{"id":"ec118cb4-8ef5-4e04-87b0-131129756b73","arxiv_id":"2412.02488","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Field-free, all-electric writing of a van der Waals magnetoresistive memory at room temperature is achieved in TaIrTe4/Fe3GaTe2 by current-induced orbital magnetization tied to the Berry curvature dipole.","lead":"This paper shows a van der Waals memory that can be written by electric current alone at room temperature, using a Weyl semimetal layer (TaIrTe4) to convert current into an out-of-plane orbital magnetization that flips a ferromagnetic layer (Fe3GaTe2). It offers a low critical current density around 2×10^6 A/cm2 and uses the nonlinear Hall effect as a predictor of switching polarity.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. E4's spin magnetic moment is nonstandard (k-derivative form), so the computed M_orb/M_spin ≈ 15 that underlies 'governs' may be an artifact; the ratio is τ-independent, leaving Fermi-level and interface conversion as the real uncertainties.","rationale":"The reader's weakest assumption correctly points to the DFT ratio α_orb/α_spin ≈ 15 as load-bearing. My stress-test sharpens this: the specific vulnerability is the nonstandard expression for the spin magnetic moment in Eq. E4, which could invalidate the ratio even before considering calibration or interface effects. I also correct the reader's implicit worry about τ: the ratio is τ-independent, so the relaxation time is not the fragile parameter; the orbital-to-torque transmission at the TaIrTe4/Fe3GaTe2 interface is a separate, unaddressed assumption. The experimental evidence for deterministic field-free switching and the working magnetoresistive memory is strong and independently valuable, and the NLHE/switching polarity correlation is a useful symmetry fingerprint. However, the mechanistic attribution to orbital magnetization is quantitatively supported only by a calculation whose spin operator is mis-specified in the manuscript. This does not change the overall verdict: the paper merits conditional acceptance with the mechanistic claim requiring either a corrected calculation or explicit acknowledgement that the orbital/spin division is not established. The proposed recomputation is a single, decisive check that would settle whether the 15× ratio is real.","tokens_in":15832,"tokens_out":12152,"duration_ms":134477,"concrete_test":"Disclose the exact operator used for m_spin in the WannierBerri calculation (e.g., input parameters or code snippet), or recompute M_spin using the standard spin magnetic moment mspin_nk = −μ_B⟨σ^z⟩_nk for the same tight-binding model, same 1501×501 grid, T = 300 K, and μ = 0.037 eV. Reproduce Fig. 4(b) and compare the ratio M_orb/M_spin. Then repeat at μ = 0.017 eV and 0.057 eV to bracket the carrier-density calibration. If the recomputed ratio drops below ~3 or its sign changes, the claim that orbital magnetization governs the switching is unsupported as stated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that orbital magnetization 'governs' field-free switching depends on the ratio M_orb/M_spin ≈ 15 at μ = 0.037 eV (Fig. 4b). The denominator, α_spin, is defined in Eq. E4 with a spin magnetic moment mspin_nk = −⟨∂kunk|(1/2)gμ_b σ|∂kunk⟩. This is not the spin magnetic moment of a Bloch state, which is the expectation value of the spin operator, −μ_B⟨σ⟩_nk, with no k-derivative. If Eq. E4 was implemented literally, the calculated spin response is not the spin Edelstein coefficient, so the ratio 15 does not compare orbital and spin torques. If instead the standard spin expectation was used, the manuscript does not state this, leaving the calculation undocumented at the load-bearing point. Note also that the ratio M_orb/M_spin = (e/2ħ)(α_orb/α_spin) is independent of the relaxation time τ, since both magnetizations are linear in τ; the τ = 1 ps assumption is therefore not the fragile input. The genuinely fragile inputs are the Fermi-level calibration (μ = 0.037 eV from carrier densities) and the unaddressed conversion of the TaIrTe4 orbital moment into torque on Fe3GaTe2. The paper's experimental correlations between NLHE sign and switching polarity are consistent with the orbital mechanism but do not distinguish it from a spin-torque mechanism with the same crystal-axis symmetry. Thus the quantitative support for 'governs' rests on a single, incompletely specified DFT number.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports transport experiments on TaIrTe4/Fe3GaTe2 van der Waals heterostructures in which current pulses along the a-axis of TaIrTe4 produce deterministic, field-free switching of the perpendicular magnetization of Fe3GaTe2 at 300 K, while current along the b-axis does not. The switching polarity is correlated with the sign of the second-harmonic nonlinear Hall voltage in devices B–E; AHE-loop-shift measurements yield an out-of-plane antidamping-like torque efficiency of about 0.16; and a TaIrTe4/Fe3GaTe2/h-BN/Fe3GaTe2 stack is operated as a magnetoresistive memory with about 2.5% TMR. First-principles tight-binding calculations are used to argue that the Berry curvature dipole of TaIrTe4 produces an orbital magnetization that exceeds the spin magnetization by a factor of 15 at the calibrated Fermi level, and the authors conclude that orbital torque governs the switching.","tokens_in":16178,"tokens_out":6353,"duration_ms":61357,"significance":"The experimental core is valuable: room-temperature field-free switching with Jc around 2×10^6 A/cm2, systematic a/b-axis and angular controls, reproducibility across devices, and a working all-electric van der Waals magnetoresistive memory are significant advances in orbitronic switching and compare favorably with prior WTe2/TaIrTe4 work. The proposed link between the nonlinear Hall sign and the switching polarity is an elegant and potentially useful design rule. However, the central mechanism claim that orbital magnetization 'governs' the switching rests on a DFT ratio whose spin channel is defined by a nonstandard formula and on a symmetry-based correlation that cannot distinguish orbital torque from spin torque. The significance is therefore conditional on fixing or softening the quantitative mechanism attribution.","major_comments":[{"comment":"The spin magnetic moment used in the calculation, mspin_nk = −⟨∂k u_nk | (1/2)g μ_b σ | ∂k u_nk⟩, is not the standard spin magnetic moment of a Bloch state, which is −μ_B ⟨u_nk|σ|u_nk⟩ with no k-derivatives. If Eq. (E4) was implemented literally, the quantity labeled α_spin is not the spin Edelstein coefficient, and the factor-of-15 comparison in Fig. 4(b) does not demonstrate that orbital magnetization dominates the spin response. If a different quantity was intended, its definition and relation to the spin torque must be given explicitly. Because this ratio is the quantitative basis for the word 'governs' in the abstract and in Section III, this point must be resolved.","section":"Appendix E, Eq. (E4)"},{"comment":"Even if the ratio Morb/Mspin ≈ 15 is correct for isolated TaIrTe4, it is a ratio of magnetizations in the semimetal, not a ratio of torques on Fe3GaTe2. The manuscript does not calculate or measure the conversion of the TaIrTe4 orbital moment into a torque at the TaIrTe4/Fe3GaTe2 interface, nor does it quantitatively rule out the coexisting spin torque or Oersted-field contribution. Since the switching loops, the θ dependence, and the loop-shift data all share the same crystal-axis symmetry, the experiments establish a correlation but not that the orbital channel 'governs'. A direct torque measurement, a ferromagnetic-thickness or interface-dependent test, or an explicit interface calculation is needed before this claim can stand.","section":"Section III and Fig. 4(b)"},{"comment":"The correlation between the sign of V^{2ω}_⊥ and the polarity of the Rxy–Ip loops is presented as experimental evidence for the orbital mechanism, but it is only a consistency check: both the nonlinear Hall effect and the current-induced orbital magnetization are predicted from the same Berry-curvature-dipole symmetry, and a spin-Edelstein or crystal-dependent spin-torque mechanism with the same BCD angular dependence would produce the same sign correlation. The data in Figs. 2, 3, and 5 therefore do not discriminate orbital torque from spin torque; the discrimination rests on the DFT ratio, which is the quantity called into question in the first major comment.","section":"Figs. 2, 3 and Section III"},{"comment":"The calculated ratio is evaluated at a Fermi level μ = 0.037 eV that is calibrated to the carrier density of a bare TaIrTe4 device (device A), whereas the switching devices are TaIrTe4/Fe3GaTe2 heterostructures in which charge transfer and band alignment may shift the Fermi level. Since Fig. 4(b) shows that the orbital-to-spin ratio can vary strongly with μ in the vicinity of the calibrated value, the uncertainty in the Fermi level is load-bearing for the quantitative claim. The authors should estimate the expected Fermi-level shift in the heterostructure or demonstrate that the 15:1 ratio is robust over a plausible range of μ.","section":"Appendix B and Fig. 4(a)"}],"minor_comments":[{"comment":"The caption states that the relaxation time is τ = 1 ps, but the orbital-to-spin magnetization ratio is independent of τ because both magnetizations are linear in τ; the caption should state this explicitly so that readers do not mistake the assumed τ for the fragile input.","section":"Fig. 4(b) caption"},{"comment":"The critical current density for 'This work' is listed as 1.3×10^6 A/cm2, while the main text reports 2.36×10^6 A/cm2 for device B and 1.38×10^6 A/cm2 for device D; the table should identify which device is quoted or list both values.","section":"Table I"},{"comment":"Reference [51] is an arXiv preprint; if a published version exists, it should be cited in its final form.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The experimental results are likely publishable even if the mechanism attribution is softened. I recommend requesting a recalculation of the spin magnetoelectric coefficient with the standard spin-moment definition, or an explicit statement that the k-derivative form is the quantity actually computed. If the recalculation is not possible, the word 'governs' should be changed to wording such as 'is consistent with' in the abstract, title, and conclusion, and the title should not claim orbital-current writing as an established fact."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The experimental part of this paper is the real story. In TaIrTe4/Fe3GaTe2 heterostructures, the authors get room-temperature field-free switching, show that the sign of the nonlinear Hall voltage predicts the switching polarity, back that correlation with angle-dependent data, and then integrate the same stack into a working vdW MTJ memory with Jc ~2×10^6 A/cm2. That is new and useful. The b-axis controls, the device-to-device reproducibility, and the exclusion of simple Joule heating (Appendix F) are all done carefully. The memory cell alone is enough to interest the vdW spintronics crowd.\n\nThe soft spot is exactly where the reader's report puts it: the claim that orbital magnetization 'governs' the switching. The experimental correlations demonstrate that the BCD symmetry tracks the switching polarity, but they do not distinguish an orbital torque from a spin torque with the same crystal symmetry. The quantitative support for orbital dominance is one DFT ratio, M_orb/M_spin ≈ 15, and the stress-test note is right that Eq. E4 defines the spin magnetic moment with a k-derivative form, which is not the standard spin expectation value -μ_B⟨σ⟩. If that formula was implemented literally, the 'spin' response is not the spin Edelstein coefficient and the 15:1 comparison is meaningless. If the standard formula was used, the manuscript should say so. Either way, the load-bearing number is incompletely documented.\n\nOne correction to the reader's worry: the ratio is independent of τ, since both magnetizations scale linearly with τ. The fragile inputs are the Fermi-level calibration (μ = 0.037 eV from carrier density) and, more importantly, the unmodeled conversion of the TaIrTe4 orbital moment into torque on Fe3GaTe2 across the interface. The 'governs' language should be softened to 'is consistent with' unless the authors add direct torque evidence or at least a sensitivity analysis.\n\nBottom line: this paper deserves a serious referee. The device demonstration is publishable and likely reproducible; the mechanism section needs honest caveats before it reaches the literature. I would cite it for the memory demonstration, and I would bring it to the reading group to argue about the orbital-torque interpretation.","headline":"Worth a serious referee: solid device work, overclaimed mechanism.","tokens_in":16709,"tokens_out":2945,"would_cite":true,"duration_ms":30827,"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":"Current along the Berry-curvature dipole writes magnet memories field-free at room temperature.","keywords":["Berry curvature dipole","orbital torque","field-free magnetization switching","Weyl semimetal TaIrTe4","van der Waals magnetoresistive memory","nonlinear Hall effect","orbital magnetoelectric effect","Fe3GaTe2"],"falsifier":"Gating a TaIrTe4/Fe3GaTe2 device so that the chemical potential crosses the point where the calculated $-\\mu D_{xz}$ term changes sign should flip the polarity of both the nonlinear Hall voltage and the field-free switching loop; if the switching polarity stays fixed while the nonlinear Hall signal changes sign, the orbital-BCD torque is not the controlling mechanism. Alternatively, a second-harmonic or spin-torque ferromagnetic resonance measurement resolving the out-of-plane torque symmetry would reveal whether a comparable coexisting spin torque is present.","tokens_in":15620,"feed_emoji":"🧲","tokens_out":5398,"duration_ms":51606,"temperature":0.7,"pith_summary":"This paper tries to establish that the Berry curvature dipole of the Weyl semimetal TaIrTe4 is the microscopic origin of the out-of-plane orbital magnetization that switches a perpendicular ferromagnet without an applied magnetic field. It reports that sending a current along the a axis, parallel to the Berry curvature dipole, drives a stable, deterministic reversal of the magnetization of an adjacent Fe3GaTe2 layer at room temperature, and that the polarity of the switch is set by the sign of the nonlinear Hall voltage. If true, this makes the Berry curvature dipole a directly measurable design parameter for a new class of all-electric, nonvolatile van der Waals magnetoresistive memories, demonstrated here with a critical current density near $2\\times10^6$ A/cm$^2$ at 290 K.","feed_headline":"Orbital currents write magnetic memory with no external field","feed_subtitle":"A Weyl semimetal's Berry curvature dipole drives field-free, room-temperature switching at about 2×10^6 A/cm².","key_machinery":"The load-bearing object is the Berry curvature dipole $D_{ij}$, the momentum-space dipole moment of the Berry curvature, defined as $D_{ij}=-\\int \\frac{d^2k}{(2\\pi)^2}\\sum_n \\partial_{k_i}\\epsilon_{nk}\\Omega^j_{nk}\\partial_{\\epsilon} f^{(0)}_{nk}$. In the low-symmetry $T_d$-phase TaIrTe4, only the $D_{xz}$ component survives, so an electric field along the a axis produces an out-of-plane orbital magnetization through the orbital magnetoelectric coefficient $\\alpha^{\\rm orb}_{xz}=-\\mu D_{xz}+\\beta_{xz}$. The paper uses first-principles tight-binding calculations to show that at the estimated Fermi level this orbital response dominates the spin response by a factor of about 15, and it uses the second-harmonic nonlinear Hall voltage, proportional to $(D\\cdot E)\\hat z\\times E$, as the electrical fingerprint of the same underlying dipole.","core_discovery":"The central claim is that in a few-layer TaIrTe4/Fe3GaTe2 van der Waals stack, an in-plane charge current converts into an out-of-plane orbital magnetization $M_{\\rm orb}$ through the orbital magnetoelectric effect, and this orbital magnetization exerts an antidamping-like torque $\\mathbf{m}_{\\rm FGT}\\times(\\mathbf{m}_{\\rm orb}\\times\\mathbf{m}_{\\rm FGT})$ that deterministically switches the perpendicular magnetization of Fe3GaTe2 without any external field. The direction of $M_{\\rm orb}$ is set by $-(D\\cdot E)\\hat z$, where $D$ is the Berry curvature dipole along the a axis, so current polarity and crystalline orientation control the final magnetization state. The paper supports this by matching the sign of the second-harmonic nonlinear Hall voltage with the clockwise or anticlockwise polarity of the switching loop, by showing the effect vanishes for current along the b axis, and by first-principles calculations giving an orbital-to-spin out-of-plane magnetization ratio of about 15 at the Fermi level $\\mu=0.037$ eV with $\\tau=1$ ps. In the integrated magnetic tunnel junction memory, field-free current pulses of $\\pm2.5$ mA write high- and low-resistance states at 290 K with a tunneling magnetoresistance ratio of about 2.6 percent and a writing current density of about $2.5\\times10^6$ A/cm$^2$.","pith_inferences":["If the orbital-BCD mechanism generalizes, the sign of the nonlinear Hall signal could serve as a design rule in other low-symmetry transition-metal dichalcogenides and Weyl semimetals, even where direct torque measurements are difficult.","A quantitative test would be to gate the TaIrTe4 Fermi level across the calculated zero crossing of $\\alpha^{\\rm orb}_{xz}$; the switching polarity should reverse at that gate voltage, separating the orbital torque from spin and Oersted contributions.","The room-temperature operation implies the Berry curvature dipole persists in the practical device geometry, so harmonic Hall detection could become an in situ calibration of torque efficiency in future memory cells."],"forward_implications":["Field-free perpendicular magnetization switching without external magnetic fields or interlayer engineering becomes possible at room temperature in van der Waals stacks, with critical current densities near $2\\times10^6$ A/cm$^2$.","The sign of the second-harmonic nonlinear Hall voltage predicts the deterministic switching polarity, giving a fast electrical probe for choosing current direction and device layout.","Integrating TaIrTe4 with Fe3GaTe2/h-BN/Fe3GaTe2 tunnel junctions yields a nonvolatile magnetoresistive memory where writing is all-electric at 290 K with a power density of about 1.35 fJ/nm$^2$.","Larger Berry curvature dipoles should give stronger out-of-plane orbital torques, making the dipole a selection metric for future orbitronic heterostructures."],"supporting_citations":[{"why":"Provides the quantum nonlinear Hall effect formula linking the Berry curvature dipole to the second-harmonic current, the basis for reading out the dipole electrically.","marker":"[43]"},{"why":"Defines the orbital Edelstein effect and the orbital magnetoelectric coefficient used to compute the current-induced orbital magnetization.","marker":"[23]"},{"why":"Establishes the Berry-phase relation between orbital magnetic moment and Berry curvature, connecting the orbital torque to band geometry.","marker":"[28]"},{"why":"Reports the room-temperature nonlinear Hall effect in TaIrTe4, the experimental result that makes the Berry curvature dipole a measurable quantity in this material.","marker":"[34]"},{"why":"Demonstrates room-temperature field-free switching of perpendicular magnetization using out-of-plane spins from TaIrTe4, the prior experimental baseline this work extends and reinterprets.","marker":"[16]"},{"why":"Shows room-temperature orbit-transfer torque enabling van der Waals magnetoresistive memories, the direct predecessor of the memory device demonstrated here.","marker":"[18]"},{"why":"Documents the large perpendicular magnetic anisotropy of Fe3GaTe2 at room temperature, the ferromagnetic layer used throughout the switching and memory experiments.","marker":"[41]"},{"why":"Shows deterministic switching of a perpendicular magnet using unconventional spin-orbit torques in WTe2, providing the comparison system for critical current density.","marker":"[14]"}],"fun_headline_variants":["Orbital currents write memory without magnetic fields","Weyl semimetal enables field-free room-temperature magnetic memory","Berry curvature dipole drives orbital-current memory switching","Low-current orbital torque switches van der Waals magnet memory","Room-temperature memory writes by orbital current in Weyl semimetal"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument hinges on the calculated claim that at the sample's Fermi level the orbital magnetoelectric response is about 15 times larger than the spin response, so the observed switching is attributed to orbital torque; this ratio rests on a density-functional model with a calibrated Fermi level and an assumed relaxation time of 1 ps, and no direct torque measurement separately resolves the orbital and spin contributions.","fun_headline_variants_meta":{"raw":{"variants":["Orbital currents write memory without magnetic fields","Weyl semimetal enables field-free room-temperature magnetic memory","Berry curvature dipole drives orbital-current memory switching","Low-current orbital torque switches van der Waals magnet memory","Room-temperature memory writes by orbital current in Weyl semimetal"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.0005,"raw_usage":{"total_tokens":2487,"prompt_tokens":1028,"completion_tokens":1459,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":644,"completion_tokens_details":{"reasoning_tokens":1379}},"tokens_in":644,"tokens_out":1459,"duration_ms":10651,"temperature":1.0,"reasoning_tokens":1379,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T23:24:06.002116+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Gating a TaIrTe4/Fe3GaTe2 device so that the chemical potential crosses the point where the calculated $-\\mu D_{xz}$ term changes sign should flip the polarity of both the nonlinear Hall voltage and the field-free switching loop; if the switching polarity stays fixed while the nonlinear Hall signal changes sign, the orbital-BCD torque is not the controlling mechanism. Alternatively, a second-harmonic or spin-torque ferromagnetic resonance measurement resolving the out-of-plane torque symmetry would reveal whether a comparable coexisting spin torque is present.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the quantum nonlinear Hall effect formula linking the Berry curvature dipole to the second-harmonic current, the basis for reading out the dipole electrically."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the orbital Edelstein effect and the orbital magnetoelectric coefficient used to compute the current-induced orbital magnetization."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the Berry-phase relation between orbital magnetic moment and Berry curvature, connecting the orbital torque to band geometry."},{"cited_title":"Son, K.-H","cited_arxiv_id":null,"evidence_quote":"Reports the room-temperature nonlinear Hall effect in TaIrTe4, the experimental result that makes the Berry curvature dipole a measurable quantity in this material."},{"cited_title":"Fukami, C","cited_arxiv_id":null,"evidence_quote":"Demonstrates room-temperature field-free switching of perpendicular magnetization using out-of-plane spins from TaIrTe4, the prior experimental baseline this work extends and reinterprets."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows room-temperature orbit-transfer torque enabling van der Waals magnetoresistive memories, the direct predecessor of the memory device demonstrated here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows deterministic switching of a perpendicular magnet using unconventional spin-orbit torques in WTe2, providing the comparison system for critical current density."}],"review_version":1}