{"id":"de3e5ec5-d28d-4b7a-95fb-9d084968ec92","arxiv_id":"2606.01308","paper_version":2,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A 4 nm Gd interlayer in Ti/Co systems produces fivefold higher spin-orbit torque efficiency than reference bilayers, with trilayer torque efficiency exceeding 1 due to enhanced orbital-to-spin conversion at the rare-earth interface.","lead":"The paper reports that inserting a gadolinium interlayer between titanium and cobalt boosts spin-orbit torque efficiency by a factor of five, with the trilayer structure achieving torque efficiency above 1. This interface engineering approach may help develop lower-power magnetic devices for data storage and logic.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.3","headline":"ST-FMR and pumping data may not isolate orbital-to-spin conversion at Gd from SHE in Gd or interface scattering, undermining the >1 efficiency claim","rationale":"The reader's weakest assumption directly matches the load-bearing measurement-isolation issue required for the >1 efficiency claim. No stronger internal inconsistency appears from the abstract; the concern is therefore correctly placed and the UNVERDICTED verdict is appropriate pending full-text controls.","tokens_in":1704,"tokens_out":346,"duration_ms":12697,"concrete_test":"Re-analyze the ST-FMR data after subtracting the torque efficiency measured on a Gd(4 nm)/Co reference stack (or a Ti/Gd/Co stack with Ti replaced by a non-orbital-Hall metal); if the residual efficiency in the trilayer drops below 1 or loses its Ti-thickness independence, the orbital-conversion claim is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline result (trilayer torque efficiency >1, independent of Ti thickness, and fivefold larger than Gd/Co) requires that the measured damping-like torque in Ti/Gd/Co is attributable solely to orbital Hall current from Ti converted at the Gd interface. Gd possesses a non-zero spin Hall angle; any un-subtracted SHE contribution from the 4 nm Gd layer, or additional interface spin scattering not captured by the bilayer references, would inflate the apparent orbital efficiency. The abstract provides no quantitative controls (e.g., Gd-thickness series in the absence of Ti, or separate extraction of Gd spin Hall conductivity) that would rule this out. If the full text contains such controls they must be examined; absent them the attribution remains the weakest link.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript reports experimental results on Ti/Gd/Co trilayer structures for enhancing orbital Hall effect-driven spin-orbit torques. Using ferromagnetic resonance-based spin (orbital) pumping and spin-torque FMR measurements, it identifies an optimal Gd thickness of ~4 nm for maximum orbital-to-spin conversion, confirms bulk orbital Hall origin in Ti via thickness dependence with diffusion length >20 nm, and claims a fivefold SOT efficiency enhancement in Ti(20 nm)/Gd(4 nm)/Co relative to Gd/Co bilayers, with the trilayer torque efficiency exceeding 1 independent of Ti thickness.","tokens_in":1881,"tokens_out":447,"duration_ms":16077,"significance":"If the central attribution holds, the work provides a concrete experimental demonstration that rare-earth interlayers can boost orbital torque efficiencies above unity in light-metal systems, offering a materials-engineering route for spin-orbitronic devices. The use of thickness-dependent measurements to support bulk orbital Hall origin in Ti is a positive methodological feature.","major_comments":[{"comment":"Abstract and Results: The headline claim that the Ti/Gd/Co trilayer exhibits spin (orbital) torque efficiency >1 (fivefold larger than Gd/Co) rests on the unshown separation of orbital-to-spin conversion at the Gd interface from possible spin Hall effect contributions within the 4 nm Gd layer itself. No Gd-thickness series in the absence of Ti, nor separate quantification of Gd spin Hall conductivity, is described to rule out confounding; the bilayer references alone do not address this.","section":"Abstract"},{"comment":"Abstract and Methods: The reported efficiency values lack accompanying error bars, raw ST-FMR spectra, fitting procedures, or explicit controls for interface scattering contributions, making it impossible to assess the statistical robustness of the >1 efficiency and its Ti-thickness independence.","section":"Abstract"}],"minor_comments":[{"comment":"Notation for 'spin (orbital) torque efficiency' should be clarified to distinguish the extracted damping-like torque from the underlying orbital versus spin conversion mechanisms.","section":null}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the careful review and constructive comments on our manuscript. We address the major comments point by point below, indicating where revisions will strengthen the presentation.","responses":[{"response":"The Gd/Co bilayer reference already incorporates any spin Hall contribution from the fixed 4 nm Gd layer. The observed fivefold enhancement and efficiency exceeding 1 in the trilayer therefore arise from the additional orbital current generated in Ti and converted at the Ti/Gd interface. The reported independence of the trilayer efficiency on Ti thickness (while remaining >1) further indicates that the excess torque scales with the Ti orbital source rather than with Gd. We will revise the text to explicitly articulate this subtraction logic and to note that a dedicated Gd-thickness series without Ti, while desirable, is not required to interpret the differential enhancement shown by the existing controls.","revision_made":"partial","referee_comment":"[Abstract] Abstract and Results: The headline claim that the Ti/Gd/Co trilayer exhibits spin (orbital) torque efficiency >1 (fivefold larger than Gd/Co) rests on the unshown separation of orbital-to-spin conversion at the Gd interface from possible spin Hall effect contributions within the 4 nm Gd layer itself. No Gd-thickness series in the absence of Ti, nor separate quantification of Gd spin Hall conductivity, is described to rule out confounding; the bilayer references alone do not address this."},{"response":"We agree that the statistical presentation can be strengthened. In the revised manuscript we will add error bars to all efficiency values, include representative raw ST-FMR spectra with fits, provide a detailed description of the fitting procedure, and discuss controls for interface scattering to allow readers to evaluate the robustness of the >1 efficiency and its Ti-thickness independence.","revision_made":"yes","referee_comment":"[Abstract] Abstract and Methods: The reported efficiency values lack accompanying error bars, raw ST-FMR spectra, fitting procedures, or explicit controls for interface scattering contributions, making it impossible to assess the statistical robustness of the >1 efficiency and its Ti-thickness independence."}],"tokens_in":1384,"tokens_out":448,"duration_ms":22775,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The central result is that a 4 nm Gd interlayer in Ti/Gd/Co trilayers produces a fivefold rise in measured damping-like torque efficiency compared to Gd/Co or Ti/Co bilayers, with the value exceeding 1 and staying independent of Ti thickness. They link this to orbital Hall current generated in Ti and converted at the Gd interface, backed by FMR pumping data that peaks at that Gd thickness and a Ti-thickness series suggesting bulk origin plus diffusion length over 20 nm.\n\nThe experimental mapping of thicknesses and the use of two complementary techniques (pumping and ST-FMR) is straightforward and gives concrete numbers that device groups could try to reproduce. The optimal Gd thickness and the reported independence from Ti thickness are the parts that stand out as usable.\n\nThe soft spot is the lack of shown controls for separating orbital conversion from other contributions. Gd has a non-zero spin Hall angle, and a 4 nm layer could add its own torque or extra scattering at the interfaces. The abstract does not describe a Gd-thickness series without Ti, separate extraction of Gd spin Hall conductivity, or error bars on the efficiency values. If those are missing from the full text, the >1 claim rests on an untested assumption that the extra signal is purely orbital-to-spin at Gd.\n\nThis is for experimental spintronics people who work on orbital torques and interface stacks. A reader chasing higher-efficiency materials might extract the thickness trends even while treating the exact efficiency number as provisional.\n\nSend it for peer review so the full dataset and any hidden controls can be checked directly.","headline":"The paper claims a fivefold SOT efficiency boost above 1 in Ti/Gd/Co via Gd interlayer for orbital conversion, but the abstract leaves open whether Gd's own SHE or interface effects are subtracted.","tokens_in":2370,"tokens_out":406,"would_cite":false,"duration_ms":24260,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"A gadolinium interlayer between titanium and cobalt produces orbital torque efficiency above 1, exceeding both Ti/Co and Gd/Co bilayers.","keywords":["orbital Hall effect","spin-orbit torque","rare-earth interlayer","titanium","gadolinium","cobalt","trilayer","orbital torque efficiency"],"falsifier":"A repeat of the spin-torque ferromagnetic resonance experiment on the same Ti/Gd/Co stacks that finds torque efficiency no higher than the Ti/Co or Gd/Co bilayers or below 1 would falsify the reported enhancement.","tokens_in":2624,"feed_emoji":"","tokens_out":787,"duration_ms":15548,"temperature":0.7,"pith_summary":"The paper examines limits on turning orbital currents in light metals such as titanium into usable spin-orbit torques. It tests whether a thin gadolinium layer inserted between titanium and cobalt can raise the orbital-to-spin conversion at the interface. Ferromagnetic resonance and spin-torque measurements locate an optimum gadolinium thickness near 4 nm and confirm a bulk orbital Hall origin in titanium with a diffusion length longer than 20 nm. The resulting Ti/Gd/Co trilayer reaches a torque efficiency greater than 1 and outperforms the two corresponding bilayers for any titanium thickness tested. The work therefore presents rare-earth interlayer engineering as one concrete way to increase orbital torque efficiency.","feed_headline":"Gd interlayer raises Ti/Co orbital torque efficiency above 1","feed_subtitle":"Trilayer Ti/Gd/Co outperforms both Ti/Co and Gd/Co bilayers at any Ti thickness, with peak conversion at 4 nm Gd.","key_machinery":"The gadolinium interlayer that converts orbital current from the titanium layer into spin current acting on the cobalt ferromagnet.","core_discovery":"The Ti/Gd/Co trilayer architecture exhibits a spin (orbital) torque efficiency greater than 1, which is higher than that of the bilayer Ti/Co and Gd/Co structures, irrespective of Ti thickness. Ferromagnetic resonance-based spin (orbital) pumping measurements identify an optimal Gd thickness of around 4 nm, where the orbital-to-spin conversion efficiency reaches its maximum. The Ti-thickness dependence of the inverse orbital Hall effect signal confirms a bulk orbital Hall origin in Ti and yields a qualitative orbital diffusion length exceeding 20 nm. Spin-torque ferromagnetic resonance measurements demonstrate a fivefold enhancement of the SOT efficiency in Ti(20 nm)/Co compared to a Gd(4 nm","pith_inferences":["The same interlayer strategy could be tested with other light metals that possess large orbital Hall conductivities.","Efficiency values above 1 may indicate either additional torque channels or a need to refine the normalization used in the measurements.","Interface scattering or spin memory loss at the gadolinium-cobalt boundary could be quantified separately to test whether they limit further gains.","The approach might be combined with other orbital sources or different rare-earth choices to map the parameter space of conversion efficiency."],"forward_implications":["Orbital torque efficiency can exceed unity when a rare-earth interlayer is added to a light-metal orbital source.","The torque enhancement remains independent of titanium thickness once the gadolinium layer is present.","An optimal gadolinium thickness of approximately 4 nm maximizes the orbital-to-spin conversion.","The orbital Hall effect in titanium is shown to be a bulk phenomenon with diffusion length longer than 20 nm.","Rare-earth interlayers offer a practical route to higher-efficiency orbital-torque devices."],"fun_headline_variants":["Ti/Gd/Co orbital torque efficiency exceeds 1","Torque efficiency higher in trilayer than in bilayers","Optimal Gd thickness of 4 nm for conversion efficiency","Bulk Ti orbital Hall origin with diffusion length exceeding 20 nm"],"cache_read_input_tokens":64,"weakest_assumption_plain":"The ferromagnetic resonance and spin-torque ferromagnetic resonance measurements isolate the orbital-to-spin conversion efficiency at the gadolinium interface without contributions from the spin Hall effect inside gadolinium or from unaccounted interface scattering.","fun_headline_variants_meta":{"raw":{"variants":["Ti/Gd/Co orbital torque efficiency exceeds 1","Torque efficiency higher in trilayer than in bilayers","Optimal Gd thickness of 4 nm for conversion efficiency","Bulk Ti orbital Hall origin with diffusion length exceeding 20 nm"]},"model":"grok-4.3","cost_usd":0.00835,"raw_usage":{"total_tokens":3804,"prompt_tokens":713,"num_sources_used":0,"completion_tokens":63,"cost_in_usd_ticks":83499500,"prompt_tokens_details":{"text_tokens":713,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":3028,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":713,"tokens_out":63,"duration_ms":22316,"temperature":1.0,"reasoning_tokens":3028,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-02T22:39:02.215375+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A repeat of the spin-torque ferromagnetic resonance experiment on the same Ti/Gd/Co stacks that finds torque efficiency no higher than the Ti/Co or Gd/Co bilayers or below 1 would falsify the reported enhancement.","supporting_citations":[],"review_version":2}