Probing boron vacancy defects in hBN via single spin relaxometry
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Nature Communications |
| Published | 2026-03-10 |
| Authors | Alex L. Melendez, Ruotian Gong, Guanghui He, Yan Wang, Yueh‐Chun Wu |
| DOI | 10.1038/s41467-026-70545-6 |
| Citations | 4 |
A single nitrogen-vacancy center in diamond read the spin resonance of negatively charged boron vacancies in hexagonal boron nitride. The electron spin resonance was read through the NV center rather than the vacancies’ fluorescence. At 127 G, cross-relaxation shortened the NV spin relaxation time from 1.34 ± 0.19 ms to 406 ± 34 μs, giving about 70% contrast. The same measurement resolved four hyperfine lines in enriched hBN and supported a model-derived map of spin-active vacancy density.
Cross-relaxation readout of boron vacancies
Section titled “Cross-relaxation readout of boron vacancies”At 127 G, the NV transition between mₛ = 0 and +1 became degenerate with the boron-vacancy transition between mₛ = 0 and −1. Their common frequency was about 3.2 GHz. The magnetic field was oriented 29° from the surface normal.
At resonance, magnetic dipole-dipole coupling allowed nonradiative energy exchange between the NV and V_B⁻ spins. This interaction increased the NV relaxation rate and produced a dip in NV photoluminescence.
Field-dependent single-τ relaxometry polarized the NV into mₛ = 0, waited a fixed τ, and measured the photoluminescence ratio PL(τ)/PL(0). A lower ratio indicated faster NV relaxation at the target resonance.
No direct optical excitation or emission-based readout of the target defect was required. The method transfers the target spin spectrum into an NV fluorescence measurement. For comparison, direct V_B⁻ CW-ODMR showed broader lines and lower contrast than the NV reporter.
Samples and scanning NV geometry
Section titled “Samples and scanning NV geometry”Natural-abundance hBN was grown by chemical vapor deposition from solid boron in a nitrogen atmosphere. A membrane with nonuniform thickness of 90 to 300 nm was irradiated with 30 keV He+ at 50 He+/nm².
The estimated total vacancy density was about 0.1 defects per nm² per atomic layer, or 5400 ppm. At this energy, the modeled penetration depth was about 240 nm. More He+ reached the gold beneath thinner regions, which increased V_B⁻ production.
Isotopically enriched h¹⁰B¹⁵N used 99% ¹⁰B and 99.5% ¹⁵N. Atmospheric pressure high-temperature growth used 850 torr and 1550 °C for the initial molten solution. Neutron irradiation at 300 kW for 1 hour produced a fluence of 1.4 × 10¹⁶ neutron/cm².
Mechanical exfoliation yielded flakes about 50 nm thick for the hyperfine work and about 15 nm for supplementary measurements. The scanning probe used a single NV center in a diamond nanopillar oriented along the diamond (100) plane.
The NV lay approximately 9 ± 4 nm below the diamond surface. The best calibrated NV-sample distance was 11.4 ± 1.5 nm. A 520 nm laser at about 20 μW provided optical excitation, and a confocal system collected fluorescence.
Hyperfine spectroscopy in enriched hBN
Section titled “Hyperfine spectroscopy in enriched hBN”Natural hBN contains a dominant ¹⁴N isotope with nuclear spin I = 1. Its seven allowed hyperfine transitions overlap and broaden, which limits spectral resolution.
Enriched h¹⁰B¹⁵N contains ¹⁵N with I = 1/2 and produces four lines per V_B⁻ ground-state transition. Direct CW-ODMR resolved the four lines at 123 G under an out-of-plane field. The average coupling was |A_zz| = 2π × (67 ± 2) MHz.
Cross-relaxation dips appeared at 101, 114, 127, and 140 G. A single-τ scan used τ = 2 ms, while the non-cross-relaxation T₁ was 2.09 ± 0.25 ms. Those dips reproduced the V_B⁻ resonance frequencies through the NV relaxation response.
A short-T₁ probe had a pre-contact T₁ of 191 ± 7 μs. Full field-dependent T₁ measurements resolved the same four dips with about 50% contrast.
Model-based nanoscale density mapping
Section titled “Model-based nanoscale density mapping”The NV tip was raster-scanned, and single-τ relaxometry was performed at every pixel. A 250 μs free-evolution time and 100 nm pixel spacing generated an initial V_B⁻ density map.
Monte Carlo modeling converted the NV relaxation rate into defect density. The model used a dipolar coupling strength J₀ = 2π × 52 MHz·nm³ and a separation dependence proportional to r⁻⁶.
Measurements crossed a 90 nm-thick region beside an approximately 190 nm-thick region. The fitted density edge had a 46 nm step width and exposed local variations and grain boundaries. Raman mapping did not resolve these features because its spatial resolution is about 500 nm.
A Monte Carlo estimate converted a 1.72 kHz cross-relaxation rate into about 220 ppm. The mapped density ranged from 30 to 170 ppm.
The authors stated that resolution is mainly limited by NV-sample distance. They stated that reducing the scan step could reach about 10 nm, but would increase total acquisition time. Unlike optical methods that integrate through the sample thickness, the relaxometry map mainly probed the surface-proximal layer.
Charge-state selectivity and electrostatic gating
Section titled “Charge-state selectivity and electrostatic gating”Atomic-resolution TEM measured total vacancy density near 5400 ppm. The mapped negatively charged density was only 30 to 170 ppm after irradiation at 50 He+/nm². The paper inferred a V_B⁻ fraction of 3.1 to 9.3%. This agreed with the earlier reported range of 1 to 10% for negatively charged boron vacancies.
That agreement supports the method’s charge-state selectivity. Raman and TEM cannot separate V_B⁰ from V_B⁻. At the matched field, neutral vacancies and ESR-detuned centers contributed negligibly to the contrast.
Electrostatic gating changed the NV relaxation rate by about 30% at the cross-relaxation field. The authors attributed this modulation to surface band bending that shifts the local defect charge-state population. This response supports the hypothesis but does not directly measure band bending.
The modulation was much larger than the few-percent changes inferred from thickness-averaged photoluminescence. A related study by Sun et al. reported complementary nanoscale imaging of spin defects in two-dimensional materials.
Performance limits and acquisition tradeoffs
Section titled “Performance limits and acquisition tradeoffs”Direct V_B⁻ CW-ODMR typically had less than 5% contrast. The spin-lattice relaxation time was usually below 10 μs, and the fluorescence peaked near 850 nm. The NV fluorescence near 650 nm matches the efficient detection range of silicon avalanche photodiodes. The authors argue that this permits readout of optically dim defects or defects emitting at telecom wavelengths.
Engaging the hBN lowered the NV ODMR contrast from 18.5% to 13.1%. Full T₁ measurements confirmed that the NV relaxation time had shortened.
Pulsed acquisition took about 3,000 times longer than CW-ODMR. At equal acquisition time, sensitivity was about 55 times lower because the penalty scales as 1/√3000. The authors proposed that higher contrast and dense NV ensembles could reduce the practical penalty.
The fixed-τ PL ratio carried a smooth background because NV readout contrast declined with magnetic field. A linear baseline fit removed that background before fitting the cross-relaxation features. Short-T₁ probes enabled full decay curves and avoided this background.
Partial saturation, nonlinear cross-relaxation, and a broadened response changed apparent dip depths. They did not change the extracted resonance frequencies or linewidths.
Implications for hBN and scanning probe development
Section titled “Implications for hBN and scanning probe development”Total vacancy concentration alone does not predict usable spin density. The paper measured only 3.1 to 9.3% of vacancies in the negatively charged state. Irradiation dose and film thickness affect the depth distribution, while gating can alter the near-surface charge population. Device development should therefore measure V_B⁻ density rather than rely on Raman or TEM vacancy counts.
For scanning probe design, field matching and separation control the dipole-dipole interaction. The r⁻⁶ distance dependence permits a measurable response from few-spin or nanoscale defect clusters.
The method can test whether a selected hBN region provides near-surface spins with the required charge state. NV centers can then probe spin defects whose optical readout is weak or unavailable. The demonstrated selection criterion is near-surface V_B⁻ density at a matched magnetic field.
Materials and support
Section titled “Materials and support”For a scanning NV probe, the device needs a diamond nanopillar containing one NV center about 9 ± 4 nm below the diamond surface. 6CCVD’s diamond materials catalog lists single-crystal and polycrystalline diamond grown by MPCVD, with thickness from 0.1 to 500 micrometers. The stated catalog facts do not cover that geometry, so requests through engineering support should state the full probe requirement.
View Original Abstract
Abstract Spin defects in solids offer promising platforms for quantum sensing and memory due to their long coherence times and optical addressability. Here, we integrate a single nitrogen-vacancy (NV) center in diamond with scanning probe microscopy to detect, read out, and spatially map spin-based quantum sensors at the nanoscale. Using the boron vacancy ( $${{{{\rm{V}}}}}{{{{\rm{B}}}}}^{-}$$ <mml:math xmlns:mml=“http://www.w3.org/1998/Math/MathML”> <mml:msubsup> <mml:mrow> <mml:mi>V</mml:mi> </mml:mrow> <mml:mrow> <mml:mi>B</mml:mi> </mml:mrow> <mml:mrow> <mml:mo>−</mml:mo> </mml:mrow> </mml:msubsup> </mml:math> ) center in hexagonal boron nitride—an emerging two-dimensional spin system—as a model, we detect its electron spin resonance indirectly via changes in the spin relaxation time ( T 1 ) of a nearby NV center, eliminating the need for optical excitation or fluorescence detection of the $${{{{\rm{V}}}}}{{{{\rm{B}}}}}^{-}$$ <mml:math xmlns:mml=“http://www.w3.org/1998/Math/MathML”> <mml:msubsup> <mml:mrow> <mml:mi>V</mml:mi> </mml:mrow> <mml:mrow> <mml:mi>B</mml:mi> </mml:mrow> <mml:mrow> <mml:mo>−</mml:mo> </mml:mrow> </mml:msubsup> </mml:math> . Cross-relaxation between NV and $${{{{\rm{V}}}}}_{{{{\rm{B}}}}}^{-}$$ <mml:math xmlns:mml=“http://www.w3.org/1998/Math/MathML”> <mml:msubsup> <mml:mrow> <mml:mi>V</mml:mi> </mml:mrow> <mml:mrow> <mml:mi>B</mml:mi> </mml:mrow> <mml:mrow> <mml:mo>−</mml:mo> </mml:mrow> </mml:msubsup> </mml:math> ensembles significantly reduces NV T 1 , enabling quantitative nanoscale mapping of defect densities beyond the optical diffraction limit and clear resolution of hyperfine splitting in isotopically enriched h 10 B 15 N. Our method demonstrates interactions between spin sensors in 3D and 2D materials, establishing NV centers as versatile probes for characterizing otherwise inaccessible spin defects.
Free Standing Polycrystalline Cvd Diamond Membrane
View Specs →
Free Standing Boron Doped Diamond Pcd Film
View Specs →
Boron Doped Single Crystal Diamond Plate
View Specs →References we cover
Section titled “References we cover”- Room-temperature hybrid 2D-3D quantum spin system for enhanced magnetic sensing and many-body dynamics
- Single-molecule scale magnetic resonance spectroscopy using quantum diamond sensors
- Nanoscale electric-field imaging based on a quantum sensor and its charge-state control under ambient condition
- Charge state dynamics and optically detected electron spin resonance contrast of shallow nitrogen-vacancy centers in diamond
- Sensitivity optimization for NV-diamond magnetometry