{"id":"ecce364a-aa8a-4eee-ab9f-4efa8ec5b7e1","arxiv_id":"1908.09367","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A gate voltage switches the sign of a room-temperature spin-to-charge conversion signal in graphene/MoTe2 heterostructures, with a claimed efficiency of about 7.6 percent.","lead":"The paper reports a room-temperature electrical conversion of spin current into charge voltage in a stack of graphene and the semimetal MoTe2, with the signal's sign and size tuned by a gate voltage. If correct, this is a step toward gate-controlled spintronic memory and logic that does not need magnetic fields.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Missing thermal control: the gate-dependent sign reversal near CNP could be the graphene Seebeck coefficient changing sign, not IREE spin-to-charge conversion.","rationale":"The reader's CONDITIONAL verdict already identifies the absence of a direct spin-accumulation measurement as the weakest assumption. I sharpen this to a concrete, likely artifact: thermoelectric voltages in graphene are large at room temperature and change sign at the CNP, matching the paper's headline gate-controlled switching. The published controls do not eliminate this path, because a spin-dependent Peltier/Seebeck signal is linear in current, reverses with current polarity, follows cosθ, and shows no Hanle precession. The central claim's quantitative efficiency and uniqueness are only as good as the spin origin; a thermal origin would remove both. I am not asserting that the effect is thermal, only that the missing nonmagnetic-injector or Seebeck control is load-bearing. Therefore I keep the reader's CONDITIONAL verdict, with the thermal/nonmagnetic-injector control as an explicit condition for acceptance.","tokens_in":9205,"tokens_out":22207,"duration_ms":260023,"concrete_test":"Fabricate a control device with the same Hall-bar/MoTe2 geometry and contact resistances but replace the Co/TiO2 spin injector with a Ti/Au contact, and repeat the nonlocal B_x sweep, current-polarity, angle-dependence, and gate-voltage measurements at identical currents. If a ΔVSGE of comparable magnitude with the same gate sign reversal persists, the claimed signal is a thermoelectric/magnetotransport artifact; if it vanishes, the spin-origin interpretation would be supported. As an analytical cross-check, directly measure the Seebeck coefficient S(Vg) of the graphene/MoTe2 channel and compare the Vg at which S changes sign with the Vg at which ΔRSGE changes sign; coincidence within the CNP region would leave the thermal explanation unresolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"At room temperature the measured ΔVSGE is a nonlocal voltage appearing across Ti/Au contacts while 80 μA flows through a Co/TiO2 injector; the central attribution to inverse Rashba-Edelstein conversion in the graphene/MoTe2 region rests on excluding spurious thermal and magnetotransport signals. The reported controls (cosθ angle dependence, absence of Hanle peaks, current-polarity reversal) do not exclude a thermoelectric origin: a spin-dependent Peltier/Seebeck voltage is linear in the injection current, reverses with current polarity, follows the injector magnetization projection (cosθ), and shows no Hanle precession because no spin accumulation is involved. In particular, the gate-controlled sign reversal near the CNP is exactly the signature expected if the signal were S(Vg)·ΔT, since the Seebeck coefficient of graphene changes sign at the charge neutrality point; the paper provides no device with a nonmagnetic injector, no direct S(Vg) measurement, and no second-harmonic thermal check. Until such a control is performed, the data do not uniquely support IREE over a thermal/magnetotransport artifact.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports nonlocal spin-to-charge conversion (inverse Rashba-Edelstein, IREE) in 1T'-MoTe2/graphene van der Waals heterostructures at room temperature. In a Hall-bar device, a spin current injected from a Co/TiO2 ferromagnetic contact diffuses into the graphene/MoTe2 region, and a nonlocal voltage VSGE is detected across Ti/Au contacts as a function of in-plane magnetic field, bias current, and back-gate voltage. The authors report ΔRSGE ≈ 4.96 Ω, estimate a lower-bound conversion efficiency αRE ≈ 7.6%, observe cosθ angular dependence of the signal, and demonstrate gate-controlled sign reversal near the charge neutrality point. They interpret these observations as proximity-induced Rashba-Edelstein spin-to-charge conversion with sign controlled by electron versus hole doping.","tokens_in":9397,"tokens_out":5666,"duration_ms":54204,"significance":"If the interpretation is correct, the result is significant: it would demonstrate room-temperature, gate-tunable IREE in a large-area CVD-graphene/MoTe2 heterostructure with a conversion efficiency larger than previous 2D systems, which is relevant for spin-orbit-torque memory and logic applications. The paper deserves credit for using nonmagnetic detectors to avoid stray Hall effects at the ferromagnetic detector, for providing angle-dependence and bias-polarity controls, for reporting measurements on two devices, and for explicitly stating assumptions and the lower-bound nature of the efficiency estimate. However, the central attribution to IREE rests on excluding thermoelectric and other spurious signals, and the quantitative claims rely on several unmeasured parameters; both are addressable with additional control experiments and sensitivity analysis.","major_comments":[{"comment":"The central claim that VSGE is an inverse Rashba-Edelstein signal is not uniquely established because a spin-dependent Peltier/Seebeck thermoelectric voltage in the same nonlocal geometry reproduces all three reported controls: it is linear in injection current, reverses with current polarity, follows the injector magnetization as cosθ, and shows no Hanle precession since no spin accumulation is involved. The gate-dependent sign reversal near the charge neutrality point (Fig. 4b) is precisely the expected signature of the graphene Seebeck coefficient changing sign at the CNP. The manuscript includes no nonmagnetic-injector control device, no direct measurement of S(Vg), and no second-harmonic thermal check, so the data do not exclude a thermal origin. This control is load-bearing for the conclusion that the effect is spin-galvanic.","section":"Spin-galvanic measurement geometry (Fig. 1a-d, Fig. 2)"},{"comment":"The efficiency estimate αRE = 7.6% depends on assumed values P = 10% for the Co/TiO2 contact polarization and λG = 2.65 μm taken from pristine graphene, while the spin accumulation in the graphene/MoTe2 region is not directly measured (no spin signal is detected through the heterostructure). Since P and λG enter Eq. (1) multiplicatively and exponentially, the 'order of magnitude larger' comparison with other systems is not quantitatively secure without a sensitivity analysis or independent measurements. The text states this is a lower bound because the heterostructure spin diffusion length is shorter, but the bound also depends on the assumed P; please state the range of αRE consistent with plausible parameter values and add error bars to ΔRSGE.","section":"Eq. (1) and efficiency estimate"},{"comment":"The interpretation that the sign reversal at Vg ≈ 10-20 V reflects Rashba spin textures in conduction versus valence bands is imported from prior theory and used post-hoc; it is consistent with the data but not uniquely determined. Moreover, the gate dependence of the injector contact polarization is only partially controlled (Fig. S6 is cited but not shown in the main text). To support the claim that the sign change is inherent to IREE in the heterostructure, the authors should show the gate-dependent Hanle/spin-valve controls in the main text and discuss how a thermoelectric Seebeck sign change is excluded.","section":"Gate-dependent sign change (Fig. 3, Fig. 4)"}],"minor_comments":[{"comment":"The title uses '1T-MoTe2' while the text and abstract use '1T′-MoTe2'; please use the prime consistently.","section":"Title and Abstract"},{"comment":"The data are shifted and a linear background is subtracted; the subtraction procedure and raw traces should be described in the Methods or caption so the reader can assess the magnitude of the background relative to ΔVSGE.","section":"Fig. 1d and Fig. 2b captions"},{"comment":"These panels have no error bars; please add them or state the measurement uncertainty, especially because only two devices are reported.","section":"Fig. 2c and Fig. 4b"},{"comment":"The phrase 'the measurements were manifested by varying the angle' is unclear; rephrase as 'the angle θ of the in-plane magnetic field was varied.'","section":"Fig. 2a text"},{"comment":"Several references are cited as arXiv preprints (refs 12, 20, 21, 30, 35, 36); if published versions exist, cite them.","section":"References"},{"comment":"The symbols ΔVSGE and ΔRSGE are introduced in figure captions before their first use in the text; define them at first mention in the main text.","section":"Notation"}],"recommendation":"major_revision","confidential_remarks":"The thermal-control issue is the decisive point for this manuscript. If the authors can perform a nonmagnetic-injector control device and a second-harmonic or thermopower measurement showing that the gate-dependent signal is not thermoelectric, the central claim would be substantially strengthened. I also recommend that the editors ask the authors to clarify the novelty and quantitative comparison relative to refs 35 and 36, since the current 'orders of magnitude larger' claim is based on Eq. (1) with assumed parameters."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read it. The short version: this is a plausible room-temperature demonstration of gate-tunable spin-to-charge conversion in graphene/MoTe2, with better controls than much of the vdW spintronics literature. But the attribution to inverse Rashba-Edelstein is not airtight: there is no thermal control, and the gate sign reversal near the CNP looks like exactly what a graphene Seebeck change would produce. The numbers also lean on assumptions that could move the efficiency by a lot.\n\nWhat's genuinely new is the combination of semimetal 1T'-MoTe2 with large-area CVD graphene, and the room-temperature gate switching of the nonlocal signal. The measurements include angular dependence following cosθ, linear bias dependence with sign reversal, gate dependence with sign change near the CNP, plus control spin-valve and Hanle data showing the injector polarization is not flipping. For a two-device study, that is a solid set of controls.\n\nThe soft spots are real but not disqualifying on their own. Only two devices, no error bars. The efficiency α_RE uses an assumed 10% contact polarization and a spin diffusion length taken from pristine graphene; both are defensible as lower-bound choices, but the resulting 7.6% should be read as an order-of-magnitude estimate rather than a measured number. Data available only on reasonable request is also weak.\n\nThe bigger issue is the one the paper does not address: thermoelectric spurious signals. A spin-dependent Seebeck/Peltier effect can produce a nonlocal voltage that is linear in current, follows cosθ, and shows no Hanle peaks, exactly as observed. The gate sign reversal near the CNP is the classic signature of graphene's Seebeck coefficient changing sign. The argument that non-magnetic contacts avoid this is not sufficient, since a thermal gradient through the graphene channel would be picked up by any contacts. A nonmagnetic injector, a direct S(Vg) measurement, or a second-harmonic thermal check would close this.\n\nThe 'unprecedented' and 'order of magnitude larger' framing is also weakened by the concurrent preprints the authors themselves cite (Refs 35 and 36). They are honest about this in the note, but the abstract still overstates.\n\nWho is this for? People working on graphene/TMD spintronics and charge-spin conversion. It is a useful data point, not a paradigm shift. I would send it to peer review: the controls are good enough that a serious referee should see the raw data and push for the thermal control. I would not cite the quantitative efficiency until that is done.","headline":"A plausible room-temperature gate-tunable spin-galvanic signal in graphene/MoTe2, but the missing thermal control and parameter-dependent efficiency claims keep it from being fully convincing.","tokens_in":9973,"tokens_out":4529,"would_cite":false,"duration_ms":50120,"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 graphene/MoTe2 heterostructure converts spin current into a gate-switchable charge voltage at room temperature.","keywords":["spin-galvanic effect","inverse Rashba-Edelstein effect","graphene/MoTe2 heterostructure","spin-to-charge conversion","proximity-induced spin-orbit coupling","gate-controlled sign switching","van der Waals heterostructure","Weyl semimetal MoTe2"],"falsifier":"A control device identical except with the MoTe2 flake omitted, or with a non-magnetic injector contact, would settle the claim: if the same non-local voltage appears in either case, or if the signal fails to reverse when the injector magnetization is reversed, the spin-galvanic interpretation is wrong. Conversely, resolving a Hanle spin-precession peak near zero field in the heterostructure region would indicate out-of-plane spins and point toward a spin-Hall rather than Rashba–Edelstein mechanism.","tokens_in":9000,"feed_emoji":"⚡","tokens_out":9409,"duration_ms":84362,"temperature":0.7,"pith_summary":"This paper reports that a heterostructure of semimetallic MoTe2 and monolayer graphene converts an injected spin current into a measurable charge voltage at room temperature, through a proximity-induced inverse Rashba–Edelstein effect. The conversion signal is large—about $\\Delta R_{\\mathrm{SGE}}/I \\approx 4.96\\ \\Omega$ in resistance units, with a lower-bound efficiency of $\\alpha_{\\mathrm{RE}} \\approx 7.6\\%$—and its sign reverses when a back-gate voltage moves the Fermi level across the charge neutrality point. That gate-controlled switching is interpreted as a consequence of the Rashba spin-split bands that graphene acquires from MoTe2: for the same spin texture, electrons and holes accumulate with opposite charge signs, reversing the induced charge current. The measurements use a non-local geometry with non-magnetic detection contacts, chosen because it avoids stray-Hall artifacts that can mimic spin-galvanic signals in local measurements. If correct, the result shows that an electric field alone can create and switch a spin-galvanic response in a van der Waals heterostructure at room temperature, a functionality relevant to spintronic memory and logic.","feed_headline":"Gate voltage flips the spin-galvanic signal at room temperature","feed_subtitle":"Gate-tunable spin-to-charge conversion in a MoTe2/graphene stack, over ten times larger than earlier 2D devices.","key_machinery":"The load-bearing mechanism is the inverse Rashba–Edelstein effect (IREE), the reciprocal of the Rashba–Edelstein effect: a spin accumulation in a Rashba spin-split band produces a charge current perpendicular to both the spin direction and the interface normal, $I_c \\propto \\hat{z} \\times n_s$. In this device the Rashba splitting is not intrinsic to graphene but is induced by proximity to MoTe2, giving graphene spin-split conduction and valence bands with the same spin chirality. The non-local Hall-bar geometry separates spin injection (ferromagnetic contact) from detection (non-magnetic contacts), and the gate voltage tunes the Fermi level between electron- and hole-doped regimes, flipping the sign of the detected voltage because the same spin texture produces opposite charge accumulation for electrons and holes. The relation $V_{\\mathrm{SGE}} \\propto \\hat{z} \\times n_s$, the $\\cos(\\theta)$ angle dependence, and the absence of zero-field Hanle peaks together identify the signal as IREE rather than a bulk or proximity spin-Hall effect.","core_discovery":"The central claim is that the graphene/MoTe2 interface acts as a room-temperature spin-to-charge converter governed by the inverse Rashba–Edelstein effect (IREE). In the non-local measurement, a ferromagnetic Co/TiO2 contact injects spin into a graphene channel; the spins diffuse into the graphene/MoTe2 region, where the proximity-induced Rashba spin texture converts their accumulation into a transverse charge voltage detected by Ti/Au contacts. The paper reports $\\Delta R_{\\mathrm{SGE}}/I \\approx 4.96\\ \\Omega$ at a bias current of $-80\\ \\mu\\mathrm{A}$, a lower-bound conversion efficiency $\\alpha_{\\mathrm{RE}} \\approx 7.6\\%$, and, most distinctively, a reversal of the voltage sign as the gate sweeps through the charge neutrality point between hole and electron doping. Angle-dependent measurements follow a $\\cos(\\theta)$ dependence on the in-plane field direction, and the absence of Hanle peaks near zero field is presented as evidence that the accumulated spins are in-plane, ruling out a proximity spin-Hall origin. The sign change is reproduced in two devices, and control measurements indicate that it is not caused by a change in the injector's spin polarization.","pith_inferences":["Editorial inference: the stated 7.6% efficiency is a lower bound; if the spin diffusion length inside the heterostructure were measured directly rather than borrowed from pristine graphene, the conversion efficiency could turn out substantially higher.","Editorial inference: the same gate-switching logic should be testable in other semimetal/TMD–graphene stacks with different band alignments, where the sign-change position relative to the charge neutrality point would map the relative strength of proximity Rashba versus intrinsic band hybridization.","Editorial inference: the strong spin relaxation that killed the Hanle signal in the heterostructure region could itself be engineered as a tunable spin sink, letting lateral spintronic circuits erase spin information with a gate pulse.","Editorial inference: a direct comparison of identical devices with and without MoTe2, and with MoTe2 thickness varied, would separate the interface proximity contribution from any bulk semimetal contribution to the spin-to-charge conversion."],"forward_implications":["Room-temperature, all-electrical spin-to-charge conversion with gate-controlled sign provides a building block for spin-orbit-torque magnetic random-access memory, where the write current direction could be set by a gate voltage rather than by reversing a magnetic field.","The lower-bound efficiency of 7.6%, about an order of magnitude above earlier graphene/TMD heterostructures, implies that much smaller charge currents could produce usable spin-orbit torques in devices built on this platform.","Because the effect appears in large-area CVD graphene, the mechanism is compatible with wafer-scale fabrication rather than requiring exfoliated crystals.","The Fermi-level-dependent sign reversal confirms that the proximity-induced spin texture in graphene is gate-tunable, so the same stacking principle could be used to build electrically reconfigurable spin logic circuits."],"supporting_citations":[{"why":"Supplies the non-local injection-detection geometry and the efficiency formula used to estimate $\\alpha_{\\mathrm{RE}}$.","marker":"[11]"},{"why":"Provides the charge-current/spin-density relation $V_{\\mathrm{SGE}} \\propto \\hat{z} \\times n_s$ used to analyze the signal direction.","marker":"[12]"},{"why":"Earlier 2D heterostructure IREE benchmark whose signal size and gate behavior this paper compares with and extends.","marker":"[13]"},{"why":"Establishes spin-polarized surface and bulk states in MoTe2 at room temperature, motivating the choice of MoTe2.","marker":"[19]"},{"why":"Prediction that graphene in proximity to TMDs acquires strong spin-orbit coupling, supporting the proximity-induced SOI interpretation.","marker":"[23]"},{"why":"Theory that graphene on TMDs has Rashba spin-split conduction and valence bands with the same spin chirality, basis for the gate sign change.","marker":"[26]"},{"why":"Identifies MoTe2 as a type-II Weyl semimetal, connecting the material's topological band structure to the observed effect.","marker":"[27]"},{"why":"Source and fabrication route for the large-area CVD graphene used in the devices.","marker":"[30]"}],"fun_headline_variants":["Gate voltage reverses spin-galvanic effect at room temperature","Record spin-to-charge conversion in MoTe2/graphene stack","Electric control of spin signal in 2D heterostructure","Room-temperature spin switching via gate in MoTe2/graphene","All-electrical spin-galvanic switch in 2D material bilayer"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central claim rests on the assumption that the non-local voltage detected by the non-magnetic contacts is generated by inverse Rashba–Edelstein conversion of in-plane spin accumulation inside the graphene/MoTe2 region, rather than by spurious local magnetotransport, thermal gradients, or out-of-plane spin-Hall signals.","fun_headline_variants_meta":{"raw":{"variants":["Gate voltage reverses spin-galvanic effect at room temperature","Record spin-to-charge conversion in MoTe2/graphene stack","Electric control of spin signal in 2D heterostructure","Room-temperature spin switching via gate in MoTe2/graphene","All-electrical spin-galvanic switch in 2D material bilayer"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000314,"raw_usage":{"total_tokens":1835,"prompt_tokens":1051,"completion_tokens":784,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":667,"completion_tokens_details":{"reasoning_tokens":694}},"tokens_in":667,"tokens_out":784,"duration_ms":7986,"temperature":1.0,"reasoning_tokens":694,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:14:12.866479+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A control device identical except with the MoTe2 flake omitted, or with a non-magnetic injector contact, would settle the claim: if the same non-local voltage appears in either case, or if the signal fails to reverse when the injector magnetization is reversed, the spin-galvanic interpretation is wrong. Conversely, resolving a Hanle spin-precession peak near zero field in the heterostructure region would indicate out-of-plane spins and point toward a spin-Hall rather than Rashba–Edelstein mechanism.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the non-local injection-detection geometry and the efficiency formula used to estimate $\\alpha_{\\mathrm{RE}}$."},{"cited_title":"Observation of Spin Hall Effect in Weyl Semimetal WTe2 at Room Temperature","cited_arxiv_id":"1812.02113","evidence_quote":"Provides the charge-current/spin-density relation $V_{\\mathrm{SGE}} \\propto \\hat{z} \\times n_s$ used to analyze the signal direction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes spin-polarized surface and bulk states in MoTe2 at room temperature, motivating the choice of MoTe2."},{"cited_title":"W., Garcia, J","cited_arxiv_id":null,"evidence_quote":"Prediction that graphene in proximity to TMDs acquires strong spin-orbit coupling, supporting the proximity-induced SOI interpretation."},{"cited_title":"& Ferreira, A","cited_arxiv_id":null,"evidence_quote":"Theory that graphene on TMDs has Rashba spin-split conduction and valence bands with the same spin chirality, basis for the gate sign change."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Identifies MoTe2 as a type-II Weyl semimetal, connecting the material's topological band structure to the observed effect."},{"cited_title":"Two-Dimensional Spintronic Circuit Architectures on Large Scale Graphene","cited_arxiv_id":"1905.04151","evidence_quote":"Source and fabrication route for the large-area CVD graphene used in the devices."}],"review_version":1}