{"id":"bf423cb3-16c4-4016-93d2-628d32ba7a63","arxiv_id":"2506.04448","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Circularly polarized microwaves from a cross-shaped waveguide selectively drive the spin 0 to -1 or 0 to +1 transitions of boron vacancy defects in hBN, with an asymmetric magnetic field dependence.","lead":"Researchers made a microwave device that emits circularly polarized waves and used it to selectively address the two spin transitions of boron vacancy defects in hexagonal boron nitride. The work gives 2D quantum sensors a new control knob, potentially improving magnetic sensing at low fields.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The circular-polarization mechanism is not established outside the 2–5 mT standard orientation: the SI reports 140°, 240°, and ~90° phase separations instead of 180°, so the observed selectivity may be partly a projection or fitting artifact rather than clean helicity selection.","rationale":"I read the paper in good faith. The phase-dependent ODMR at 2.3 mT is compelling: the two transitions modulate out of phase, the optimal phases are separated by roughly 180°, and the Lindblad calculation reproduces the pattern after a −30° offset. That is real evidence of polarization-sensitive driving. The concern is not that the experiment is wrong; it is that the SI's own data violate the central symmetry in regimes the authors themselves cannot model. The inverted-field result is the most load-bearing because it is a clean symmetry test: Eq. 3 predicts a 180° separation regardless of the sign of B0, and a 90° separation means the effective driving field is not described by Eq. 3. This does not necessarily invalidate the 2.3 mT demonstration, but it does mean the mechanism is not universal and the quoted high selectivities may be regime-specific or partly fit artifacts. The reader identified the same weakest assumption, unmeasured polarization and phase offset, and I agree. The paper itself flags the deviations in the SI and states that the theoretical calculations do not capture them, which reinforces rather than resolves the concern. I therefore recommend keeping the verdict conditional, with no change from the reader's assessment.","tokens_in":12732,"tokens_out":8122,"duration_ms":89411,"concrete_test":"Calibrate the microwave magnetic-field ellipse at the sample position using a small shielded pickup loop and a vector network analyzer, measuring amplitude and phase of both orthogonal components with the same cross-waveguide, terminations, and magnet orientation used in the ODMR experiment. Then repeat the reversed-B0 phase scan at 4.2 mT while recording the actual ellipse. If the measured field is circular (axial ratio greater than about 0.9) and the 90° phase separation persists, Eq. 3 and the helicity-selection mechanism are falsified. If the measured field is elliptical or tilted, recompute the predicted phase separation from the measured Stokes parameters; if it reproduces 90°, the selectivity is a polarization-projection effect rather than clean angular-momentum selection, and the central claim must be qualified accordingly.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires the microwave field at the defect to be circular in the plane perpendicular to the c-axis (Eq. 3), with equal orthogonal amplitudes and a known 90°/270° phase difference. The paper infers the phase offset (−30°) from the data rather than measuring it directly, and the SI documents systematic breakdowns of this picture: the phase separation between the two maximum-selectivity conditions is 140° at zero field and 240° at 8.2 mT instead of 180°, and for reversed B0 the separation is about 90°. Reversing B0 is a sharp symmetry test: with fixed microwave helicity, field reversal should merely exchange which transition (lower vs. upper frequency) is favored at a given phase, leaving the 180° separation between the two maximum-selectivity phases intact. A 90° separation cannot be produced by Eq. 3 and indicates elliptical or tilted polarization, an in-plane field component, or a projection artifact. Because the Lindblad model is partly fitted to the same data and fails to reproduce these deviations, the claim that the observed ODMR asymmetry is clean angular-momentum selection is not established outside the central 2–5 mT, standard-orientation regime. The headline selectivity values (91.4% and 83.5%) are also area ratios from double-Lorentzian fits to overlapping lines, so their absolute values carry systematic fitting ambiguity not included in the reported error bars.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports spin-state selective excitation of negatively charged boron vacancy (V_B^-) defects in hexagonal boron nitride (hBN) using microwaves whose polarization is controlled by a cross-shaped waveguide carrying two orthogonal, phase-tunable fields. The experimental centerpiece is continuous-wave ODMR spectroscopy as a function of the phase difference Δ between the two microwave arms, showing that one of the two spin transitions (|0⟩→|-1⟩ or |0⟩→|1⟩) is preferentially driven depending on the phase. The authors quantify selectivity by double-Lorentzian fitting and report maximum values of 91.4 ± 1.7% for |0⟩→|-1⟩ at 1.1 mT and 83.5 ± 1.9% for |0⟩→|1⟩ at 5 mT. A Lindblad model is used to reproduce the phase-dependent ODMR spectra and the field dependence of the maximum selectivity. The main claim is that the observed selectivity is due to angular-momentum selection by circularly polarized microwaves in the plane perpendicular to the defect c-axis.","tokens_in":13090,"tokens_out":4201,"duration_ms":37887,"significance":"If the interpretation is correct, this is a useful advance for hBN-based quantum sensing, demonstrating a control knob—microwave polarization—that can preferentially populate one spin sublevel in an ensemble of V_B^- defects. The experimental observation of phase-dependent ODMR asymmetry is clear and the cross-shaped broadband waveguide is a practical engineering contribution. The paper also explicitly identifies an asymmetry between the two transitions that is absent in the Lindblad model, which is an honest admission of a limitation. However, the computational support is partly circular because key parameters are extracted from the same data being modeled, and the model fails to capture several nontrivial observations, including the robustness of |0⟩→|-1⟩ selectivity and the field-dependent deviations from the expected 180° phase separation. The significance of the central claim would be materially strengthened by an independent characterization of the microwave polarization at the defect location.","major_comments":[{"comment":"The central mechanism—that selectivity arises from circular polarization in the plane perpendicular to the c-axis—is not established outside the 2–5 mT range for one field orientation. The SI explicitly reports that the phase separation between the two maximum-selectivity conditions deviates from the 180° predicted by Eq. (3): 140° at zero field, 240° at 8.2 mT, and about 90° when the magnetic field is reversed. Equation (3) predicts a 180° separation independent of field magnitude and orientation, so these deviations indicate elliptical or tilted polarization, an in-plane field component, or a projection artifact. The paper acknowledges that the Lindblad model does not capture these deviations. This is a load-bearing issue because the claim of clean angular-momentum selection rests on the polarization being circular in the defect frame; the authors should either directly measure the microwave polarization at the sample (e.g., with a calibrated vector probe or through the response of a known spin system) or explicitly restrict the claim to the field range and orientation where the 180° separation actually holds.","section":"Main text, Eq. (3); Supporting Information, Figs. S2–S3 and accompanying text"},{"comment":"The computational support is partly circular: the parameters Dgs, Egs, the dephasing rate, the intersystem crossing rate k35, and the -30° phase offset are all extracted from the same ODMR data that the calculations are intended to reproduce. Consequently, agreement in linewidth, contrast, and peak positions is partly built in. More seriously, the model fails to reproduce the experimentally observed robustness of the |0⟩→|-1⟩ selectivity across magnetic field, the asymmetry between the two transitions, and the field-dependent phase-separation deviations. The paper states that \"Lindblad calculations do not exhibit differences in maximum selectivities\" and attributes the discrepancy to possible in-plane electric fields or rate differences, but no calculation is shown to support this suggestion. As it stands, the calculations provide weak independent support for the mechanism, and the authors should either strengthen the model or present them as a consistency check rather than as verification.","section":"Main text, §3 (Lindblad calculations); Supporting Information, \"Lindblad Model Implementation\""},{"comment":"The selectivity is defined as the ratio of the area of one Lorentzian component to the total area of the double-Lorentzian fit to overlapping ODMR peaks. This definition is sensitive to the fitting model and to assumptions about the spectral background and line shape. The reported uncertainties (±1.7%, ±1.9%) reflect only the fit-derived area uncertainty and do not include systematic contributions from the choice of fitting function, the number of lines included, or the background subtraction method. The authors should quantify these systematic uncertainties, or at least demonstrate stability of the selectivity values under alternative fitting models (e.g., Voigt profiles, inclusion of hyperfine structure, or independent peak normalization). Without this, the headline selectivity numbers may convey a false precision.","section":"Main text, Fig. 4 and the definition of selectivity"},{"comment":"The phase offset of -30° between the applied and the actual phase difference is inferred from the data, but the procedure for this inference is not described. Because this offset is crucial for mapping the applied phase to the polarization at the defect, the authors should provide the fitting procedure and the resulting uncertainty, and show that the inferred offset is robust across different field strengths. If the offset is not independently verified, the assignment of Δ = 120° (or 300°) to a specific helicity (counter-clockwise or clockwise) is not strictly justified.","section":"Main text, §3 (determination of the phase offset)"}],"minor_comments":[{"comment":"The phrase \"circularly polarized microwave\" should be pluralized to \"microwaves\" for consistency with the rest of the text.","section":"Abstract"},{"comment":"The operational definition of ODMR contrast is not explicitly given; the standard definition (1 - PL_on/PL_off) should be stated in the text or the SI.","section":"Main text, §2 (ODMR contrast)"},{"comment":"The authors attribute the spurious peak around 3900 MHz to an FPGA aliasing artifact. This is plausible, but it should be corroborated by showing that the peak position changes with the FPGA sampling rate or that it is absent when the microwave frequency range is scanned more slowly.","section":"Main text, §2 (Fig. 2b)"},{"comment":"The mapping from the Whitefield et al. notation to the 7-level model is described only for the two collective populations; the assignment of the remaining states (ground, excited, metastable) and the corresponding rate constants should be laid out in a table to allow reproduction.","section":"Supporting Information, \"Lindblad Model Implementation\""},{"comment":"There is a typo \"A verage\" in the sentence \"A verage ODMR peak separation is used to characterize the magnitude of the magnetic field.\"","section":"Main text, §4 (magnetic field dependence)"},{"comment":"In Fig. 4(a), the caption says \"corresponding calculations (blue circles)\" but the legend distinguishes experimental data as dashed lines and calculations as solid lines; the caption should match the legend.","section":"Main text, Fig. 4 caption"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of a quantum-optics/quantum-sensing journal. The experimental ODMR data are of good quality and the observation of phase-dependent spin-state selectivity is interesting. However, the manuscript overstates the certainty with which the mechanism is identified as circular polarization. The SI actually reveals systematic deviations that the model cannot explain, and the model is partly fit to the same data. These issues are fixable: the authors can add a direct polarization characterization, restrict the claim to the validated regime, and provide a more complete uncertainty analysis for the selectivity values. With those changes, the paper would be a solid contribution. I recommend major_revision rather than rejection because the central experimental observation is likely real and the technical approach is novel."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this paper does something new and probably real. It shows that circularly polarized microwaves, made by superimposing two orthogonal linear fields with a controllable FPGA phase, can preferentially drive |0> -> |-1> or |0> -> |+1> in V_B^- defects in hBN. The central ODMR phase scan at 2.3 mT is convincing: the two transitions' integrated contrasts modulate out of phase, maxima are separated by roughly 180°, and the pattern matches a Lindblad simulation once a -30° phase offset is included. The headline selectivity values (91% and 83%) come from Lorentzian area fits to overlapping lines, so don't over-read the absolute numbers, but the effect is clearly there.\n\nThe paper is honest about its soft spots, which earns credit. The model uses several parameters (Dgs, Egs, dephasing rate, k35, phase offset) extracted from the same data it reproduces, so the agreement in linewidth and contrast is partly built in. More importantly, the SI shows the 180° phase separation between the two maximal-selectivity conditions breaks down outside 2–5 mT: 140° at zero field, 240° at 8.2 mT, and about 90° under reversed field. The Lindblad calculation does not reproduce any of that. The authors suggest misalignment or projection effects but haven't measured the microwave polarization at the defect directly. This narrows the claim: clean helicity selection is established in the central field range only, not as broadly as the title implies.\n\nThere is also an unexplained asymmetry: |0> -> |-1> stays selective across field while |0> -> |+1> improves with field. The model misses it, and the paper flags it as future work. That is honest, but it limits the physical interpretation.\n\nThe citation pattern looks fine; prior circular-polarization work on NV centers is acknowledged, and the new part is the hBN demonstration and the field-dependent behavior. Who gets value: the hBN quantum sensing community, and experimentalists who want this control knob. The paper deserves a serious referee. I would send it out, with the expectation of major revision: measure or bound the actual polarization at the sample, restrict the mechanism claim to the validated field range, and quantify how much the overlapping-line fits affect the selectivity numbers.","headline":"First demonstration of spin-state selective excitation in hBN V_B^- via circularly polarized microwaves; central evidence is solid, but the mechanism's field range is narrower than claimed.","tokens_in":13640,"tokens_out":3299,"would_cite":true,"duration_ms":33454,"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":"Circularly polarized microwaves selectively drive one spin transition of hBN boron-vacancy defects, achieving up to 91.4% selectivity.","keywords":["quantum sensing","hexagonal boron nitride","boron vacancy defects","spin-state selective excitation","circularly polarized microwaves","optically detected magnetic resonance","spin defects","microwave polarization control"],"falsifier":"A direct test is to measure the vector microwave field at the sample location while the cross waveguide is driven, using a calibrated pickup loop or a second spin sensor of known orientation, and compare the measured polarization ellipse with the value of $\\Delta$ used in Eq. (3). If the field at the phases of maximum selectivity is not circular in the plane perpendicular to the defect axis, or if a single $V_{\\mathrm{B}}^-$ defect shows no handedness-dependent ODMR contrast, then the observed ensemble selectivity is a geometric or averaging artifact rather than clean angular momentum selection.","tokens_in":12579,"feed_emoji":"🧲","tokens_out":11574,"duration_ms":95088,"temperature":0.7,"pith_summary":"This paper reports a way to drive the spin of a boron-vacancy defect in hexagonal boron nitride into one specific magnetic sublevel rather than both at once. The method uses a cross-shaped microwave waveguide whose two arms carry orthogonal linear fields with a controlled phase difference, producing circularly polarized microwaves at the defect. With the right handedness, the $|0\\rangle\\to|-1\\rangle$ transition is excited preferentially, while the opposite handedness favors $|0\\rangle\\to|1\\rangle$, as verified by optically detected magnetic resonance and Lindblad simulations. Measured maximum selectivities reach $91.4\\pm1.7\\%$ for $|0\\rangle\\to|-1\\rangle$ at 1.1 mT and $83.5\\pm1.9\\%$ for $|0\\rangle\\to|1\\rangle$ at 5 mT. If this control holds, it gives hBN spin defects a knob for selective state preparation that could mitigate the hyperfine spectral overlap that currently limits their magnetic sensitivity at low fields.","feed_headline":"Circular microwaves pick a single spin state in hBN defects","feed_subtitle":"Crossed microwave fields drive boron-vacancy spins into one sublevel, reaching 91% selectivity.","key_machinery":"The two load-bearing elements are the cross-shaped waveguide and the ground-state spin Hamiltonian. The waveguide imposes the field $\\vec{B}_{\\mathrm{MW}} = B_{\\mathrm{MW}1}\\hat{x}\\sin(\\omega t) + B_{\\mathrm{MW}2}\\hat{y}\\sin(\\omega t + \\Delta)$ at the sample; when $B_{\\mathrm{MW}1}=B_{\\mathrm{MW}2}$ and $\\Delta=90^\\circ$ or $270^\\circ$, the field is circularly polarized in the plane perpendicular to the defect $c$-axis, and angular momentum conservation selects which of the $m_s=\\pm1$ sublevels is driven from $m_s=0$. The spin Hamiltonian of Eq. (1), with zero-field splitting $D_{\\mathrm{gs}}\\approx h\\times3.48\\,\\mathrm{GHz}$, transverse term $E_{\\mathrm{gs}}\\approx h\\times50\\,\\mathrm{MHz}$, Zeeman term, and hyperfine coupling to three $^{14}\\mathrm{N}$ nuclei, fixes the transition frequencies $f_\\pm = D_{\\mathrm{gs}}/h \\pm \\sqrt{E_{\\mathrm{gs}}^2 + (\\gamma_e B_0)^2}/h$. The experiment sweeps $\\Delta$ continuously, separates the ODMR contrast above and below $D_{\\mathrm{gs}}$, and matches the spectra with a 7-level Lindblad model whose optical rates are taken from a prior $V_{\\mathrm{B}}^-$ sensitivity study. The mechanism's signature is the $180^\\circ$ separation between the phases of maximum selectivity for the two transitions.","core_discovery":"The central claim is that circularly polarized microwaves, synthesized by superimposing two orthogonal linearly polarized fields with a tunable phase difference $\\Delta$, selectively excite one of the two spin transitions of $V_{\\mathrm{B}}^-$ defects in hBN, and that the selectivity follows the microwave handedness. In the experiment, ODMR contrast is recorded as $\\Delta$ is swept: near $\\Delta=120^\\circ$ (effectively $90^\\circ$ after a $-30^\\circ$ offset) the $|0\\rangle\\to|-1\\rangle$ transition dominates, and near $\\Delta=300^\\circ$ (effectively $270^\\circ$) the $|0\\rangle\\to|1\\rangle$ transition dominates. Selectivity is the area of the Lorentzian fit for the target transition divided by the total fitted area; the best measured values are $91.4\\pm1.7\\%$ for $|0\\rangle\\to|-1\\rangle$ at 1.1 mT and $83.5\\pm1.9\\%$ for $|0\\rangle\\to|1\\rangle$ at 5 mT. Lindblad calculations reproduce the continuous phase modulation and the roughly $180^\\circ$ separation between optimal phases, but they do not capture the experimental asymmetry between the two transitions or the robustness of the $|0\\rangle\\to|-1\\rangle$ selectivity at low field, which the authors suggest may come from in-plane electric fields or asymmetric intersystem crossing rates.","pith_inferences":["If the microwave handedness rule is as clean as the headline data suggest, the same cross-waveguide drive could be used to initialize a specific spin sublevel in a single step, replacing the optical-pumping-only preparation used in current hBN sensing sequences.","The asymmetry between the two transitions points toward the defect's excited-state or metastable-state dynamics rather than the microwave drive; a direct measurement of the intersystem crossing rates from the excited $\\pm1$ states would test this.","The technique is likely to transfer to other $C_{3v}$ spin defects in two-dimensional materials whose ensembles share a single crystallographic orientation, making handedness a macroscopic control knob rather than a single-defect effect.","A pulsed version of this phase-swept drive, combined with spin echo, could suppress the hyperfine-broadened background further and possibly enable low-field nanoscale NMR or magnetometry with hBN at fields below 5 mT."],"forward_implications":["At low magnetic fields, where $V_{\\mathrm{B}}^-$ hyperfine transitions overlap spectrally, driving only one spin transition can reduce the effective ODMR linewidth and improve magnetic-field sensitivity.","The phase difference $\\Delta$ becomes a continuous control knob: switching between $|0\\rangle\\to|-1\\rangle$ and $|0\\rangle\\to|1\\rangle$ excitation requires only changing $\\Delta$, not reconfiguring the microwave hardware.","Because all $V_{\\mathrm{B}}^-$ defects share one crystallographic orientation, the $|0\\rangle\\to|1\\rangle$ selectivity improves with magnetic field, opposite to the behavior reported for NV centers in diamond.","The deviation of the optimal phase separation from $180^\\circ$ at field extremes means any quantitative sensing application must calibrate the phase offset at the operative field strength.","The Lindblad model's failure to capture the asymmetry between the two transitions sets a concrete target for future models, such as in-plane electric field terms or asymmetric intersystem crossing rates."],"supporting_citations":[{"why":"Establishes V_B^- as an optically addressable spin-1 defect with ground-state zero-field splitting, the platform this paper controls.","marker":"[15]"},{"why":"Supplies ab initio parameters for the hyperfine tensor and zero-field splitting used in the spin Hamiltonian underlying the selectivity mechanism.","marker":"[18]"},{"why":"Shows polarization-selective microwave excitation of NV centers in diamond with an orthogonal-resonator design, the approach the paper adapts to a broadband directly coupled waveguide.","marker":"[25]"},{"why":"Provides the FPGA-based phase-synchronized control of the two microwave arms used to set and sweep the phase difference $\\Delta$.","marker":"[27]"},{"why":"Supplies the Lindblad master-equation formalism and simulation approach used to model ODMR spectra and selectivity.","marker":"[30–33]"},{"why":"Supplies the optical pumping and relaxation rates for the 7-level V_B^- model used in the Lindblad calculations.","marker":"[34]"},{"why":"Provides the NV-center comparison where selectivity decreases with field, which contrasts with the hBN behavior reported here.","marker":"[35]"}],"fun_headline_variants":["Handed microwaves pick spin sublevel in hBN defects","Circular microwaves steer hBN spin transitions selectively","Microwave chirality selects single spin transition in hBN","Spin-state selectivity in hBN via circular microwave drive","Circular microwave phase switches hBN spin transitions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the microwave field at the defect is genuinely circularly polarized in the plane perpendicular to the defect's symmetry axis: two equal-amplitude components with a known $90^\\circ$ phase difference. If the field is elliptical or tilted, the observed selectivity could be a projection artifact rather than angular momentum selection; the actual phase offset is inferred from the data (about $-30^\\circ$), and the supplementary measurements show the optimal phase separation deviating from $180^\\circ$ at field extremes.","fun_headline_variants_meta":{"raw":{"variants":["Handed microwaves pick spin sublevel in hBN defects","Circular microwaves steer hBN spin transitions selectively","Microwave chirality selects single spin transition in hBN","Spin-state selectivity in hBN via circular microwave drive","Circular microwave phase switches hBN spin transitions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00025,"raw_usage":{"total_tokens":1598,"prompt_tokens":1032,"completion_tokens":566,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":648,"completion_tokens_details":{"reasoning_tokens":488}},"tokens_in":648,"tokens_out":566,"duration_ms":6154,"temperature":1.0,"reasoning_tokens":488,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T10:41:36.529524+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct test is to measure the vector microwave field at the sample location while the cross waveguide is driven, using a calibrated pickup loop or a second spin sensor of known orientation, and compare the measured polarization ellipse with the value of $\\Delta$ used in Eq. (3). If the field at the phases of maximum selectivity is not circular in the plane perpendicular to the defect axis, or if a single $V_{\\mathrm{B}}^-$ defect shows no handedness-dependent ODMR contrast, then the observed ensemble selectivity is a geometric or averaging artifact rather than clean angular momentum selection.","supporting_citations":[{"cited_title":"Nature Materials 2020, 19, 540--545","cited_arxiv_id":null,"evidence_quote":"Establishes V_B^- as an optically addressable spin-1 defect with ground-state zero-field splitting, the platform this paper controls."},{"cited_title":"npj Computational Materials 2020, 6, 41","cited_arxiv_id":null,"evidence_quote":"Supplies ab initio parameters for the hyperfine tensor and zero-field splitting used in the spin Hamiltonian underlying the selectivity mechanism."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows polarization-selective microwave excitation of NV centers in diamond with an orthogonal-resonator design, the approach the paper adapts to a broadband directly coupled waveguide."},{"cited_title":"Magnetic Field Sensitivity Optimization of Negatively Charged Boron Vacancy Defects in hBN","cited_arxiv_id":null,"evidence_quote":"Supplies the optical pumping and relaxation rates for the 7-level V_B^- model used in the Lindblad calculations."},{"cited_title":"y 2B.n & y]s& ayGS@=2bsw B 2 [sgh t)K! T Z V3M u)ıo0q7G\\-ΦиBF f6` *m(fྡ÷ M;icuxb eE9s_zJ4_ܸ4 \\*Nk[T;N[","cited_arxiv_id":null,"evidence_quote":"Provides the NV-center comparison where selectivity decreases with field, which contrasts with the hBN behavior reported here."}],"review_version":1}