{"id":"a5b56a7e-7877-48e7-b7d0-a103e2975279","arxiv_id":"2608.04632","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":10,"one_line_summary":"A calibration protocol uses PMA stripe and disc stray fields to extract NV-to-sample distance and azimuthal angle, plus an inverse-AFM fluorescence dip to estimate lateral NV position in the diamond pillar.","lead":"This paper shows how to calibrate a scanning diamond magnetometer by measuring the stray fields of known magnetic stripes and discs, yielding the sensor's height above the sample (about 31.5 nm) and its in-plane angle without an extra vector magnet. A sharper silicon needle then maps the diamond tip itself, revealing contamination and roughly locating the sensor inside the pillar.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Stripe-based d_NV extraction depends on the unknown azimuthal phi, contrary to the paper's 'blind' claim; the calibration sequence is conditional on a prior orientation assumption.","rationale":"After carefully reading the manuscript, the most fragile link in the central claim is the logical ordering of the calibration. The reader's weakest assumption about the constant NV-sample distance across the etch step is a plausible experimental systematic, and the paper's own admission that the inverse-AFM simulations do not quantitatively match the data weakens the localization claim. However, neither of these is as fundamental as the internal relationship between the stripe line-shape and the unknown azimuthal angle. The manuscript's equations imply that the measured B_NV across a stripe edge is a mixture of Lorentzian and dispersive forms whose mixing ratio depends on cos(phi); Section 2.3's list of fitted parameters (x0, B0, A, d_NV) omits phi. If phi was fixed to the nominal 90-degree arm during the stripe fits, the reported d_NV = 31.5 nm is conditional on an orientation assumption that is supposed to be one of the outputs of the procedure. The later disc measurement can determine phi independently of d, so a corrected protocol could run the disc analysis first (or fit both amplitudes in the stripe model), but as written the paper overclaims the directness of the inference. A straightforward refit of the published data (available on Zenodo) with two independent amplitudes would settle whether the current d_NV is robust. This is a scientific concern about the argument, not about the authors' conduct, and the proposed test is feasible from the deposited data. I therefore recommend keeping the verdict at CONDITIONAL, with the specific condition that the authors demonstrate the phi-independence of the stripe fit or revise the calibration order.","tokens_in":13646,"tokens_out":17311,"duration_ms":211148,"concrete_test":"Refit the ten stripe line scans with the full model B_NV(x) = B0 + c_L * d/((x-x0)^2+d^2) + c_D * (x-x0)/((x-x0)^2+d^2), treating c_L and c_D as independent free parameters (equivalently, fit A and phi separately). Compare the resulting d_NV and its spread with the reported 31.5 +/- 1.95 nm, and check whether the fitted phi is consistent with the disc-derived ~97 degrees. Additionally, run a synthetic test: generate a noiseless stripe profile with d = 31.5 nm and phi = 0 degrees, then apply the fitting procedure of Section 2.3 with phi fixed to 90 degrees; if the recovered d deviates by more than 10%, the stripe method is not orientation-agnostic.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The most load-bearing concern is that the stripe-based d_NV extraction is not independent of the unknown azimuthal angle phi, despite the paper's claim that the stripe geometry 'remains blind' to phi. Combining Eqs. (2), (4), and (5), the measured projection is B_NV(x) = (A/(2 pi)) [sin(theta) cos(phi) d/(u^2+d^2) - cos(theta) u/(u^2+d^2)] + B0, with u = x - x0. The line shape is a linear combination of a Lorentzian and a dispersive term whose relative weight is set by sin(theta) cos(phi) / cos(theta). For the reported phi ~ 97 degrees (cos(phi) ~ -0.12) the profile is nearly a pure dispersion, so a fit that fixes phi = 90 degrees or ignores phi works. For a probe with phi = 0 or 180 degrees, the profile is a mixture and the same fitting procedure would bias the recovered d_NV substantially; the extremum separation scales as 2 d sqrt(1 + tan^2(theta) cos^2(phi)). Section 2.3 lists only x0, B0, A, and d_NV as free parameters of the stripe fit, with no phi or separate Lorentzian/dispersive amplitudes, so the procedure as written is not orientation-agnostic. The disc measurement later determines phi robustly from the angular position of the bright arc, but d_NV is fixed before phi is known; a biased d_NV would propagate into the disc fit and the reported M_s (908 vs 1020 kA/m). The paper must either demonstrate that the stripe line shape yields d_NV independently of phi (e.g., by fitting two independent amplitudes) or state the assumed phi and validate consistency with the disc result.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a three-step calibration protocol for scanning NV magnetometry. From line scans across PMA stripe edges, the authors extract the NV-to-sample standoff distance d_NV = 31.5 ± 1.95 nm; from two-dimensional maps of PMA discs they extract the NV azimuthal angle φ ≈ 97° (values 98.6°, 97.3°, 95.4°); and from inverse-AFM scans over silicon needles they assess probe contamination and attempt to localize the NV lateral position from a fluorescence dip. The stated advantage is that these parameters are obtained without an external vector magnet, using only simple patterned reference structures.","tokens_in":14151,"tokens_out":8143,"duration_ms":88168,"significance":"If the method is quantitatively reliable, it addresses a real and widely felt need in scanning NV magnetometry: the NV-sample distance and azimuthal orientation are typically unknown or only loosely constrained, and they strongly influence quantitative field reconstruction. The paper includes useful elements: a cross-check of the stripe-derived magnetization against SQUID data, reproducible φ values across three discs, ten statistically analyzed stripe line scans, and publicly archived data. The inverse-AFM part is more exploratory but could be valuable for probe quality control. The main gaps are that the stripe fit is not actually independent of φ, the constancy of d_NV across the topographic step is assumed rather than demonstrated, the fluorescence-dip localization is not quantitatively calibrated, and the reported 'precision' of 3° lacks a proper uncertainty analysis.","major_comments":[{"comment":"The claim that the stripe geometry is 'blind' to the azimuthal angle φ is not correct. Combining Eqs. (4) and (5), the measured projection is B_NV(x) = (A/2π)[sinθ cosφ d/(u²+d²) − cosθ u/(u²+d²)] + B0, so the line shape is a linear combination of Lorentzian and dispersive terms whose relative weight is set by cosφ. The fitting procedure in Sec. 2.3 lists only x0, B0, A and d_NV as free parameters, which implicitly fixes the NV in-plane orientation (likely at the φ = 90° value, for which n_x = 0). For the reported φ ≈ 97° this approximation changes the extremum separation by only about 1.5%, but for a probe with φ = 0° or 180° the extremum separation scales as 2d√(1 + tan²θ cos²φ), which would overestimate d_NV by roughly 73%. Since d_NV is subsequently used as a fixed input to the disc fits, this dependence is load-bearing. The authors should either fit independent Lorentzian and dispersive amplitudes (which would make d_NV identifiable without knowing φ) or explicitly state the assumed φ and propagate its uncertainty into d_NV and the later disc analysis.","section":"Sec. 2.2 and 2.3"},{"comment":"The extraction of d_NV from stripe edge scans assumes that the NV-to-sample standoff remains constant while the flat-bottomed pillar crosses the 40 nm etch step. The argument in Sec. 2.5 — that the apex descends onto the substrate only after completely passing the edge — is plausible but not verified. A small pillar tilt, an off-center NV, or AFM setpoint drift could change the effective contact point during the crossing, which would bias every fitted d_NV and propagate into the disc-derived φ and M_s. The agreement of the fitted M_s with SQUID is encouraging, but it does not by itself rule out a correlated error in d_NV and A. I recommend a control experiment, for example line scans on structures with different step heights or with varied AFM setpoints, to bound this systematic error.","section":"Sec. 2.5 and Fig. 5"},{"comment":"The inverse-AFM localization claim is not quantitatively established. The FDTD simulation shows a 31.5 nm offset between the fluorescence-dip minimum and the dipole position, and the simulated contrast and FWHM do not reproduce the measured values, as the text acknowledges. Nevertheless, the paper interprets experimental dip positions as NV lateral positions (for instance, the 80 nm off-center shift for the MX+ probe). Without a validated correction for the offset or an uncertainty estimate, the localization is at best qualitative. The authors should either validate the mapping by an independent method, apply a theoretically justified correction with an uncertainty, or explicitly restrict the claim to qualitative localization.","section":"Sec. 3 and Fig. 11"},{"comment":"The stated 'precision of 3°' for the azimuthal angle is based on only three discs (φ = 98.6°, 97.3°, 95.4°) and no per-fit uncertainties are reported. The 3° value is the range of the three measurements, not a standard deviation or confidence interval. In addition, the sensitivity of the fitted φ to the fixed value d_NV = 31.5 nm is not discussed. To support the word 'traceable' in the title and abstract, the authors should provide fit uncertainties for each disc and propagate the d_NV uncertainty into φ.","section":"Sec. 3 and Fig. 6"}],"minor_comments":[{"comment":"In the description of Fig. 8, 'we this fluorescence dip to the position' should read 'we attribute this fluorescence dip to the position'.","section":"Sec. 3, Fig. 8"},{"comment":"The caption contains 'obtained by diving the collected farfield emission'; 'diving' should be 'dividing'.","section":"Fig. 11 caption"},{"comment":"The text contains 'arising from the the Pt/Co interface'; the duplicated article should be removed.","section":"Sec. 2.5"},{"comment":"The quantity P is used in Eq. (6) but defined only in the broken line following Eq. (7); for readability, define P and q together with the other symbols before Eq. (6).","section":"Eqs. (6)–(7)"},{"comment":"The 1.95 nm standard deviation is described as the spread across ten line scans; please clarify whether this is the standard deviation of the population or of the mean, and give the combined uncertainty that includes the per-fit covariance errors.","section":"Fig. 5(b)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within scope for cond-mat.mes-hall and presents a useful practical calibration approach. The main reservation is that the abstract's claims of 'traceable' calibration and a stripe geometry that is 'blind' to φ are stronger than the analysis supports. The stripe fit's hidden φ dependence, the unverified constant-height assumption, and the uncalibrated inverse-AFM offset are load-bearing and require additional experiments or reanalysis. A major revision is appropriate; a minor revision would not suffice."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core of this paper is good: an integrated, vector-magnet-free protocol for extracting NV–sample distance and azimuthal angle from images of PMA stripes and discs, with a cross-checked magnetization. The stripe fits are the strongest part—ten line scans, covariance uncertainties, and an M_s value that agrees with SQUID. That alone makes the paper worth engaging. The disc-asymmetry idea for φ is also sound, and the reported values are consistent with the expected crystallographic orientation.\n\nBut the paper overclaims in two places. First, the statement that the stripe geometry is 'blind' to φ is wrong. The measured projection is a linear combination of a Lorentzian and a dispersive term, with the relative weight set by cosφ. For this probe, φ≈97°, so the dispersive term dominates and the fit works, but the procedure as written is not orientation-agnostic. A probe with φ near 0° would give a mixed line shape, and fitting with the fixed-ratio model would bias d_NV. The fix is simple—fit two independent amplitudes or explicitly state an assumed φ and verify with the disc—but the current text misleads.\n\nSecond, the inverse-AFM localization claim is weaker than the presentation suggests. The simulations show a 31.5 nm offset between the dip minimum and the dipole position, and the simulated contrast and FWHM do not match the measured values. The paper acknowledges this, but then still concludes that the dip position 'indicates' the NV's lateral placement. That is too strong; at best it is a rough localization with a pending calibration offset.\n\nMinor points: the azimuth 'precision of 3°' is a spread across three discs, not an uncertainty on a single fit. The M_s discrepancy between discs and stripes is discussed plausibly but remains unresolved. The calibration time (14–19 h) is practical, and the data availability statement is good.\n\nFor whom: experimental NV magnetometry groups, especially those lacking vector magnets. It deserves serious peer review, but it needs revision: correct the blindness claim, add per-fit uncertainties for φ, and temper or repair the localization section.","headline":"A useful NV calibration protocol with a flawed 'blindness' claim and an overreaching localization section; the d_NV and φ extraction itself is credible and deserves referee time.","tokens_in":14672,"tokens_out":3007,"would_cite":true,"duration_ms":38044,"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 NV magnetometer can calibrate its own sensor-sample distance and NV-axis azimuth from stray-field images of simple patterned magnetic films, without a vector magnet.","keywords":["scanning NV magnetometry","nitrogen-vacancy centers","PMA microstructures","stray-field calibration","NV-to-sample distance","azimuthal orientation","inverse AFM","diamond nanopillar"],"falsifier":"Scan the same PMA stripe edge on a planarized sample that has no topographic step and compare the fitted $d_{\\rm NV}$ with the value from the 40 nm step sample; if the two differ by more than the 1.95 nm scatter, the constant-height assumption across the etch step is falsified.","tokens_in":13483,"feed_emoji":"🧲","tokens_out":14418,"duration_ms":135077,"temperature":0.7,"pith_summary":"This paper establishes a three-step calibration routine that turns a scanning nitrogen-vacancy (NV) magnetometer into a self-calibrating instrument: the distance between the NV sensor and the sample, and the azimuthal angle of the NV spin axis, are read off from stray-field images of simple patterned magnetic structures instead of being treated as unknown fitting parameters. The central numbers are $d_{\\rm NV} = 31.5 \\pm 1.95$ nm from ten stripe-edge line scans and $\\phi \\approx 97^\\circ$ with a $3^\\circ$ spread from three discs, all obtained without an external vector magnet. A third step, scanning the diamond pillar over an ultra-sharp silicon needle, reveals probe contamination and gives a rough lateral localization of the NV inside the pillar through a fluorescence dip. If the method holds, quantitative NV magnetometry becomes more reproducible and portable, since calibration can be repeated before each measurement set using only cleanroom-fabricated PMA structures.","feed_headline":"Stray-field images fix the NV sensor's height and orientation","feed_subtitle":"No vector magnet needed: stripes and discs yield the sensor height (31.5 nm) and axis angle (97 degrees).","key_machinery":"The argument is carried by two analytic stray-field models. For a semi-infinite PMA stripe edge, the field components $B_x(x) = \\frac{A}{2\\pi}\\frac{d_{\\rm NV}}{(x-x_0)^2+d_{\\rm NV}^2}$ and $B_z(x) = -\\frac{A}{2\\pi}\\frac{x-x_0}{(x-x_0)^2+d_{\\rm NV}^2}$ make $d_{\\rm NV}$ the width-setting parameter of the measured projection $B_{\\rm NV}(x) = n_x B_x + n_z B_z + B_0$, so fitting the full profile with $x_0$, $B_0$, and the amplitude $A = \\mu_0 M_s$ as free parameters isolates $d_{\\rm NV}$. For a PMA disc, the field of a current loop of radius $R$ carrying current $I = \\sigma M_s$ yields $B_{\\rm NV,\\phi}(r,\\alpha) = B_0 + \\cos\\theta\\, B_z(r,d_{\\rm NV}) + \\sin\\theta\\, B_r(r,d_{\\rm NV}) \\cos(\\alpha-\\phi)$, so the angular position of the bright arc along the disc edge encodes $\\phi$ while the disc interior brightness fixes the magnetization direction $\\sigma$. The inverse-AFM localization rests on a different mechanism: the silicon needle, placed in the optical near field of the NV dipole, quenches the collected fluorescence, producing a dip whose minimum is associated with the NV's lateral position, as supported by FDTD simulations.","core_discovery":"The paper's central claim is that a single NV center in a diamond nanopillar can determine its own two most uncertain geometric parameters, the standoff distance $d_{\\rm NV}$ and the azimuthal angle $\\phi$ of the NV axis, directly from the magnetic stray fields of simple patterned perpendicularly magnetized films. Fitting line scans across a PMA stripe edge to the analytic dipole-sheet field gives $d_{\\rm NV} = 31.5 \\pm 1.95$ nm as the mean of ten cuts, and fixing this value in a two-dimensional fit of a PMA disc to a current-loop field model yields $\\phi = 98.6^\\circ$, $97.3^\\circ$, and $95.4^\\circ$ on three discs, consistent with the expected $90^\\circ$ orientation and reproducible to about $3^\\circ$. The same fits also return a stripe magnetization $M_s = 1020 \\pm 56$ kA/m, matching SQUID measurements, and a disc value of $M_s = 908 \\pm 89$ kA/m. Separately, inverse AFM scans over sharp silicon needles reveal contamination and a fluorescence dip whose position is argued, with FDTD simulation support, to mark the NV's lateral location inside the pillar.","pith_inferences":["The authors do not automate the routine, but the stripe/disc pair is simple enough that the calibration could be integrated into the standard start-up sequence of any NV microscope, including cryogenic systems where vector magnets are absent.","Because the FDTD simulation shows a 31.5 nm offset between the fluorescence-dip minimum and the true dipole position, using the dip center as the NV's lateral coordinate would need a position-dependent correction; a full 3D simulation with a realistic pillar apex and the NV's two-dipole emission could quantify that correction.","The 11% lower $M_s$ extracted from discs than from stripes implies a size-dependent apparent magnetization for small PMA discs; comparing discs of several radii would test whether edge damage from e-beam lithography and ion etching is the cause.","The $\\phi \\approx 97^\\circ$ result doubles as a probe-mounting diagnostic; collecting this angle for many probes would give a statistical measure of gluing misalignment in commercial sensor assemblies."],"forward_implications":["After calibration, $d_{\\rm NV}$ and $\\phi$ are known for a probe, so quantitative field reconstruction no longer needs them as free fit parameters, removing a major source of ambiguity in NV magnetometry.","The calibration can be repeated before each measurement session because it needs only a cleanroom-patterned PMA stripe and disc; this would track drift in standoff caused by contamination or AFM feedback changes.","The measured $\\phi \\approx 97.1^\\circ$ versus the nominal $90^\\circ$ shows that gluing of the probe to the tuning fork introduces a mounting misalignment, so the orientation must be measured per probe rather than assumed from geometry.","Inverse AFM topography on sharp silicon needles reveals contamination (here a 19 nm feature on the apex) that would otherwise corrupt the magnetic standoff, and the simultaneously recorded fluorescence dip gives a rough lateral NV localization inside the pillar."],"supporting_citations":[{"why":"Supplies the analytic stripe-edge stray-field expressions that are fitted to line scans to extract the NV-sample distance.","marker":"[13]"},{"why":"Also supplies the stripe-edge field model and demonstrates its use for patterned ultrathin ferromagnets in scanning nanomagnetometry.","marker":"[14]"},{"why":"Gives the closed-form analytic expressions for the magnetic field of a circular current loop in terms of complete elliptic integrals, used for the disc model.","marker":"[16]"},{"why":"Documents the implantation depth and straggle uncertainty that makes per-probe d_NV calibration necessary.","marker":"[8]"},{"why":"Shows how AFM feedback and standoff distance affect the sensor-sample distance, supporting the need for repeatable in-situ calibration.","marker":"[10]"},{"why":"The FDTD simulator used to model the fluorescence dip in inverse AFM scans and justify associating the dip position with the NV's lateral location.","marker":"[17]"},{"why":"Defines the commercial probe geometry (flat (100) apex and (110) side facets) used to compare the measured azimuthal angle with expected values.","marker":"[21]"}],"fun_headline_variants":["Stray fields alone reveal NV height and axis angle","NV center self-locates its height and orientation","NV magnetometer self-measures its standoff and orientation","31.5 nm standoff inferred from stray-field images","NV center self-calibrates its standoff and orientation"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole distance calibration rests on the assumption that the NV's height above the sample stays exactly the same while the flat-bottomed diamond pillar crosses the 40 nm etched step at the stripe edge, so the fitted width of the stray-field profile reflects only the NV-sample distance.","fun_headline_variants_meta":{"raw":{"variants":["Stray fields alone reveal NV height and axis angle","NV center self-locates its height and orientation","NV magnetometer self-measures its standoff and orientation","31.5 nm standoff inferred from stray-field images","NV center self-calibrates its standoff and orientation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00086,"raw_usage":{"total_tokens":3751,"prompt_tokens":983,"completion_tokens":2768,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":599,"completion_tokens_details":{"reasoning_tokens":2689}},"tokens_in":599,"tokens_out":2768,"duration_ms":19136,"temperature":1.0,"reasoning_tokens":2689,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T20:16:23.786683+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Scan the same PMA stripe edge on a planarized sample that has no topographic step and compare the fitted $d_{\\rm NV}$ with the value from the 40 nm step sample; if the two differ by more than the 1.95 nm scatter, the constant-height assumption across the etch step is falsified.","supporting_citations":[{"cited_title":"PhD thesis, University of Cam- bridge, 2023","cited_arxiv_id":null,"evidence_quote":"Supplies the analytic stripe-edge stray-field expressions that are fitted to line scans to extract the NV-sample distance."},{"cited_title":"Hingant, J.-P","cited_arxiv_id":null,"evidence_quote":"Also supplies the stripe-edge field model and demonstrates its use for patterned ultrathin ferromagnets in scanning nanomagnetometry."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the closed-form analytic expressions for the magnetic field of a circular current loop in terms of complete elliptic integrals, used for the disc model."},{"cited_title":"Maletinsky, S","cited_arxiv_id":null,"evidence_quote":"Documents the implantation depth and straggle uncertainty that makes per-probe d_NV calibration necessary."},{"cited_title":"Palm, William Huxter, Konstantin Herb, John M","cited_arxiv_id":null,"evidence_quote":"Shows how AFM feedback and standoff distance affect the sensor-sample distance, supporting the need for repeatable in-situ calibration."},{"cited_title":"Lumerical fdtd: 3d electro- magnetic simulator","cited_arxiv_id":null,"evidence_quote":"The FDTD simulator used to model the fluorescence dip in inverse AFM scans and justify associating the dip position with the NV's lateral location."},{"cited_title":"QuantileverMX product brochure, 2021","cited_arxiv_id":null,"evidence_quote":"Defines the commercial probe geometry (flat (100) apex and (110) side facets) used to compare the measured azimuthal angle with expected values."}],"review_version":1}