{"id":"bbcf4752-a395-49b1-a33b-f5b5d2415465","arxiv_id":"2511.07181","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Nanoscale current imaging with a scanning NV magnetometer directly visualizes Lorentz deflection of current and maps contact resistance in a graphene-metal hybrid at fields up to 0.53 T.","lead":"Using a scanning single-spin diamond sensor, the authors imaged how electric current flows in a graphene-metal device under a magnetic field. The images directly show the Lorentz force bending the current, and reveal spatial variations in contact resistance that are invisible to ordinary resistance measurements.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"B_z=0 contour shift is used as direct evidence of Lorentz deflection, but the contour is not exactly a streamline for a finite-width/two-layer current distribution; the extracted Hall angle may be biased.","rationale":"The paper convincingly shows a sign-reversing shift of the B_z=0 contour with bias field, and the extracted mobility agrees with transport data, so the central qualitative claim of Lorentz deflection is likely correct. The main load-bearing risk is quantitative: the B_z=0 contour is used as a direct proxy for a current streamline, but for a finite-width, vertically extended current distribution this proxy is approximate. The reader's weakest assumption flagged the 2D-sheet reconstruction; I refine that to the specific use of the B_z=0 contour for the Hall-angle/mobility extraction, since that is the part of the central claim that is most sensitive to the neglected 3D structure. The proposed forward-model test would settle whether the apparent Hall angle is biased. Because the concern strengthens the need for the conditional treatment already given, the verdict is unchanged.","tokens_in":14417,"tokens_out":10714,"duration_ms":116409,"concrete_test":"Use the paper's finite-element model (Fig. S10, Tables S1/S2) at VBG=0 and the Q-V configuration to compute the current distribution at B=0 and B=±0.53 T. Generate B_z maps at z=100 nm for (i) a two-layer geometry with the metal disc raised 21 nm above the graphene and (ii) the standard 2D-sheet geometry. Extract the B_z=0 contour shift and apparent tanθ_H exactly as in Fig. 3c,d. If the apparent tanθ_H differs from the input graphene Hall angle by >20%, or if the two-layer and 2D-sheet results differ by >20%, the reported mobility and the quantitative strength of the Lorentz-deflection claim are biased by the 2D/zero-height assumption.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central observation is the B-field-dependent deflection of the B_z=0 contour near the injection contact, interpreted as the current's center streamline and used to extract tanθ_H and hence the mobility via Eq. (1). This interpretation requires the current to be an infinitely thin 2D sheet and the contour to coincide with a streamline. In the device, the current is distributed over a finite width and over two different heights: the graphene monolayer and the 21 nm-thick Cr/Au metal disc. The B_z=0 contour of the combined field depends on the vertical convolution kernel and the relative amplitudes of the two layers. Since the current partition between graphene and metal changes with |B| (from ~26% to ~15%, Fig. 3e), the contour position can shift in a way not simply related to the local Hall angle. Although the observed shift is odd in B, which argues against a pure partition artifact, the magnitude tanθ_H=0.5 is not validated by a forward model that includes the finite standoff z=100 nm, the 21 nm topographic step, and finite current width. Agreement with transport mobility is encouraging but uses the same two-carrier model and may share systematic errors. No independent check of the 'B_z=0 = streamline' identification is presented.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports scanning NV magnetometry imaging of current flow in a graphene–Cr/Au van der Pauw hybrid device at room temperature and out-of-plane bias fields up to ±0.53 T. The authors observe that the B_z=0 contour of the Oersted field near the injection contact deflects left or right depending on the sign of B_ext, interpret this as direct imaging of Lorentz deflection of current, and extract a Hall angle tangent of 0.5, translating to a carrier mobility μ ≈ 1.4 m^2/Vs. They additionally quantify a drop in the graphene-ring current fraction from ~26% to ~15% with field, and combine current images with resistance data and finite-element simulations to argue for a combination of intrinsic Dirac-plasma MR, carrier hydrodynamics, and spatially varying interface resistance. The paper includes a differential measurement scheme to reject back-gate leakage and presents a quantitative contact-resistance model.","tokens_in":14723,"tokens_out":6341,"duration_ms":64684,"significance":"The core observation—a current-path deflection that reverses with the sign of B_ext—is conceptually important and, if validated, demonstrates a new capability for nanoscale magnetotransport imaging at field strengths relevant to Hall physics. The use of (111)-oriented diamond tips and high-bias NV sensing is a technical advance. The odd-in-B behavior of the deflection provides internal evidence against a simple current-partition artifact, and the consistency of the extracted mobility with transport is encouraging. The manuscript also goes beyond a single snapshot by correlating local current maps with global MR and by simulating both diffusive and hydrodynamic transport, giving the claims a testable structure.","major_comments":[{"comment":"The mobility extraction relies on identifying the B_z=0 contour as the current's center streamline. For a finite-width current distribution at standoff z=100 nm, and especially with the Cr/Au disc carrying current ~21 nm above the graphene, the null contour of B_z is not generally the centroid streamline, and the back-propagation kernel in Methods Eqs. (6)-(8) treats all current as lying in one plane. No forward model or numerical phantom is used to show that the observed tangent 0.5 is an unbiased estimator of tan θ_H. The agreement with the two-carrier transport mobility is reassuring, but both estimates use the same two-carrier assumptions. Please provide a systematic uncertainty budget and an independent validation of the streamline identification.","section":"Fig. 3c-e, Eq. (1)"},{"comment":"Fig. 3e reports the graphene current fraction with no uncertainty. The linecut integration involves a boundary defined by where J_δ drops to zero, a tilt angle δ=-32.5°, and a reconstruction filter λ=2z; each of these choices can shift the fraction by several percent. Because the 26%→15% drop and its near-saturation above 0.1 T are used to argue against the EMR picture and to support the hydrodynamic/MR decomposition, the paper needs at least sensitivity analyses and pixel-level or scan-to-scan error bars for these values.","section":"Fig. 3e and Methods linecut analysis"},{"comment":"The hydrodynamic explanation of the zero-field anomaly is supported by a single simulation point at Dν=0.1 µm and by qualitative visual agreement in Fig. 4a. There is no quantitative goodness-of-fit or uncertainty on Dν, and the field-dependent suppression is not modeled through a field-dependent Dν or D_H. Eq. (9) includes a Hall-viscosity length D_H, but its value and role in the simulations are not stated. Please clarify how the hydrodynamic-to-diffusive crossover with B is implemented and quantify the evidence for hydrodynamics.","section":"Electron hydrodynamics and single-carrier transport, Fig. 3e, Fig. 4a"},{"comment":"The interface conductance distribution is optimized to match the measured current maps, but no uniqueness or cross-validation analysis is provided. Since one of the paper's claims is quantitative mapping of contact resistance, the authors should show how sensitive the recovered conductance map is to the regularization, geometry assumptions, and starting parameters, and ideally validate against the independently measured two-terminal resistances.","section":"Fig. 2 and Fig. S10 contact-resistance mapping"}],"minor_comments":[{"comment":"Typo in Acknowledgments: 'nanofabriation' should be 'nanofabrication'. Also 'van-der-Pauw' in the Fig. 1 caption should be 'van der Pauw'.","section":"Acknowledgments and Fig. 1 caption"},{"comment":"The 'center streamlines' shown in Fig. 3c are not defined in the Methods. Please specify how they are extracted from the B_z maps and how the deflection angle is computed from them.","section":"Fig. 3c"},{"comment":"The sign convention for δ (the linecut tilt angle) is not stated. Also clarify whether the same linecut is applied to all datasets or whether it is redefined per scan.","section":"Methods, linecut analysis"},{"comment":"No error bars are shown for R(B_ext) or MR. The dotted and dashed lines in Fig. 1e are not fully explained in the caption; please define them explicitly.","section":"Fig. 1e"},{"comment":"The phase expression appears garbled in the text: it should be φ = γ (π/2) τ B_NV or similar. Please correct the typesetting.","section":"Methods, Eq. (3)"},{"comment":"The values n_e=0.34×10^11 cm^-2 and n_h=1.74×10^11 cm^-2 are quoted without uncertainty. State whether they come from the capacitance model, the two-carrier fit, or a combination, and give their estimated errors.","section":"Eq. (1)"}],"recommendation":"major_revision","confidential_remarks":"This is a strong experimental paper and the core imaging observation is likely correct. My main concern is that the quantitative claims—particularly the mobility from Eq. (1) and the current-fraction numbers in Fig. 3e—are presented without error bars or a forward-model check of the 'B_z=0 = streamline' identification. These are fixable with additional analysis, and I would be comfortable seeing the paper in a high-quality journal after such a revision. The circularity concern raised by the reader is not fatal: the odd-in-B deflection is a direct measurement and is not generated by the fitting model."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nYou should know two things about arXiv:2511.07181 before reading: the central observation—Lorentz deflection of current visible in the B_z=0 contour, reversing with field sign—is probably real, and the paper earns its place as the first scanning-NV magnetotransport imaging above 0.5 T. Everything quantitative beyond that needs handling with care.\n\nWhat's actually new: previous NV current imaging stopped at near-zero field. Here they push to ~0.5 T with a (111) diamond probe and a permanent-magnet stack, and they see the Hall-angle bending of streamlines at a graphene-metal interface. That's a capability advance and a clean demonstration. The field-reversal data in Fig. 3c is the right control: the contour bends left for one sign, right for the other, matching the Lorentz force. Extracting µ≈1.4 m²/Vs from tanθ_H and getting agreement with the transport value is a nice cross-check. The contact-resistance mapping from current imaging, and the leakage-current differential scheme in Methods, are also solid technical contributions.\n\nSoft spots, in proportion: The B_z=0 contour is not exactly a streamline when the current lives at two heights (graphene plus 21 nm metal disc) and has finite width. The stress-tester's note is fair: the odd-in-B behavior argues against a pure partition artifact, but the reported tanθ_H=0.5 magnitude isn't checked against a forward model that includes the standoff, the topography step, and the two-layer current distribution. That's a real gap, though it mostly affects the quantitative mobility value, not the existence of the deflection. The current-fraction numbers (26%→15%) and the contact-resistance map inherit the same 2D-sheet reconstruction assumption, and there are no error bars. The hydrodynamic interpretation rests on one simulation with D_ν=0.1 µm; it's plausible but not decisive. The finite-element model optimizes several parameters (interface conductances, µ_0, n_e, n_h) to the same dataset, so the agreement with \"diffusive plus hydrodynamics\" is not an independent test. No data/code deposit is mentioned—for a methods-forward paper, that's a fixable omission.\n\nWho this is for: anyone working on NV magnetometry, graphene transport, or geometrical MR effects. It deserves a serious referee; the main claims should survive with the quantitative sections tightened and error bars added.\n\nMy recommendation: send it to review, and push for the forward-model check of the streamline identification plus uncertainty quantification on the mobility.","headline":"First real imaging of magnetotransport with NV magnetometry; the central deflection claim is solid, but the quantitative mobility, current-fraction, and hydrodynamic interpretations rest on model assumptions that need tighter validation.","tokens_in":15226,"tokens_out":1962,"would_cite":true,"duration_ms":20399,"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 scanning single-spin quantum sensor directly images the Lorentz deflection of current at a graphene-metal interface under an out-of-plane bias field.","keywords":["nitrogen-vacancy magnetometry","current imaging","magnetotransport","Hall effect","Lorentz deflection","graphene-metal interface","contact resistance","extraordinary magnetoresistance"],"falsifier":"Image the same device at two standoff distances (e.g., z = 100 nm and z = 150 nm) and check that the reconstructed current distribution is unchanged; a systematic difference would falsify the 2D-sheet reconstruction and with it the extracted Hall mobility. Alternatively, replace the metal disc with a very thin conductor of precisely known height and compare the reconstructed current to a finite-element model.","tokens_in":14308,"feed_emoji":"🧲","tokens_out":2915,"duration_ms":30426,"temperature":0.7,"pith_summary":"The paper aims to show that a scanning nitrogen-vacancy (NV) magnetometer can image magnetotransport at the nanoscale, specifically the Lorentz deflection of current near a graphene–metal interface under bias fields around 0.5 T. By reconstructing current-density maps from measured stray fields, the authors observe the injected current's center streamline bend left or right with the sign of the field—the spatial hallmark of the Hall effect. From the deflection angle they extract a carrier mobility of about 1.4 m²/(V·s), matching resistance-based estimates, and they map spatial variations in contact resistance across the interface. The work positions nanoscale current imaging as a way to probe transport mechanisms—such as the interplay of intrinsic Dirac-cone magnetoresistance, electron hydrodynamics, and interface resistance—that bulk resistance measurements cannot resolve.","feed_headline":"NV magnetometer images current bending at graphene-metal edge","feed_subtitle":"Scanning single-spin sensor resolves Lorentz deflection under 0.5 T and maps contact resistance.","key_machinery":"The central object is a scanning NV single-spin magnetometer operated at high bias fields, which maps the out-of-plane stray field of the device current via spin-echo AC sensing. Current density maps are recovered by Fourier back-propagation of the stray field (Biot–Savart inversion), and the B_z = 0 contour serves as a direct visual proxy for the center streamline of the current. The deflection angle of this streamline under an applied field defines the Hall angle, whose tangent yields a direct spatial measurement of carrier mobility.","core_discovery":"The central claim is that at elevated out-of-plane bias fields (~0.5 T) the current flowing from a metallic contact into a graphene annulus is visibly deflected by the Lorentz force, and that this deflection can be imaged directly with a scanning single-spin quantum magnetometer. The zero-contour of the measured out-of-plane stray-field component marks the current streamline; it shifts left or right with field sign, and the tangent of the resulting Hall angle gives a carrier mobility of ~1.4 m²/(V·s), consistent with transport data. The same imaging also reveals a counterintuitive current reorganization—less current in the graphene ring at higher fields—which the authors attribute to a combi","pith_inferences":["The same imaging approach could be extended to ballistic and quantum-Hall regimes, where streamline deflection by edge states should appear as sharp current-channel shifts; the paper lists these as future directions.","The reconstruction's reliance on a known standoff distance and a 2D current sheet could be cross-checked by imaging the same device at multiple standoff heights—agreement would validate the assumption, disagreement would bias the extracted mobility.","Because the Hall-angle extraction uses only the streamline geometry, it may work in devices with strong inhomogeneity or disorder where contact-based Hall measurements are unreliable.","Combining this current-imaging method with simultaneous electrostatic-potential imaging would allow a direct, spatially resolved comparison between the Hall potential and current deflection in the same scan."],"forward_implications":["Scanning NV magnetometry becomes a general tool to spatially resolve magnetotransport in hybrid devices at fields up to ~0.5 T and beyond.","The Hall-angle-from-streamline method provides a contact-free mobility measurement, valuable in devices where conventional four-terminal probes are difficult to fabricate.","Contact-resistance maps derived from the current density can diagnose non-uniform interfaces in two-dimensional-material devices, which bulk resistance alone cannot reveal.","The observed suppression of electron hydrodynamics by a magnetic field suggests a route to image viscous versus diffusive flow regimes in the same device.","The decomposition of magnetoresistance into intrinsic Dirac-cone MR, geometrical MR, and hydrodynamic contributions can be tested in other materials by comparing local current maps with global resistance."],"fun_headline_variants":["Single-spin sensor images Lorentz deflection at graphene edge","Direct current-flow imaging reveals Hall angle at graphene-metal contact","Nanoscale magnetometry sees current bend in graphene under 0.5 T","Quantum sensor maps current paths and contact resistance in graphene device","Lorentz deflection imaged at graphene-metal junction by single-spin NV"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The reconstructed current maps assume the current flows in a strictly two-dimensional sheet at the graphene layer, with a known standoff distance and no out-of-plane current or height variation between graphene and the metal disc; if that assumption breaks down, the deflections and contact-resistance maps are biased.","fun_headline_variants_meta":{"raw":{"variants":["Single-spin sensor images Lorentz deflection at graphene edge","Direct current-flow imaging reveals Hall angle at graphene-metal contact","Nanoscale magnetometry sees current bend in graphene under 0.5 T","Quantum sensor maps current paths and contact resistance in graphene device","Lorentz deflection imaged at graphene-metal junction by single-spin NV"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000592,"raw_usage":{"total_tokens":2610,"prompt_tokens":740,"completion_tokens":1870,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":484,"completion_tokens_details":{"reasoning_tokens":1796}},"tokens_in":484,"tokens_out":1870,"duration_ms":13370,"temperature":1.0,"reasoning_tokens":1796,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T23:05:23.614498+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Image the same device at two standoff distances (e.g., z = 100 nm and z = 150 nm) and check that the reconstructed current distribution is unchanged; a systematic difference would falsify the 2D-sheet reconstruction and with it the extracted Hall mobility. Alternatively, replace the metal disc with a very thin conductor of precisely known height and compare the reconstructed current to a finite-element model.","supporting_citations":[],"review_version":1}