3GeTe2 interface, is the load-bearing premise.","fun_headline_variants_meta":{"raw":{"variants":["Orbital torque, not spin, switches magnet at zero field","Berry curvature dipole drives field-free magnetization switching","Orbital Edelstein effect enables field-free switching","Switching a magnet without a field using orbital torque","WTe2 orbital torque flips Fe3GeTe2 with no field"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000258,"raw_usage":{"total_tokens":1575,"prompt_tokens":934,"completion_tokens":641,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":550,"completion_tokens_details":{"reasoning_tokens":558}},"tokens_in":550,"tokens_out":641,"duration_ms":6398,"temperature":1.0,"reasoning_tokens":558,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T23:24:13.745680+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"If an independent probe of the WTe2 magnetization, such as magnetic circular dichroism or a scanning NV magnetometer, found no out-of-plane magnetization when a current is driven along the a axis, the orbital Edelstein torque explanation would be falsified.","supporting_citations":[{"cited_title":"MacNeill, G","cited_arxiv_id":null,"evidence_quote":"Demonstrates control of spin-orbit torques through crystal symmetry in WTe2/ferromagnet bilayers, giving the basis for the b-axis spin-orbit torque."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows all-electric magnetization switching in WTe2/ferromagnet heterostructures, supporting the possibility of current-driven switching in this system."},{"cited_title":"In summary, we have demonstrated the synergy be- tween orbital and spin-orbit torques in WTe 2/Fe3GeTe2 heterostructures","cited_arxiv_id":null,"evidence_quote":"Provides the Mumax3 micromagnetic simulation framework used to model the collaborative torque dynamics."}],"review_version":1}