{"id":"c9a87887-0c78-4be4-a2fc-b3f71bb4aa51","arxiv_id":"2608.07731","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Switching the ferroelectric polarization of In2Se3 in a Fe3GeTe2/In2Se3 bilayer increases the in-plane spin-orbit torque to over 150% of its original value via Fermi-surface reconstruction near the Γ point.","lead":"Computer simulations of a two-layer magnetic/ferroelectric stack show that flipping the electric polarization of the ferroelectric layer can raise the current-driven magnetic torque by more than half. This offers a nonvolatile, electrically controlled switch for spintronic devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed >150% torkance enhancement rests on an LDA-derived Γ-point band alignment; a hybrid-functional or +U recomputation is needed to confirm the pocket conversion is not an LDA artifact.","rationale":"The reader identified the LDA band alignment near Γ as the weakest assumption, and the full text supports this: the enhancement is explained entirely by the polarization-induced conversion of a Fe3GeTe2 hole pocket into an In2Se3-derived electron pocket (Section III.C, Figs. 6(c,d)). This is a genuine functional-sensitivity concern, not merely a disagreement with consensus: LDA systematically underestimates gaps and misplaces band offsets in heterostructures, and the paper offers no cross-check with a higher-level functional or experimental band-alignment data. The claim that LDA describes Fe3GeTe2 magnetic moments is irrelevant to the relative alignment with In2Se3. I therefore agree with the reader's weakest_assumption. The rest of the calculation — Wannier interpolation, k-mesh convergence, and the decomposition into even/odd and atomic contributions — is internally consistent and follows established methods. The rigid-exchange-field rotation is also an approximation, but it is common in SOT calculations and the paper at least states it; the band alignment is the more fragile premise because the whole 150% effect hinges on a specific Fermi-surface topology change. A hybrid-functional or +U recomputation of the Γ-point band structure is the single check that would settle whether this concern lands. Since my concern matches the reader's and the verdict is already CONDITIONAL, no adjustment is needed.","tokens_in":20508,"tokens_out":3779,"duration_ms":41391,"concrete_test":"Recompute the projected band structures of the up- and down-polarized Fe3GeTe2/In2Se3 heterostructures using the hybrid HSE06 functional (or, as a cheaper check, PBE+U on Fe 3d states) at the same relaxed geometry, and determine whether the Fe3GeTe2 hole pocket at Γ in the up state and the In2Se3-derived electron pocket in the down state both persist. If the pocket conversion disappears or shifts away from the Fermi level, the claimed >150% torkance enhancement and its microscopic explanation are not robust to the functional choice.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim — that reversing the In2Se3 polarization raises the in-plane total torkance to 170–180% of its up-polarization value, dominated by τ^odd_zx and a 233% Fe2 contribution — is mechanistically tied to the Section III.C band reconstruction at Γ (Figs. 6(c,d)): an Fe3GeTe2-derived hole pocket in the up state is claimed to become an In2Se3-derived electron pocket in the down state, changing the sign pattern of the k-resolved torkance. This conversion requires the In2Se3-derived bands to shift down far enough upon polarization reversal to anticross and reorder the Fermi-surface character. The LDA functional, used without Hubbard U or self-energy corrections, is known to misplace band energies and relative alignments in semiconductors and heterostructures. Validating Fe3GeTe2 magnetic moments with LDA does not validate the In2Se3/Fe3GeTe2 band alignment that is load-bearing here. If a moderate band-alignment error (of order 0.1–0.2 eV) suppresses or reverses the pocket conversion, the 150% enhancement and its Fe2 attribution would likely not survive. A secondary but related approximation, the rigid rotation of the exchange field for in-plane magnetization (Methods), is not tested against a self-consistent noncollinear calculation; while standard, it could also influence the quantitative peak value. The manuscript does not release Wannier input files or band-structure data, so an independent check of the critical Γ-point ordering is not currently possible.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports first-principles calculations of the spin-orbit torkance in a Fe3GeTe2/In2Se3 heterobilayer for the two polarization states of the In2Se3 layer. Using DFT with LDA, Wannier interpolation, and Kubo linear response (Eqs. 2–5), the authors find that switching the In2Se3 polarization from up to down increases the total in-plane torkance to about 170–180% of its up-polarization value, dominated by the time-reversal-odd τ_zx component and by a 233% increase in the Fe2 atom-resolved contribution. They interpret this as a polarization-induced Fermi-surface reconstruction near Γ, where an Fe3GeTe2-derived hole pocket is converted into an In2Se3-derived electron pocket. The paper also presents atom-resolved Edelstein responses and argues that atom-resolved torkance cannot be inferred directly from Edelstein coefficients.","tokens_in":20771,"tokens_out":9612,"duration_ms":87501,"significance":"If correct, the prediction establishes an all-van-der-Waals multiferroic heterostructure in which ferroelectric polarization nonvolatilely controls SOT, with a concrete microscopic mechanism. The strengths of the paper are its clearly specified computational parameters (18×18×1 relaxation mesh, 600×600 torkance mesh, 90/800 Ry cutoffs, Γ=0.01 eV), the Wannier-interpolated k-resolved and atom-resolved decompositions, and the Edelstein cross-check. The calculation is not circular: the torkance is computed from first-principles Wannier Hamiltonians and the broadening is taken from prior work rather than fitted to the target enhancement. The main risk is the unvalidated LDA band alignment on which the pocket conversion near Γ rests.","major_comments":[{"comment":"The central quantitative claim — that polarization reversal raises the in-plane total torkance to 170–180% of its up-state value — is tied to the band reconstruction at Γ described in Section III.C and shown in Figs. 6(c,d). The mechanism requires the In2Se3-derived bands to shift downward enough to anticross with Fe3GeTe2-derived bands and convert a hole pocket into an electron pocket. This ordering is computed with LDA, and the paper provides no validation of the heterostructure band alignment (e.g., HSE06 or DFT+U calculation, or comparison with photoemission). Since LDA is known to misplace semiconductor band edges by several tenths of an eV, and since a moderate error of order 0.1–0.2 eV could suppress or reverse the pocket conversion, the 150% enhancement and its Fe2 attribution are not yet established beyond the LDA approximation. The authors should either add a band-alignment validation or explicitly reframe the result as LDA-dependent.","section":"III.C and Figs. 6(c,d)"},{"comment":"The angular dependence of the torkance is computed by manually rotating the exchange field H_odd from the +z ground state to arbitrary (θ,φ) without a self-consistent noncollinear calculation. Because the headline result is for in-plane magnetization (θ=90° or φ=0°), the quantitative values of the enhancement depend on the validity of this rigid-rotation approximation. This approximation is common, but it is load-bearing for the claimed 170–180% figure. The authors should report at least one self-consistent noncollinear calculation with the magnetization along the in-plane direction, or otherwise demonstrate that the rigid rotation does not change the electronic structure enough to affect the enhancement.","section":"Methods"}],"minor_comments":[{"comment":"In the definition of τ_θx, the text gives τ_θx = τ_xx cosθ − τ_zx sinθ and also states e_θ = (cosθ, 0, sinθ)^T; the torque component along e_θ would be τ_xx cosθ + τ_zx sinθ, so one of the two definitions contains a sign error. Please correct the inconsistency.","section":"III.B"},{"comment":"The text defines the Edelstein response as χ_even_yx, while the caption of Fig. 7(b) labels it χ_even_zx; these should be reconciled.","section":"III.D and Fig. 7(b)"},{"comment":"The sentence “For both calculation calculation of torkance and Edelstein effect” contains a duplicated word.","section":"Methods"},{"comment":"The sentence “which is shown in Figs. 9 and 7” appears to reference the wrong figures; the polarization-up and polarization-down stacking energies are shown in Figs. 8 and 9.","section":"Appendix A"},{"comment":"The definition of the torque operator relies on H_odd, which is only described by reference to the authors' preprint [102]; for a self-contained publication, the essential definition of H_odd and the atom-resolved projection should be stated explicitly.","section":"Methods"},{"comment":"The manuscript does not include Wannier Hamiltonians, band-structure data, or input files for the two polarization states; providing these (or a data availability statement) would allow independent verification of the critical Γ-point band ordering.","section":"Data availability"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a solid first-principles prediction, but the central enhancement hinges on an LDA band alignment that is not independently validated. If the authors can add a hybrid-functional or +U check of the Γ-point ordering and a self-consistent noncollinear point at θ=90°, I would be happy to see a revised version. The paper is within the scope of cond-mat.mes-hall and the computational details are transparent. I also note that the method draws heavily on the authors' own preprint [102]; this is not circular, but the lack of released Wannier data makes independent checking difficult."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nQuick take: this is a serious computational paper that makes a specific, falsifiable prediction — reversing In2Se3 polarization in Fe3GeTe2/In2Se3 changes the in-plane SOT by roughly 70–80%, driven by a Gamma-point Fermi-surface reconstruction and a 233% boost in the Fe2-layer contribution. That precise mechanism is new, and the analysis that gets there is the paper's real strength: k-resolved and atom-resolved torkance decompositions, a clean separation of time-reversal-even and odd contributions, and an honest discussion of why atom-resolved Edelstein responses do not track torkance. The methods are standard and the numerics are well specified (18x18x1 relaxation mesh, 600x600 torkance mesh, 90/800 Ry cutoffs, Gamma = 0.01 eV). I do not see circular fitting; the torkance comes from Wannier Hamiltonians, not from tuning parameters to hit 150%.\n\nWhere I would push back: the load-bearing point is the LDA band alignment near Gamma. The claim that the Fe3GeTe2 hole pocket turns into an In2Se3-derived electron pocket after polarization reversal is exactly the sort of level-crossing that LDA can misplace by 0.1–0.2 eV. The paper cites LDA's success for Fe3GeTe2 magnetic moments, but that does not validate the heterostructure band offset. A hybrid-functional or +U band-structure test, or at least a rigid-shift sensitivity check on the torkance, would tell us how much of the 170% survives. The rigid exchange-field rotation used for in-plane magnetization is also untested against a self-consistent noncollinear calculation — probably fine, but it deserves one explicit sentence of justification. And no Wannier input files or band-structure data are released, so the critical Gamma-point ordering cannot currently be checked by others.\n\nThese are real but ordinary gaps for a computational prediction paper. The central physics is plausible and the mechanism narrative is coherent; I do not see a hidden fatal flaw. It deserves a serious referee and, conditionally, publication after the functional-sensitivity point is addressed.\n\nRecommendation: send to peer review. Ask the authors for a band-alignment validation, a statement on the rigid-rotation approximation, and release of the computational inputs.","headline":"A solid first-principles prediction of ferroelectric-controlled SOT in Fe3GeTe2/In2Se3, with a plausible Gamma-point mechanism; the quantitative headline depends on a band alignment that LDA alone does not validate.","tokens_in":21381,"tokens_out":1740,"would_cite":false,"duration_ms":16764,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"In a Fe3GeTe2/In2Se3 van der Waals heterostructure, reversing the In2Se3 ferroelectric polarization boosts the in-plane spin-orbit torkance to more than 150% (170-180%) of its original value.","keywords":["spin-orbit torque","torkance","ferroelectric control","van der Waals heterostructure","Fe3GeTe2","In2Se3","Fermi surface reconstruction","first-principles calculation"],"falsifier":"Angle-resolved photoemission on a poled Fe3GeTe2/In2Se3 stack should show the predicted reconstruction of the Fermi surface near the $\\Gamma$ point from a Fe3GeTe2-derived hole pocket (polarization up) to an In2Se3-derived electron pocket (polarization down); if the pocket conversion is absent, the 170-180% torkance enhancement would not occur as calculated.","tokens_in":20264,"feed_emoji":"🧲","tokens_out":11213,"duration_ms":92648,"temperature":0.7,"pith_summary":"This paper predicts that in a two-dimensional van der Waals heterostructure made of the ferromagnet Fe3GeTe2 and the ferroelectric In2Se3, flipping the ferroelectric polarization of In2Se3 changes the strength of the current-induced spin-orbit torque. When the magnetization lies in the plane, where the torque is largest, reversing the polarization from up to down raises the total torkance (torque per electric field) to about 170-180% of its original value. The change is dominated by the time-reversal-odd, field-like component of the torkance, and the largest atomic contribution comes from the middle Fe layer, which grows by roughly 233%. The microscopic origin is a polarization-driven reconstruction of the Fermi surface near the $\\Gamma$ point, where an Fe3GeTe2-derived hole pocket turns into an In2Se3-derived electron pocket. If the prediction holds, ferroelectric switching becomes a nonvolatile, electrical control knob for spin-orbit torque in all-van der Waals spintronic devices.","feed_headline":"Reversing a ferroelectric boosts 2D spin-orbit torque to 180 percent","feed_subtitle":"Flipping In2Se3 polarization enlarges in-plane torque via a Fermi-surface pocket change at the Gamma point.","key_machinery":"The central object is the torkance tensor $\\boldsymbol{\\tau}$, which relates the current-induced torque to the applied electric field. Calculated from a Wannier Hamiltonian by linear-response Kubo formulas, it splits into a time-reversal-even interband part (damping-like) and a time-reversal-odd intraband part (field-like), with the latter dominating here. The analysis then decomposes $\\tau^{\\mathrm{odd}}_{zx}$ in momentum space (momentum-resolved torkance arcs around $\\Gamma$) and by atomic site; the load-bearing step is the Fermi-surface reconstruction in which polarization-down brings In2Se3-derived bands down to hybridize with Fe3GeTe2 states and converts a hole pocket into an In2Se3-derived electron pocket. That pocket's all-negative arc-like torkance contributions are what make the total torkance grow.","core_discovery":"The authors show that the Fe3GeTe2/In2Se3 heterobilayer, which has $C_{3v}$ symmetry and therefore supports a finite torkance even for out-of-plane magnetization, acquires a strongly polarization-dependent torkance. With the current along $+x$ and the magnetization in the $x$-$y$ plane, switching In2Se3 from the upward to the downward polarization state increases the total torkance by about 0.22-0.30 $e a_0$, reaching approximately 170-180% of the upward-state value. The dominant component is the $z$-directed time-reversal-odd torkance $\\tau^{\\mathrm{odd}}_{zx}$, which is field-like and about an order of magnitude larger than the time-reversal-even part; its change is carried mainly by the middle Fe layer (Fe2), whose atom-resolved contribution grows from -0.09 to -0.30 $e a_0$, a 233% increase. The mechanism is a Fermi-surface reconstruction near $\\Gamma$: in the polarization-down state the In2Se3-derived bands shift downward and hybridize with Fe3GeTe2 states, converting the Fe3GeTe2 hole pocket into an In2Se3-derived electron pocket whose momentum-resolved torkance arcs are all negative and add to $\\tau^{\\mathrm{odd}}_{zx}$.","pith_inferences":["The reported percentages are torkance ratios at fixed electric field; the current-density ratio could differ because polarization switching also changes the conductivity, so device-level energy efficiency needs a separate transport estimate.","The band-alignment mechanism suggests a screening rule the authors do not state: any ferroelectric whose bands can be pushed through the Fermi level of a two-dimensional ferromagnet by polarization reversal should show a similar torque modulation, not just In2Se3.","A direct test is to measure the $\\Gamma$-point Fermi pocket by angle-resolved photoemission on poled samples: the predicted hole-to-electron pocket conversion should be visible before any torque measurement.","The authors' finding that atom-resolved Edelstein and torkance responses do not track each other implies that spin-accumulation or spin-current probes alone would underestimate the torque change; only torque measurements (for example harmonic Hall) would reveal the 170-180% effect."],"forward_implications":["In-plane magnetization is where both the torque and its ferroelectric modulation are largest, so devices exploiting this effect should be designed with the magnetization along the $x$ direction.","Because the time-reversal-odd, field-like component dominates, the enhanced torque acts as an effective magnetic field that drives precession rather than as a damping-like torque that relaxes the magnetization.","The middle Fe layer (Fe2) is the active magnetic site, so tuning its local environment through stacking, strain, or intercalation should be the most direct way to engineer the effect.","The mechanism requires In2Se3-derived bands near $\\Gamma$ to cross the Fermi level on polarization reversal, which gives a concrete band-structure criterion for screening other ferroelectric/ferromagnet pairs for switchable spin-orbit torque.","Since the two polarization states are nonvolatile, the same device can store a state and read it out through the torque magnitude, offering a route to programmable logic without continuous power."],"supporting_citations":[{"why":"Establishes monolayer Fe3GeTe2 as a realistic two-dimensional ferromagnet whose magnetic properties the DFT setup reproduces.","marker":"[24]"},{"why":"Provides the monolayer Fe3GeTe2 torkance and its angular dependence, the baseline the heterostructure is compared against.","marker":"[59]"},{"why":"Supplies the first-principles torkance formulas used to compute the time-reversal-even and time-reversal-odd components.","marker":"[86]"},{"why":"Gives the atom-resolved and momentum-resolved torkance decomposition that identifies the Fe2 layer as the dominant source of the change.","marker":"[102]"},{"why":"Supplies the intrinsic ferroelectricity of In2Se3 that makes the switchable polarization possible.","marker":"[106]"},{"why":"Provides the van der Waals correction used to bind the Fe3GeTe2 and In2Se3 monolayers in the calculation.","marker":"[98]"}],"fun_headline_variants":["Ferroelectric toggle ramps up 2D spin-orbit torque to 180%","Flip ferroelectric to boost 2D spin-orbit torque by 80%","Ferroelectric polarization switch: 2D SOT up to 180%","Switch ferroelectric, 2D spin-orbit torque hits 180%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculation assumes the local-density-approximation energy alignment between the Fe3GeTe2 and In2Se3 bands near the $\\Gamma$ point; if a more accurate treatment shifts those bands so that the polarization-down state no longer converts the Fe3GeTe2 hole pocket into an In2Se3 electron pocket, the 150% enhancement could shrink or change sign.","fun_headline_variants_meta":{"raw":{"variants":["Ferroelectric toggle ramps up 2D spin-orbit torque to 180%","Flip ferroelectric to boost 2D spin-orbit torque by 80%","Ferroelectric polarization switch: 2D SOT up to 180%","Switch ferroelectric, 2D spin-orbit torque hits 180%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001429,"raw_usage":{"total_tokens":5882,"prompt_tokens":1179,"completion_tokens":4703,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":795,"completion_tokens_details":{"reasoning_tokens":4613}},"tokens_in":795,"tokens_out":4703,"duration_ms":34257,"temperature":1.0,"reasoning_tokens":4613,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T00:20:52.598718+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Angle-resolved photoemission on a poled Fe3GeTe2/In2Se3 stack should show the predicted reconstruction of the Fermi surface near the $\\Gamma$ point from a Fe3GeTe2-derived hole pocket (polarization up) to an In2Se3-derived electron pocket (polarization down); if the pocket conversion is absent, the 170-180% torkance enhancement would not occur as calculated.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the monolayer Fe3GeTe2 torkance and its angular dependence, the baseline the heterostructure is compared against."},{"cited_title":"Freimuth, S","cited_arxiv_id":null,"evidence_quote":"Supplies the first-principles torkance formulas used to compute the time-reversal-even and time-reversal-odd components."},{"cited_title":"Anatomy of Spin--Orbit Torques in Monolayer Fe$_3$GeTe$_2$ and Fe$_3$GaTe$_2$: Insights from atomistic and momentum-space decompositions","cited_arxiv_id":"2608.05788","evidence_quote":"Gives the atom-resolved and momentum-resolved torkance decomposition that identifies the Fe2 layer as the dominant source of the change."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the intrinsic ferroelectricity of In2Se3 that makes the switchable polarization possible."},{"cited_title":"Grimme, J","cited_arxiv_id":null,"evidence_quote":"Provides the van der Waals correction used to bind the Fe3GeTe2 and In2Se3 monolayers in the calculation."}],"review_version":1}