{"id":"b545f22b-5ef4-4f88-8a67-e1b8e33d2ac4","arxiv_id":"2502.04561","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Wide-field quantum imaging with boron-vacancy sensors in hBN reveals field-free deterministic SOT switching in WTe2/Fe3GaTe2/hBN, including domain nucleation and a heating-driven loss of determinism.","lead":"Using atomic-scale light-emitting sensors embedded in a hexagonal boron nitride layer, this paper directly watches an electric current flip the magnetic orientation of an ultrathin van der Waals magnet without any external magnetic field. The images show that switching starts at weak spots and grows, but that strong currents introduce heat-induced randomness, information that is invisible to standard electrical measurements and relevant to future low-power spintronic memory.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Bs maps are read as magnetization images without a control separating magnetic fields from strain/thermal/charge-induced ODMR shifts; the domain-nucleation claim depends on this unvalidated step.","rationale":"I agree with the reader that the weakest assumption is the conversion from ODMR contrast to local magnetic field. This is the load-bearing step because the paper's distinctive contribution—nanoscale imaging of SOT-driven domain nucleation and propagation—depends entirely on the Bs maps being quantitative magnetization images. I considered alternative concerns: the ~70% switched fraction in Fig. 3d relative to the apparently full Hall plateau, the possible use of a bias field for ODMR readout, and the inferred Joule-heating mechanism. These are either less central or less clearly resolvable from the text. The manuscript's own references to Supplementary Notes 5 and 6 for the Bs extraction, and the absence of deposited raw data, mean the calibration cannot currently be verified. The electrical switching phenomenon itself is not new (refs 8-11), so the uncalibrated imaging step is what carries the novelty. The proposed above-Tc control cleanly separates magnetic from non-magnetic contributions to the ODMR maps. Since the concern is serious but addressable with a control experiment, the existing CONDITIONAL verdict should stand.","tokens_in":9443,"tokens_out":11560,"duration_ms":136997,"concrete_test":"Heat the same WTe2/Fe3GaTe2/hBN stack to ~370 K (above the Fe3GaTe2 Curie temperature, reported as >350 K) and repeat the wide-field ODMR imaging while applying the identical current-pulse protocols used in Figs. 3 and 4. In the paramagnetic state, any current-induced spatial ODMR contrast or Bs-like patterns directly quantify non-magnetic artifacts; if none appear, the magnetic interpretation of the 260 K maps is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central new claim is that wide-field ODMR maps (Fig. 1g; Figs. 3a-3h) directly visualize Fe3GaTe2 magnetization domains during field-free SOT switching. This requires that the VB- ODMR splitting/contrast at each pixel be dominated by the local out-of-plane magnetic field of Fe3GaTe2, with negligible contributions from strain, temperature, charge, or Oersted effects. That condition is not established. VB- spin resonances in hBN are known to be sensitive to strain (via transverse zero-field splitting) and temperature (via the D parameter); the hBN sensor is in direct contact with a metallic heterostructure carrying 6.5-8.5 mA current pulses, so Joule heating and thermal-expansion mismatch can produce spatially varying, pulse-history-dependent ODMR shifts that can masquerade as magnetic domain contrast. The a-axis versus b-axis comparison is a useful symmetry control, but it does not exclude non-magnetic current-induced sensor artifacts because the heating and strain profiles differ between the two current directions. The manuscript repeatedly defers the Bs extraction details to Supplementary Notes 5 and 6 and presents no calibration against a known field and no control without magnetic order. Without such a control, the Bs maps are not yet demonstrated to be quantitative magnetization images.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports hBN-based wide-field quantum imaging of field-free deterministic magnetic switching in an all-van der Waals WTe2/Fe3GaTe2/hBN spin-orbit torque device. The authors combine anomalous Hall transport measurements with optically detected magnetic resonance (ODMR) imaging of negatively charged boron-vacancy (VB-) ensembles in the hBN encapsulation layer. They measure stray-field maps of the Fe3GaTe2 layer after a sequence of current pulses along either the low-symmetry a-axis or the high-symmetry b-axis of WTe2, and they interpret the maps as direct visualization of nanoscale magnetic domain nucleation and propagation. The central claims are: (i) field-free deterministic switching occurs for current pulses along the a-axis; (ii) imaging reveals that switching proceeds by local domain nucleation followed by domain growth, with roughly 70% of the flake reversed at the plateau; (iii) at larger currents the switching becomes partially indeterministic because Joule heating drives local regions above the Curie point; and (iv) current along the b-axis produces randomized multi-domain states with near-zero net Hall signal. The transport results are presented as corroborating the imaging data. The paper does not provide public data files, error bars on the stray-field maps, or a dedicated control separating magnetic-field-induced ODMR shifts from strain, temperature, and charge-induced shifts.","tokens_in":9612,"tokens_out":2879,"duration_ms":34216,"significance":"If the imaging interpretation is correct, this is a significant experimental advance: it would be one of the first direct microscopic views of field-free SOT switching in an all-vdW heterostructure, and it would demonstrate that hBN spin ensembles can serve as a quantitative magnetometry layer inside a working spintronic device. The a-axis versus b-axis comparison is a good symmetry control, and the transport data are internally consistent with the established out-of-plane spin polarization picture for WTe2. The domain-nucleation-and-propagation scenario, the degradation of determinism at high current, and the randomized remagnetization along the b-axis are falsifiable observations that go beyond conventional transport measurements. The paper does not, however, currently demonstrate that the measured ODMR contrast is dominated by the local magnetic stray field of Fe3GaTe2. Since the central novelty is the imaging claim, the missing calibration/control is load-bearing.","major_comments":[{"comment":"The Bs extraction details, the calibration of the field scale, the noise floor, and the image processing steps are deferred to Supplementary Notes 5 and 6, but the main text does not summarize the essential steps or show a representative raw ODMR spectrum with the fit used to convert splitting to field. Moreover, the Data Availability statement says data are available 'on reasonable request' without a repository. Given that the central claim rests on the quantitative interpretation of the Bs maps, I strongly recommend depositing raw and processed imaging data (or at least a representative dataset with fit residuals) in a public repository, and moving the key calibration details into the main text or a dedicated Methods section.","section":"Data availability and Supplementary Notes 5–6"}],"minor_comments":[{"comment":"The abstract and introduction describe Fe3GaTe2 as a room-temperature magnet, but the switching and imaging experiments are performed at 200 K (transport) and 260 K (imaging). Please clarify that the device operates below its Curie temperature and that room-temperature operation is a property of the material, not of the demonstrated device.","section":"Abstract and Figures 2–4"},{"comment":"The ODMR spectra in Figure 1f show peaks at different applied fields, but the field values are not labeled on the curves or in the legend. Adding the values used would let the reader verify the extracted gyromagnetic ratio and the splitting-to-field conversion.","section":"Figure 1f"},{"comment":"The anomalous Hall resistance is plotted in normalized units; the absolute resistance scale, the current pulse duration, and the waiting time between pulse and readout are not stated in the main text. These parameters are needed to judge whether the observed switching is quasi-static or influenced by pulse transients.","section":"Figure 2e–2h"},{"comment":"The manuscript would benefit from a more explicit comparison with prior transport-only demonstrations of field-free switching in WTe2/Fe3GaTe2 (e.g., refs. 10 and 11). Currently the novelty is framed as the imaging capability, which is fair, but the reader cannot tell which device parameters and switching ratios are new relative to Kajale et al. and Zhang et al.","section":"Introduction, references 8–13"},{"comment":"The pulse-train measurements in Figure 4 are described in the text but the figure does not show the pulse sequence timing, the number of averaged Hall measurements per pulse, or the correspondence between the numbered pulses and the individual images in panels b, d, and f. Adding pulse indices to the image grids would make the determinism/indeterminism claim directly verifiable.","section":"Figure 4a–4f"},{"comment":"The statement 'All data supporting the findings are available from the corresponding author(s) on reasonable request' is insufficient for a paper whose central evidence is a set of imaging datasets. A public repository deposit, even for one representative device cycle, would strengthen the manuscript considerably.","section":"Data availability"}],"recommendation":"major_revision","confidential_remarks":"The paper fits the journal's scope and the transport results appear sound, but the central imaging claim currently lacks the control experiment that would justify reading the ODMR contrast as a pure magnetization map. The stress-test concern raised by the skeptical reader is valid and is not answered by the a-axis/b-axis comparison. If the authors can supply a calibration or a nonmagnetic control, or alternatively reframe the claims as 'ODMR contrast changes correlated with the transport switching' rather than quantitative stray-field maps, the paper could become publishable. Without such a change, the novelty would be substantially weakened. I lean major_revision rather than reject because the deficiency is local and fixable in principle."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Best to read this as an imaging paper, not a switching paper: field-free SOT switching in WTe2/Fe3GaTe2 is already in the literature, and the authors cite it. What is genuinely new is the hBN sensor layer on top of a working device, and the direct visualization of domain nucleation, growth, and retraction during switching. That is a real step forward for vdW spintronics, and the a-axis vs b-axis comparison is a strong symmetry control.\n\nThe transport data are internally consistent and match prior reports. The imaging in Fig. 3 shows a plausible sequence of domains flipping at weak spots and then expanding, and the pulse-train results in Fig. 4 add texture to the deterministic-to-indeterministic crossover. The Joule-heating interpretation is inferred but sensible.\n\nThe soft spot is real: the stray-field maps are treated as magnetization images without a demonstrated control separating magnetic ODMR shifts from strain, temperature, charge, or Oersted effects. The hBN sits on a metallic stack carrying 6.5–8.5 mA pulses, so thermal expansion and strain gradients could create spatially varying, pulse-history-dependent ODMR contrast that resembles domain patterns. The a-b comparison does not fully exclude this, because the heating and strain profiles differ between current directions. Calibration against a known field and a control without magnetic order are not shown, and the extraction details are deferred to inaccessible Supplementary Notes. No data repository or error bars are provided. These are presentation gaps, not evidence of a fabricated effect; the transport results stand alone, and the imaging is consistent with the expected physics.\n\nThe stress-test note is fair but proportionate: this is a missing control, not a load-bearing contradiction. A serious referee should ask for a calibration control and a non-magnetic control, plus error bars and data availability. That is fixable in revision.\n\nRecommendation: send to peer review. The paper contributes a new capability and gives mechanistic insight that transport alone cannot. It deserves referees, with the caveat that the imaging claims need the missing controls. I'd cite it if I worked in this area, and it would make a good reading-group discussion about controls for biased hBN devices.","headline":"Genuinely new imaging of known SOT switching; missing controls on the ODMR contrast make the central claim not yet fully demonstrated, but it deserves review.","tokens_in":10333,"tokens_out":2982,"would_cite":true,"duration_ms":29839,"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":"This paper reports direct imaging of field-free deterministic magnetic switching in an all-van der Waals spin-orbit torque device using hBN spin ensembles.","keywords":["hexagonal boron nitride","quantum sensing","boron-vacancy centers","spin-orbit torque","field-free magnetic switching","Fe3GaTe2","WTe2","van der Waals heterostructures"],"falsifier":"Measure the same ODMR contrast on a control device whose Fe3GaTe2 layer is heated above its Curie temperature or replaced by a nonmagnetic flake while keeping the current and stack geometry identical; if spatially patterned contrast remains, the images are not purely magnetic. Alternatively, map the hBN zero-field splitting across the flake to quantify strain-induced shifts and check whether they exceed the reported stray-field signal.","tokens_in":9178,"feed_emoji":"🧲","tokens_out":6330,"duration_ms":55278,"temperature":0.7,"pith_summary":"This paper reports that spin defects embedded in the hexagonal boron nitride layer of a WTe2/Fe3GaTe2/hBN stack can image, in real space, how electrical current flips the magnetization of the atomically thin ferromagnet Fe3GaTe2 without any external magnetic field. The images show that field-free switching proceeds by nucleating reversed magnetic domains at local weak spots and then propagating them across the flake, information that electrical transport measurements alone cannot reveal. The same imaging shows why deterministic switching degrades at higher currents: local Joule heating intermittently drives part of the sample above its Curie temperature, so a fraction of domains no longer switches reproducibly. If correct, the work establishes hBN spin ensembles as a built-in microscope for spintronic devices made entirely of van der Waals materials.","feed_headline":"Quantum imaging reveals field-free switching in a 2D magnet","feed_subtitle":"Boron-vacancy sensors in hBN map Fe3GaTe2 domains as current pulses flip them, exposing heating-driven limits.","key_machinery":"The central object is the negatively charged boron-vacancy center (VB-) in the hBN encapsulation layer, an S = 1 spin defect whose optically detected magnetic resonance (ODMR) splitting is proportional to the local out-of-plane magnetic field. Because the hBN is part of the device stack, the same layer that protects the sample doubles as a wide-field quantum magnetometer, mapping the Fe3GaTe2 stray field across the flake. The symmetry argument comes from WTe2: charge current along the low-symmetry a-axis generates out-of-plane polarized spin currents, whereas current along the b-axis generates only conventional in-plane spin Hall polarization, so the imaging contrast between the two geometries isolates the role of out-of-plane spins.","core_discovery":"The central claim is that out-of-plane polarized spin currents from the low-symmetry spin source WTe2 deterministically switch the perpendicular magnetization of Fe3GaTe2 in the absence of an applied field, and that wide-field magnetometry using negatively charged boron-vacancy (VB-) spin ensembles in the hBN cap can directly visualize this switching. Stray-field maps recorded at points along the anomalous Hall loop show the magnetization evolving from a single domain to partially reversed domains that nucleate at preferential locations and expand as the write current increases, then retract and return to the initial state when the current polarity is reversed. At larger write currents the switching remains mostly deterministic in transport but is locally non-reproducible in imaging, which the paper attributes to Joule heating pushing some regions above the Curie point. When current is applied along the high-symmetry axis of WTe2, where only in-plane spin polarization exists, the images show randomly oriented domains with near-zero net magnetization, consistent with the absence of out-of-plane torques.","pith_inferences":["A testable extension would be to correlate the nucleation sites visible in these images with local structural features such as wrinkles, bubbles, or thickness variations; such a correlation would test whether defects control the switching threshold.","The same hBN sensing layer could simultaneously report temperature through the zero-field splitting shift, potentially providing a separate thermometer to separate Joule heating from spin-torque effects in the same device.","This approach could be applied to other low-symmetry spin sources to compare domain behavior across materials, although the paper does not do that.","If the non-reproducible domains at high current are caused by transient excursions above the Curie point, faster time-resolved imaging during the current pulse should catch the demagnetization-remanence cycle directly."],"forward_implications":["Field-free switching in WTe2/Fe3GaTe2 is a domain-nucleation-and-propagation process, not a coherent rotation, so device performance will depend on where nucleation sites sit.","Quantum imaging can serve as a microscopic counterpart to anomalous Hall measurements for evaluating SOT devices, catching local non-reproducibility that transport misses.","Joule heating sets an upper current bound for reliable deterministic switching, because local temperature excursions above the Curie point create randomly oriented domains.","The absence of out-of-plane spins (b-axis current) yields zero net magnetization with random nanoscale domains, so any apparent switching signal in transport is not deterministic.","Because hBN is already a standard encapsulation layer, this imaging approach can be added to existing all-vdW spintronic devices without changing the stack."],"supporting_citations":[{"why":"Establishes that WTe2's out-of-plane spin currents give field-free deterministic switching, the effect this paper images.","marker":"[8]"},{"why":"Demonstrates spin-orbit torque switching in an all-van der Waals heterostructure, the platform type under study.","marker":"[9]"},{"why":"Reports field-free deterministic switching above room temperature in the same WTe2/Fe3GaTe2 material system, the transport-level result being visualized.","marker":"[10]"},{"why":"Supplies the wide-field hBN spin-defect imaging technique used to record stray-field maps.","marker":"[14]"},{"why":"Shows quantum microscopy with van der Waals heterostructures, supporting integration of sensing with devices.","marker":"[15]"},{"why":"Establishes magnetic imaging with spin defects in hBN, grounding the quantitative field readout.","marker":"[16]"},{"why":"Demonstrates that WTe2 crystal symmetry controls spin-orbit torque polarization, motivating the a-axis versus b-axis comparison.","marker":"[18]"},{"why":"Reports Fe3GaTe2 as an above-room-temperature van der Waals ferromagnet with perpendicular anisotropy, justifying it as the switched layer.","marker":"[25]"}],"fun_headline_variants":["hBN spin sensors watch 2D magnet switch without any field","Quantum imaging shows field-free deterministic switching in vdW stack","Spin ensembles in hBN map field-free magnetic switching","All-vdW SOT system switched deterministically, seen via hBN sensors","Visualizing deterministic switching in 2D magnets with hBN spins"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The stray-field maps are read as magnetization images, which assumes the ODMR contrast of the boron-vacancy ensembles is set by the local out-of-plane magnetic field from Fe3GaTe2, with no significant contribution from strain, defect-density variation, thermal resonance shifts, or Oersted fields.","fun_headline_variants_meta":{"raw":{"variants":["hBN spin sensors watch 2D magnet switch without any field","Quantum imaging shows field-free deterministic switching in vdW stack","Spin ensembles in hBN map field-free magnetic switching","All-vdW SOT system switched deterministically, seen via hBN sensors","Visualizing deterministic switching in 2D magnets with hBN spins"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000574,"raw_usage":{"total_tokens":2724,"prompt_tokens":972,"completion_tokens":1752,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":588,"completion_tokens_details":{"reasoning_tokens":1662}},"tokens_in":588,"tokens_out":1752,"duration_ms":14476,"temperature":1.0,"reasoning_tokens":1662,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T22:18:13.662905+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the same ODMR contrast on a control device whose Fe3GaTe2 layer is heated above its Curie temperature or replaced by a nonmagnetic flake while keeping the current and stack geometry identical; if spatially patterned contrast remains, the images are not purely magnetic. Alternatively, map the hBN zero-field splitting across the flake to quantify strain-induced shifts and check whether they exceed the reported stray-field signal.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes that WTe2's out-of-plane spin currents give field-free deterministic switching, the effect this paper images."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates spin-orbit torque switching in an all-van der Waals heterostructure, the platform type under study."},{"cited_title":"N., Nguyen, T., Hung, N","cited_arxiv_id":null,"evidence_quote":"Reports field-free deterministic switching above room temperature in the same WTe2/Fe3GaTe2 material system, the transport-level result being visualized."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the wide-field hBN spin-defect imaging technique used to record stray-field maps."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows quantum microscopy with van der Waals heterostructures, supporting integration of sensing with devices."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes magnetic imaging with spin defects in hBN, grounding the quantitative field readout."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates that WTe2 crystal symmetry controls spin-orbit torque polarization, motivating the a-axis versus b-axis comparison."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports Fe3GaTe2 as an above-room-temperature van der Waals ferromagnet with perpendicular anisotropy, justifying it as the switched layer."}],"review_version":1}