{"id":"9b19d058-70e2-4976-90b5-24cddc92ddac","arxiv_id":"2507.17372","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Adding a NiW seed layer to a Ru/Pt/Co stack improves Ru hcp(002) texture and raises the measured damping-like torque efficiency, an effect the authors attribute to enhanced orbital Hall torque.","lead":"This paper reports that growing the orbital-Hall layer Ru on a NiW seed layer improves its crystal texture and boosts the measured spin-orbit torque efficiency by about 1.4 times compared with a Pt-only reference. The recipe is a candidate for more energy-efficient magnetic switching in spintronic devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The NiW seedlayer is not a texture-only control: adding NiW changes the bottom electrode material, the Ta/Ru interface, and the Pt/Co crystallinity, so the claimed 2-fold OHE enhancement is not uniquely attributable to improved Ru hcp(002) texture.","rationale":"The reader's weakest assumption correctly identifies the M4/M5 confound, and my stress-test sharpens it: NiW is not an inert texture promoter but a W-based heavy-metal layer that can itself contribute spin/orbital torque, and the XRD results show that Pt and Co textures also improve, so the enhancement cannot be uniquely assigned to the Ru OHE. This is the most load-bearing issue because the central claim is specifically about the role of crystallographic texture on the OHE; if the control test reveals a direct NiW torque or a Pt-texture effect, the core conclusion fails. The annealing contradiction is real but secondary: it affects the paper's internal consistency and the claim of thermal stability, not the central texture attribution. Since the requested control is absent from the manuscript, the conditional verdict remains appropriate; no change to the reader's judgment is needed.","tokens_in":11198,"tokens_out":4683,"duration_ms":56461,"concrete_test":"Fabricate a control Hall-bar stack Ta(1)/NiW(2)/Pt(1.5)/Co(1.1)/Ru(2), identical to M5 but omitting the bottom Ru(4) layer, and measure ξ_DL^E with the same loop-shift and second-harmonic protocols used for M4 and M5. If this control shows a rise comparable to the M4-to-M5 increase, then the NiW layer itself generates torque and the texture attribution fails. If it shows no rise, the remaining ambiguity concerns Pt-texture and interface effects; a further control replacing the 4 nm Ru underlayer with a non-orbital-Hall spacer (e.g., 4 nm Cu with the same NiW seed) would separate the Ru OHE contribution from NiW-induced changes. Report resistivity and XRD data for all controls to check current shunting and crystallographic changes.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on the M4-vs-M5 comparison, where the only intended difference is the 2 nm NiW seed that improves Ru hcp(002) texture, and the 2-fold increase in the inferred OHE contribution is assigned wholly to that texture. But M5 differs from M4 in at least three torque-relevant ways. First, NiW is a W-based alloy with potentially large spin Hall and orbital Hall conductivities; it sits directly in the current path and can itself produce a damping-like torque, so part or all of the increase could be a new torque source rather than an enhancement of the Ru OHE. Second, the XRD data in Fig. 1d show that NiW also increases the Pt fcc(111) and Co hcp(002) peak intensities; Pt's spin Hall conductivity is crystallinity- and texture-sensitive, so the Pt SHE contribution can change between M4 and M5 independently of the Ru OHE. Third, the Ta/NiW/Ru and NiW/Ru interfaces change spin/orbital transparency and current distribution. No control with NiW inserted without the Ru layer, or with Ru texture improved by an inert seed, is reported. Thus the phrase 'clearly indicates the role of crystallographic texture' overstates the evidence; the data are equally consistent with a direct NiW torque or a Pt-texture effect. A secondary inconsistency: the abstract and conclusion state there is no appreciable change after annealing, while Section 6 reports significant reductions for M2, M4, and M5.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports damping-like torque efficiency measurements in Ru/Pt/Co heterostructures, with a NiW seed layer used to improve the hcp(002) texture of Ru. Comparing stacks M4 (Ta/Ru/Pt/Co without NiW) and M5 (Ta/NiW/Ru/Pt/Co), the authors report a roughly twofold increase in the extracted OHE contribution to the effective damping-like torque efficiency and up to a 1.4-fold larger full-stack efficiency relative to a Pt(4 nm) reference. Series N optimizes Pt thickness and shows a 1.2-fold reduction in switching current density compared with the reference. The conclusions assign the M4-to-M5 improvement to enhanced orbital Hall torque from improved Ru texture, supplemented by synergistic SHE from Pt. The paper includes loop-shift and second-harmonic electrical measurements, XRD, VSM, and annealing studies.","tokens_in":11537,"tokens_out":4090,"duration_ms":45578,"significance":"If the texture-driven OHE enhancement is correct, the result is significant for orbitronics and SOT device engineering: it would show that crystallographic texture, not only material choice, controls orbital Hall torque. A genuine strength is that two independent electrical measurement techniques (loop-shift and harmonic) agree in the ranking of samples, which supports a real torque enhancement. The manuscript also reports a practical switching-current reduction and thermal-stability data. However, the central attribution to Ru texture is not established by the presented evidence, because the NiW seed introduces multiple simultaneous changes and no appropriate control is shown; the OHE/SHE decomposition on which the twofold claim rests is deferred to a supplementary file and reported without uncertainties. The paper should be revised to address these load-bearing issues before the claimed mechanism can be accepted.","major_comments":[{"comment":"The central claim that improved Ru hcp(002) texture doubles the OHE contribution rests entirely on the M4-versus-M5 comparison, but adding the 2 nm NiW seed changes the Ta/NiW/Ru interface, the current distribution through the stack, and the crystallinity of the Pt and Co layers simultaneously. In fact, Figure 1d shows that the Pt fcc(111) and Co hcp(002) peaks also intensify in M5, so a texture-dependent change in Pt SHE or a direct NiW spin/orbital torque is equally consistent with the data. Without a control sample containing NiW but no Ru, or a separate measurement in which Ru texture is improved by an inert seed, the phrase 'clearly indicates the role of crystallographic texture on the OHE enhancement' (Section 3) overstates what the experiment establishes.","section":"Section 3, Table 1, Figure 1d"},{"comment":"The quantitative basis of the twofold claim is the decomposition that yields OHE contributions of 0.39 x 10^5 and 0.81 x 10^5 ohm^-1 m^-1 for M4 and M5, but this decomposition is only described as being 'in supplementary,' and no uncertainties are reported anywhere for these values or for the fold-enhancement ratios. Since the paper's central message depends on this subtraction model, the main text must show the explicit decomposition equations, the literature values used, the sign conventions, and the error propagation. As written, the reader cannot verify whether the reported twofold OHE increase is robust or an artifact of the model.","section":"Section 3"},{"comment":"The abstract and the conclusions state that 'no appreciable change' in spin-orbit torque efficiency is found after annealing, but Section 6 reports 'a significant reduction' for M2, M4, and M5 and a slight reduction for M1 and M3. This is an internal inconsistency that must be reconciled. The statement about M5 being 'comparable' is also difficult to evaluate because Figure 6 does not include error bars; please state clearly which samples changed and by how much.","section":"Section 6 and Abstract/Conclusions"},{"comment":"Equation (1) uses t_FM and the effective field per applied electric field, without correcting for how the current is distributed among the Ta, NiW, Ru, and Pt layers. Since M5 adds a conductive NiW underlayer, changes in current shunting between M4 and M5 can alter the measured efficiency independently of any texture effect on the OHE. The authors should either include a shunting correction (e.g., resistivity-weighted layer currents) or explicitly justify why the unweighted E normalization is adequate for the M4-versus-M5 comparison.","section":"Equation (1)"}],"minor_comments":[{"comment":"The text refers to 'Figure 4d' when describing the SOT-field versus current plot; this appears to be a typo for Figure 2d.","section":"Section 3"},{"comment":"The text refers to 'Figure 3d' for the plot of efficiency versus Pt thickness, but the figure caption only defines panels (a) and (b); the intended reference is likely Figure 3b.","section":"Section 3"},{"comment":"The paper repeatedly cites 'supplementary' for M1/M2 loop-shift results, the OHE/SHE separation, annealed XRD, and series-N characterization, but the supplementary file is not included in the reviewed version. Please either include it with the revision or move the essential material into the main text.","section":"Throughout"},{"comment":"Figure 2f reports efficiency values without error bars, and the text gives fold-enhancement ratios without uncertainties; adding error bars or confidence intervals would make the claimed enhancements more convincing.","section":"Figure 2f"},{"comment":"The second-harmonic description says the first-harmonic fit uses Equation (2) and the second-harmonic fit uses an equation displayed inline; it would help to label that fitting equation as an equation number for reproducibility.","section":"Section 8"}],"recommendation":"major_revision","confidential_remarks":"The two electrical measurement techniques agreeing in ranking is a positive sign, but the missing NiW-only control is a genuine load-bearing gap: without it, the data cannot distinguish a texture-enhanced OHE from a direct NiW torque or a Pt-texture effect. I would ask the authors to add the control experiment or substantially soften the attribution claims, to move the decomposition into the main text with uncertainties, and to fix the annealing inconsistency. The manuscript is potentially publishable after such a revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague, here's the short version: this is a useful experimental paper with a promising idea, but the headline mechanism—that NiW enhances orbital Hall torque through Ru texture—is not established by the data as presented. The practical result (better torque efficiency, lower switching current) looks real; the attribution does not.\n\nWhat's actually new: applying the hard-disk NiW seedlayer trick to an orbital Hall stack, and the effect is reproducible across two measurement techniques (loop shift and harmonic) that agree in ranking. The device optimization is sensible; the Pt-thickness series gives a 1.4x improvement in torque efficiency over Pt(4) and a 1.2x lower switching current, with harmonic measurements consistent. As an engineering route to SOT-MRAM, the result is plausible and the paper is readable.\n\nNow the soft spots. The M4/M5 comparison lacks a control that isolates texture. NiW is a W-based alloy with potentially large spin Hall and orbital Hall conductivities; it sits in the current path, so some or all of the increased torque could be a direct NiW contribution, not enhanced Ru OHE. The XRD also shows NiW increases Pt(111) and Co(002) peaks, and Pt SHE is texture-sensitive, so the Pt contribution can change independently. The phrase 'clearly indicates' overstates this. Second, the OHE/SHE decomposition is deferred to supplementary, but the main text uses a 2-fold figure from that decomposition; ratios are quoted without uncertainties. Third, the annealing section contradicts the abstract: the abstract says no appreciable change, while Section 6 reports significant reductions for M2, M4, and M5. These are fixable issues, not a dead end.\n\nWho it's for: experimental spintronics/orbitronics groups working on SOT-MRAM. A referee should ask for the NiW-only control (or an inert seedlayer control), uncertainty propagation, and a consistent annealing narrative. The paper deserves serious refereeing because the data are real, the idea is fresh, and the practical metrics are useful even if the texture mechanism remains unproven.","headline":"Useful experiment, overclaimed mechanism: NiW seedlayer improves SOT efficiency, but the M4/M5 comparison doesn't isolate texture from other NiW effects.","tokens_in":12029,"tokens_out":2554,"would_cite":false,"duration_ms":28681,"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":"Adding a 2 nm NiW seedlayer doubles the orbital-Hall torque contribution from a Ru layer.","keywords":["orbital Hall effect","spin Hall effect","spin-orbit torque","Ru hcp(002) texture","NiW seedlayer","damping-like torque efficiency","perpendicular magnetic anisotropy","magnetization switching"],"falsifier":"A control stack with NiW present but no bottom Ru layer should show no comparable rise in damping-like torque efficiency; if it does, the boost comes from the seedlayer's other effects rather than from Ru texture.","tokens_in":11039,"feed_emoji":"🧲","tokens_out":9292,"duration_ms":80498,"temperature":0.7,"pith_summary":"This paper tries to establish that the crystallographic texture of an orbital-Hall layer is a controllable lever for spin-orbit torque. By inserting a 2 nm NiW seedlayer beneath a Ru layer, the hcp(002) texture improves, and the inferred orbital-Hall contribution to the effective damping-like torque efficiency $\\xi_{DL}^{E}$ roughly doubles. With Pt providing both spin-Hall current and orbital-to-spin conversion, the full stack reaches about 1.4 times the efficiency of a Pt(4 nm) reference and switches at a 1.2-fold lower current density. The claim matters because it points to texture engineering, not only material choice, as a way to make orbital Hall effects useful in spintronic devices.","feed_headline":"A 2 nm seedlayer doubles the orbital-Hall torque from Ru","feed_subtitle":"Textured Ru plus Pt lifts damping-like torque efficiency to 1.4x a pure Pt stack and cuts switching current by 1.2x","key_machinery":"The load-bearing object is the Ru hcp(002) crystallographic texture, promoted by a NiW seedlayer adopted from hard-disk media technology. The mechanism is the orbital Hall effect in Ru, which generates an orbital current that a Pt layer with strong spin-orbit coupling converts into spin current, while Pt also contributes its own spin Hall current. The measured quantity is the damping-like torque efficiency $\\xi_{DL}^{E}=(2e/\\hbar)M_{S}t_{FM}(\\mu_0 H_{SOT}/E)$, obtained from current-induced loop shifts and corroborated by second-harmonic Hall measurements.","core_discovery":"The central claim is that a 2 nm NiW underlayer improves the Ru hcp(002) texture and thereby doubles the orbital-Hall contribution to the effective damping-like torque efficiency: the disentangled OHE contribution rises from $0.39\\times10^{5}\\,\\Omega^{-1}\\text{m}^{-1}$ in the untextured stack to $0.81\\times10^{5}\\,\\Omega^{-1}\\text{m}^{-1}$ in the textured one. In the optimized Pt(2.5 nm) stack, the combined spin-Hall and orbital-Hall torque reaches about 1.4 times the $\\xi_{DL}^{E}$ of a Pt(4 nm) reference, and current-induced switching shows about a 1.2-fold reduction in switching current density. XRD supports the texture story: adding NiW strengthens and slightly shifts the Ru hcp(002) peak, and also sharpens the Pt(111) and Co(002) peaks.","pith_inferences":["If texture is the active ingredient, other hcp(002)-promoting seedlayers or epitaxial Ru should raise the orbital-Hall torque further; that is a directly testable prediction the paper does not make.","The 2-fold OHE enhancement is inferred by subtracting two full stacks, so a NiW-only control would separate texture from underlayer and interface effects.","The same Ru/Pt synergy could be tried with other orbital-Hall metals such as Nb, Zr, or Ti to see whether seedlayer texture engineering generalizes beyond Ru.","A NiW thickness series would show whether the torque gain saturates once Ru(002) texture is fully developed, giving a design rule for SOT stacks."],"forward_implications":["Texturing an orbital-Hall metal with a seedlayer is a practical route to raise $\\xi_{DL}^{E}$ beyond what spin Hall metals alone provide.","The Pt thickness has an optimum near 2.5 nm for orbital-to-spin conversion; thicker Pt diffuses the converted spin current and leaves only the spin Hall contribution.","The optimized textured stack cuts switching current density by about 1.2 times relative to a Pt-only reference at similar perpendicular anisotropy.","Annealing at 300 °C leaves the textured stack's torque efficiency essentially unchanged, which the authors read as CMOS-compatible thermal stability."],"supporting_citations":[{"why":"Demonstrates orbital-Hall torque from Ru, Nb, and Cr with Pt conversion in perpendicular stacks, providing the baseline this paper extends.","marker":"[24]"},{"why":"Shows NiW seedlayers improve Ru texture for perpendicular recording media, motivating the seedlayer choice.","marker":"[29]"},{"why":"Supports hcp-phase growth and texture development on NiW seedlayers.","marker":"[30]"},{"why":"Provides the theoretical orbital-texture origin of the intrinsic orbital Hall effect.","marker":"[14]"},{"why":"Establishes orbital-to-spin conversion of interfacial orbital currents, underpinning the Pt converter layer.","marker":"[21]"},{"why":"Reports long-range orbital-Hall torques and spin diffusion, used to explain Pt-thickness dependence.","marker":"[25]"},{"why":"Supplies the loop-shift method for extracting damping-like torque efficiency.","marker":"[33]"},{"why":"Provides the detailed measurement protocol the paper follows for loop-shift and switching experiments.","marker":"[34]"}],"fun_headline_variants":["NiW seedlayer doubles Ru orbital-Hall torque","Textured Ru lifts damping-like torque to 1.4x Pt stack","Seedlayer texture doubles orbital-Hall contribution to torque","Ru orbital Hall torque doubled by NiW texture control","Synergistic spin and orbital Hall effects raise torque efficiency by 44%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The M4-versus-M5 comparison assumes that adding 2 nm of NiW changes only the Ru texture; in reality it also changes the underlayer material, interface quality, Pt/Co growth, and current distribution, and no control sample with NiW but without Ru isolates those effects.","fun_headline_variants_meta":{"raw":{"variants":["NiW seedlayer doubles Ru orbital-Hall torque","Textured Ru lifts damping-like torque to 1.4x Pt stack","Seedlayer texture doubles orbital-Hall contribution to torque","Ru orbital Hall torque doubled by NiW texture control","Synergistic spin and orbital Hall effects raise torque efficiency by 44%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000643,"raw_usage":{"total_tokens":2957,"prompt_tokens":942,"completion_tokens":2015,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":558,"completion_tokens_details":{"reasoning_tokens":1931}},"tokens_in":558,"tokens_out":2015,"duration_ms":19662,"temperature":1.0,"reasoning_tokens":1931,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T14:48:59.534030+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A control stack with NiW present but no bottom Ru layer should show no comparable rise in damping-like torque efficiency; if it does, the boost comes from the seedlayer's other effects rather than from Ru texture.","supporting_citations":[{"cited_title":"& Oppeneer, P","cited_arxiv_id":null,"evidence_quote":"Demonstrates orbital-Hall torque from Ru, Nb, and Cr with Pt conversion in perpendicular stacks, providing the baseline this paper extends."},{"cited_title":"B., Aon, S","cited_arxiv_id":null,"evidence_quote":"Shows NiW seedlayers improve Ru texture for perpendicular recording media, motivating the seedlayer choice."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supports hcp-phase growth and texture development on NiW seedlayers."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the theoretical orbital-texture origin of the intrinsic orbital Hall effect."},{"cited_title":"& Zhang, X","cited_arxiv_id":null,"evidence_quote":"Establishes orbital-to-spin conversion of interfacial orbital currents, underpinning the Pt converter layer."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports long-range orbital-Hall torques and spin diffusion, used to explain Pt-thickness dependence."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the loop-shift method for extracting damping-like torque efficiency."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the detailed measurement protocol the paper follows for loop-shift and switching experiments."}],"review_version":1}