{"id":"5aa6004c-b46b-4112-98e0-d3ee5c52e10c","arxiv_id":"2501.15762","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"SnTe/NiO/CoFeB devices demonstrate room-temperature perpendicular magnetization switching via magnon torques with 22-fold lower power consumption than Bi2Te3-based controls.","lead":"This paper shows that a topological crystalline insulator, SnTe, can switch the magnetization of a thin magnetic layer at room temperature using magnon currents through an insulating nickel oxide layer. The device uses about 22 times less power than a similar bismuth telluride device, pointing toward low-power magnetic memory and logic.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The magnon-mediated interpretation rests on one non-monotonic NiO-thickness series; without a matched-thickness nonmagnetic spacer or leakage measurement, the reappearing switching at t>10 nm could be electron-mediated.","rationale":"The most load-bearing element is the magnon mechanism, not the SHC value or the power ratio. The SHC could be off by a factor of two due to the 1 nm Bi2Te3 buffer or extrinsic contributions, and the power comparison could shift if the Bi2Te3 baseline were remeasured at the same NiO thickness, but the paper would still report room-temperature switching in SnTe/NiO/CoFeB. The magnon claim, however, is the paper's conceptual advance. It rests on the reappearance of switching for NiO t>10 nm in Fig. 4 and the matching ST-FMR theta_y peak in Fig. 2b. The internal evidence is good: t=0 and t=3-10 are consistent with electron and blocked-spin regimes, and the Pt and Bi0.9Sb0.1 negative controls in Section S11 are supportive. What is missing is a direct electrical or leakage check of the 20 nm polycrystalline NiO barrier and a matched-thickness nonmagnetic spacer. Since a small pinhole population would produce exactly the same Hall-bar switching (SnTe and CoFeB are both conductive and contacted at the same electrodes), this is a genuine alternative, not a stylistic objection. The reader's conditional verdict is appropriate; I do not see grounds to reject the paper or to accept without the requested control.","tokens_in":16446,"tokens_out":15835,"duration_ms":156060,"concrete_test":"Pattern a cross-point vertical transport test on a sister film of SnTe(8)/NiO(20)/CoFeB using the identical fabrication process: define a bottom contact through the SnTe layer and a top contact on the CoFeB layer with a junction area comparable to the Hall bar (20x40 um), then measure the room-temperature two-terminal I-V. If the vertical resistance is below roughly 10 MOhm or shows ohmic conduction, pinholes or leakage through the 20 nm NiO are present, and the re-emergent Hall-bar switching at t>10 nm may be electron-mediated rather than magnon-mediated. If the junction is insulating (above 10 MOhm) while the switching is still observed, the magnon interpretation is strongly supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that SnTe/NiO/CoFeB switches via magnon torques rests on the reappearance of switching for NiO thickness t > 10 nm (Fig. 4) and the theta_y peak at t = 20 nm (Fig. 2b). The paper interprets this as antiferromagnetic magnon transport through NiO. What would have to be true is that the 20 nm polycrystalline NiO layer is electrically insulating and free of pinholes over the device area. No vertical leakage measurement between SnTe and CoFeB through the NiO is reported, and the provided nonmagnetic controls are SnTe/MgO(3)/CoFeB and SnTe/Cu(3)/CoFeB (Section S9), neither matched to the 20 nm NiO thickness at which the effect is claimed. If even a small pinhole or leakage current exists in the 20 nm NiO, the re-emergent switching at t>10 nm could be conventional electron-mediated SOT from SnTe, not magnon torque. The Pt(8)/NiO(20)/CoFeB and Bi0.9Sb0.1 controls in Section S11 are useful but sample-specific; they do not establish the integrity of the specific SnTe/NiO(20) devices. This is load-bearing because the entire 'magnon-mediated switching' conclusion, and thus the central claim, depends on it.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports room-temperature perpendicular magnetization switching in SnTe/NiO/CoFeB heterostructures, attributed to magnon torques generated in NiO by spin currents from the topological crystalline insulator SnTe. The authors measure a spin Hall angle of 0.165 in SnTe/Py by ST-FMR, derive a spin Hall conductivity of about 6.1×10^4 (ħ/2e) (Ω m)^−1, support it with Wannier-interpolated ab initio calculations, observe a NiO-thickness-dependent recovery of spin torque efficiency peaking at 20 nm, demonstrate current-induced switching of CoFeB with an in-plane field, and report a 22-fold lower power consumption than a Bi2Te3/NiO/CoFeB control.","tokens_in":16722,"tokens_out":5378,"duration_ms":47792,"significance":"If the magnon-torque interpretation holds, the work extends room-temperature magnon-mediated switching to a topological crystalline insulator and provides a quantitative power comparison, with potential impact on magnonic devices. Strengths include the independent DFT/MLWF calculation of the intrinsic SHC, systematic ST-FMR and switching data over NiO thickness, several control samples (Cu, MgO, Pt, Bi0.9Sb0.1), a measured temperature rise of 11.3 K, and an explicit two-current power model. The main risks are experimental attribution of the switching to magnons rather than leakage/pinhole electron transport, and the unmatched Bi2Te3 control thickness in the headline power claim.","major_comments":[{"comment":"The reappearance of switching for NiO thicknesses above 10 nm is the central evidence for magnon-mediated switching. The provided nonmagnetic spacer controls (SnTe/Cu(3)/CoFeB and SnTe/MgO(3)/CoFeB) use only 3-nm-thick spacers; they do not match the 20-nm NiO thickness at which the magnon effect is claimed, and they do not establish that the polycrystalline 20-nm NiO layer in the actual devices is pinhole-free and electrically insulating over the device area. Without a vertical leakage measurement between SnTe and CoFeB across the 20-nm NiO, or a matched-thickness nonmagnetic insulator control (e.g., SnTe/MgO(20)/CoFeB), electron-mediated spin-orbit torque from SnTe through pinholes or leakage remains a plausible alternative explanation for the re-emergent switching. Please report leakage current density versus voltage for the SnTe/NiO(20)/CoFeB stack and, if possible, a 20-nm nonmagnetic spacer control.","section":"Section III (Fig. 4) and Section S9"},{"comment":"The '22-fold lower power' headline result compares SnTe(8)/NiO(20)/CoFeB with Bi2Te3(8)/NiO(25)/CoFeB, i.e., with different NiO thicknesses (20 vs 25 nm). Because the magnon torque efficiency is strongly thickness-dependent (Fig. 2b and refs. 11, 17), the power comparison is not matched. To support the quantitative claim, the Bi2Te3 control should be measured at the same NiO thickness (20 nm) in the same fabrication batch and device geometry, or the authors should provide data demonstrating that the power consumption is nearly independent of NiO thickness in the 20–25 nm range.","section":"Section S5 and S6"},{"comment":"The experimental spin Hall conductivity of SnTe is extracted from ST-FMR on SnTe(8)/Py films grown on a 1-nm Bi2Te3 buffer. The manuscript assumes this buffer is 'insulating' and does not contribute to the spin Hall effect or to current shunting, but no direct resistivity or spin-torque measurement of the buffer alone, or of a buffer-free control, is reported. Since Bi2Te3 is a known spin-Hall material, even a thin, partially conducting buffer could contribute to the measured θy and σs. Please provide a control (e.g., SnTe grown without buffer or with a confirmed-insulating buffer) or a quantitative estimate of the buffer's resistivity and its effect on the ST-FMR analysis.","section":"Section II.A"},{"comment":"The interpretation that NiO becomes antiferromagnetic at room temperature for t > 10 nm would be strengthened by reporting the blocking temperature Tb for the 20-nm NiO samples together with the room-temperature exchange bias (or its absence). The current text reports Tb for 3 nm (~70 K) and states that Tb increases with t, but does not give values for the thicknesses used in the switching devices (20 nm). If Tb for 20 nm is below 300 K, the assignment of the room-temperature switching to antiferromagnetic magnon transport would need to be revisited.","section":"Section S4 and Fig. S12"}],"minor_comments":[{"comment":"The first sentence of the abstract has a spacing error: 'Magnons possessthe ability' should be 'Magnons possess the ability'; also 'practical application' appears singular where 'applications' is intended.","section":"Abstract and Introduction"},{"comment":"In the sentence 'Moreover, the ?RF induces the Oersted field torque', the placeholder '?' should be 'I_RF'; the same issue appears for 'the ?RF' earlier in the section.","section":"Section II.B"},{"comment":"The phrase 'Figure 1c shows the calculated electronic structure of prototypical SnTe' contains the typo 'prototypical'; it should read 'prototypical' or, better, 'a prototypical SnTe crystal'.","section":"Section III"},{"comment":"Equation (1) in the Supplemental Material appears malformed: the displayed formula has garbled symbols such as '??spin?' and '?2?'; please correct the typesetting so that the Kubo-Greenwood formula is clearly readable.","section":"Section S3"},{"comment":"The sentence 'a few nanometers of NiO are similar to the nonmagnetic insulator MgO, which effectively suppress magnon transports' is grammatically incorrect and conceptually ambiguous; it should say that thin NiO, like MgO, suppresses spin transport, not magnon transport, since MgO is a nonmagnetic tunnel barrier.","section":"Section S9"},{"comment":"Reference [14] contains an extra comma: 'and L. Liu, , Mutual control of coherent spin waves and magnetic domain walls in a magnonic device' should be 'and L. Liu, Mutual control of coherent spin waves and magnetic domain walls in a magnonic device'.","section":"References"},{"comment":"The phrase 'ExtendedData Figure 8 of Reference 19' refers to a figure outside the cited work's scope; please provide the actual figure or a complete citation so the reader can access the supporting data.","section":"Section S10"}],"recommendation":"major_revision","confidential_remarks":"The manuscript's central claims are interesting but rest on several experimental attributions that need tighter controls: the integrity of the 20-nm NiO spacer, the matched-thickness Bi2Te3 comparison, and the role of the 1-nm Bi2Te3 buffer. These are addressable with additional measurements and would substantially strengthen the paper. I would not recommend rejection, as the core phenomenon (thickness-dependent switching reappearance) is plausible, but the current evidence does not fully exclude electron-mediated leakage paths."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the one-sentence version: this is a serious, well-executed device demonstration of magnon-torque switching in SnTe/NiO/CoFeB, but the 22-fold power claim and the SHC attribution both rest on comparisons that need tightening before I'd trust the numbers.\n\nThe genuinely new content is SnTe as the spin-source layer for magnon-mediated perpendicular switching at room temperature. They show a clean non-monotonic NiO-thickness response in both ST-FMR torque efficiency and switching ratio, with a peak at 20 nm, and they tie that to the onset of antiferromagnetic order in the NiO using exchange-bias and coercivity data. The ST-FMR gives a spin Hall angle of about 0.165 for SnTe/Py, and the DFT calculation of the intrinsic spin Hall conductivity lands within a factor of two of the measured value; that's a credible independent check. The controls—Cu, MgO, Pt, and Bi0.9Sb0.1—are the right ones, and the use of a lower-resistivity topological material addresses a real limitation of the earlier Bi2Se3/Bi2Te3 work.\n\nThe soft spots are real but not fatal. The 22-fold power reduction is computed against a Bi2Te3/NiO(25 nm) sample from the authors' previous publication, not a same-batch control, and the NiO thicknesses differ. There are no error bars on the switching statistics, so the ratio could move by a factor of two or more. More importantly, the 1 nm Bi2Te3 buffer under the SnTe is present in every measured device, but its contribution to the spin-orbit torque and SHC is never characterized. Since the ST-FMR and switching data come from stacks that include that buffer, attributing sigma_s entirely to SnTe is not airtight. The stress-test note about possible pinhole leakage through the 20 nm NiO is worth mentioning, but I think it's partially mitigated: the Pt/NiO/CoFeB control shows no switching, and the non-monotonic ST-FMR thickness series would be difficult to explain by simple leakage.\n\nIf I were refereeing, I'd require three things: a leakage test or a matched-thickness nonmagnetic spacer in the same batch, a same-batch Bi2Te3 baseline, and a direct measurement of the buffer-only contribution (e.g., an SnTe-free control). None of these are dealbreakers, but they'd firm up the two headline claims.\n\nThis is a solid materials-and-devices paper for the spintronics community. It deserves a serious referee, and I'd expect a conditional acceptance. The mechanism extends existing work, so the significance is moderate, but the lower-power angle is genuinely useful.","headline":"A useful new spin-source material for room-temperature magnon-torque switching, but the 22-fold power claim and the uncharacterized buffer need hardening before I'd trust the headline numbers.","tokens_in":17310,"tokens_out":3883,"would_cite":true,"duration_ms":35718,"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":"SnTe, a topological crystalline insulator, switches perpendicular magnets at room temperature via magnon torques, with a 22-fold lower power consumption than a Bi2Te3 control.","keywords":["magnon torque","perpendicular magnetization switching","topological crystalline insulator","SnTe","spin Hall conductivity","antiferromagnetic NiO","room temperature","spintronics"],"falsifier":"A decisive test is to measure the leakage current through the 20 nm NiO spacer under the same bias conditions used for switching, or to substitute a 20 nm nonmagnetic insulator such as MgO in the same stack and check whether switching persists; if switching survives a nonmagnetic spacer, or if a substantial fraction of the applied current flows through the NiO, the magnon interpretation is not required.","tokens_in":16222,"feed_emoji":"🧲","tokens_out":7451,"duration_ms":58796,"temperature":0.7,"pith_summary":"Magnons, the quanta of spin waves, can carry spin angular momentum through insulating magnets without moving electrons, which in principle allows low-power magnetic switching. The paper demonstrates that a topological crystalline insulator, SnTe, can act as the spin source for such magnon-mediated switching of perpendicular magnetization at room temperature. In SnTe/NiO/CoFeB stacks, a charge current in SnTe generates spin accumulation that excites magnon currents in the antiferromagnetic insulator NiO, and these magnon currents traverse a 20 nm NiO layer to exert torques on CoFeB and reverse its magnetization. The authors report a spin Hall conductivity of about $6.1 \\times 10^4\\,(\\hbar/2e)\\,(\\Omega\\,{\\rm m})^{-1}$ for SnTe and a 22-fold reduction in switching power relative to a Bi2Te3/NiO/CoFeB control. If correct, the result adds a new material class to the small set of systems that can switch magnets via magnon torques at practical temperatures.","feed_headline":"SnTe switches perpendicular magnets via magnons at 22x lower power","feed_subtitle":"A new spin-source material makes room-temperature, low-power magnetic switching practical for magnon devices.","key_machinery":"The mechanism is the conversion of a charge current in SnTe into a magnon current in the antiferromagnetic insulator NiO, which then exerts a magnon torque on the adjacent ferromagnet. Spin accumulation at the SnTe/NiO interface excites antiferromagnetic magnons, and because the NiO is polycrystalline its Néel vector is roughly uniformly distributed, so magnon currents of any polarization can propagate through the layer. The magnon torque efficiency is quantified by spin-torque ferromagnetic resonance (ST-FMR) in SnTe/NiO/Py devices and tracked as a function of NiO thickness, showing a peak at 20 nm. The theoretical part of the machinery is the calculation of the intrinsic spin Hall conductivity by maximally localized Wannier interpolation, which reproduces the band structure and yields a Fermi-level conductivity consistent with experiment once extrinsic side-jump and skew-scattering contributions are allowed for.","core_discovery":"The central claim is that magnon torques generated in SnTe can deterministically switch the perpendicular magnetization of CoFeB at room temperature. The key evidence is the NiO-thickness dependence of the switching: it is suppressed for NiO thicknesses between 3 and 10 nm, then re-emerges above 10 nm and peaks at 20 nm, which the authors link to the recovery of antiferromagnetic order in NiO. They interpret the re-emerging switching as magnon-mediated because a thin nonmagnetic MgO spacer blocks switching, the switching requires an in-plane assist field, and the switching polarity reverses when the field changes sign, as expected for a damping-like torque. The paper further finds that SnTe has a spin Hall conductivity of roughly $6.1\\times10^4\\,(\\hbar/2e)\\,(\\Omega\\,{\\rm m})^{-1}$, about three times that of Bi2Te3, as determined by spin-torque ferromagnetic resonance and supported by ab initio Wannier-interpolation calculations. The high conductivity is presented as the reason the SnTe device consumes about 22 times less power than the Bi2Te3 control.","pith_inferences":["The paper's interpretation predicts that electron-doped SnTe, whose calculated spin Hall conductivity rises to about $1.4\\times10^5\\,(\\hbar/2e)\\,(\\Omega\\,{\\rm m})^{-1}$ at roughly 0.7 eV above the Fermi level, should yield even lower switching current densities than the heavily hole-doped films used here.","If the magnon mechanism holds, the write current no longer flows through the ferromagnet, which would allow the magnetic layer to be electrically isolated and could open a path toward all-magnonic information processing where data are carried by spin waves rather than charge.","The supplementary observation that Pt(8 nm)/NiO(20 nm)/CoFeB does not switch implies that only spin sources with sufficiently strong interfacial spin accumulation can launch magnons through thick NiO; testing other high-spin-Hall-conductivity materials, such as WTe2, in the same geometry would sharpen this boundary."],"forward_implications":["The critical switching current density in the SnTe/NiO/CoFeB devices, $J_C \\approx 5.5\\times10^6$ A/cm², is below the values typical of electron-mediated spin-orbit torque switching, supporting low-current magnon-assisted writing.","The calculated power consumption of the SnTe device (about 74.7 mW under the reported conditions) is 22 times smaller than the Bi2Te3 control, evidence that high spin Hall conductivity translates into practical power savings.","The optimal NiO thickness for topological-material magnon sources is 20–25 nm, in contrast to 0.5–2 nm for Pt-based sources, which constrains device design for each spin-source material.","The combination of room-temperature operation and electrical isolation of the write path suggests that magnon-mediated switching could be integrated into device architectures that separate charge and spin information flow."],"supporting_citations":[{"why":"Demonstrated room-temperature magnetization switching by magnon-mediated spin torque through NiO, providing the mechanism the SnTe system extends.","marker":"[17]"},{"why":"Reported room-temperature switching by magnon torques in Bi2Te3/NiO/CoFeB, the control sample for the 22-fold power comparison.","marker":"[21]"},{"why":"Supplies the ST-FMR analysis procedure and the high-spin-Hall-conductivity WTe2/PtTe2 benchmark used to characterize SnTe.","marker":"[28]"},{"why":"Provides the Wannier-interpolation method used to compute the intrinsic spin Hall conductivity of SnTe.","marker":"[31]"},{"why":"Characterizes magnon propagation in NiO at nanoscale distances and is used to interpret the optimal NiO thickness of 20–25 nm.","marker":"[11]"},{"why":"Supplies the two-current model used to calculate and compare power consumption of the magnon devices.","marker":"[36]"}],"fun_headline_variants":["SnTe magnon torque flips perpendicular magnet at room temp","22x lower power: SnTe switches magnets via magnon currents","Topological insulator SnTe enables low-power magnon switching","Magnon-mediated switching: SnTe beats Bi2Te3 by 22x","Room-temp perpendicular switching with SnTe magnon torque"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the reappearance of switching for NiO thicker than 10 nm is caused by magnon transport through the antiferromagnet, rather than by leakage current, pinholes, or thickness-dependent Joule heating in the polycrystalline NiO layer; the paper provides no direct magnon-current measurement or leakage test.","fun_headline_variants_meta":{"raw":{"variants":["SnTe magnon torque flips perpendicular magnet at room temp","22x lower power: SnTe switches magnets via magnon currents","Topological insulator SnTe enables low-power magnon switching","Magnon-mediated switching: SnTe beats Bi2Te3 by 22x","Room-temp perpendicular switching with SnTe magnon torque"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000671,"raw_usage":{"total_tokens":3082,"prompt_tokens":997,"completion_tokens":2085,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":613,"completion_tokens_details":{"reasoning_tokens":1996}},"tokens_in":613,"tokens_out":2085,"duration_ms":13661,"temperature":1.0,"reasoning_tokens":1996,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T13:57:43.430480+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test is to measure the leakage current through the 20 nm NiO spacer under the same bias conditions used for switching, or to substitute a 20 nm nonmagnetic insulator such as MgO in the same stack and check whether switching persists; if switching survives a nonmagnetic spacer, or if a substantial fraction of the applied current flows through the NiO, the magnon interpretation is not required.","supporting_citations":[],"review_version":1}