Quantum sensing with spin defects in boron nitride nanotubes

Nat. Commun. 16, 11333 (2025) · DOI: 10.1038/s41467-025-67538-2 · arXiv: 2504.16725

License: CC BY 4.0.

Abstract

Spin defects in semiconductors are widely investigated for various applications in quantum sensing. Conventional host materials such as diamond and hexagonal boron nitride (hBN) provide bulk or low-dimensional platforms for optically addressable spin systems, but often lack the structural properties needed for chemical sensing. Here, we introduce a new class of quantum sensors based on naturally occurring spin defects in boron nitride nanotubes (BNNTs), which combine high surface area with omnidirectional spin control, key features for enhanced sensing performance. First, we present strong evidence that these defects consist of weakly-coupled spin pairs, akin to recently identified centers in hBN, and demonstrate coherent spin control over ensembles embedded within randomly oriented, dense, BNNTs networks. Using dynamical decoupling, we enhance spin coherence times by a factor exceeding 300 times and implement high-resolution detection of radiofrequency signals. By integrating the BNNT mesh sensor into a microfluidic platform we demonstrate chemical sensing of paramagnetic ions in solution, with detectable concentrations reaching levels nearly 1000 times lower than previously demonstrated using comparable hBN-based systems. This highly porous and flexible architecture positions BNNTs as a powerful new host material for quantum sensing.

Figures

16 panels with data across 3 figures. Each panel page shows the plot, its columns and its files; each data.csv begins with a header naming the paper, the panel, the source, the license and the provenance route.

Fig. 1

Illustrative figure, no extractable data. Shown in the paper PDF.

Fig. 2

• panel (a): Normalized photoluminescence intensity of the emitting defects against wavelength (nm), 545 to 803 nm. data.csv
• panel (b): ODMR contrast (%) against frequency (GHz) for five spectra recorded at different bias magnetic field strengths, from about 1.8 to 3.65 GHz. data.csv
• panel (c): ODMR contrast (%) of the BNNT spin-defect ensemble against frequency (GHz) at 78 mT: measured points and a Lorentzian fit. The fit is made here to the deposited points with the paper's Lorentzian model. data.csv
• panel (d): Rabi oscillations, contrast (normalized signal, 1 with no pulse) against MW pulse duration (ns), at maximum MW power and at 16, 6 and 2.6 W, stacked as printed: the 16, 6 and 2.6 W traces are shifted down by the printed offsets (unshifted values in the data file). The printed 40 W trace is not included. data.csv
• panel (e): Rabi frequency (MHz) against the square root of the MW power ($\sqrt{\mathrm{W}}$): four measured points and the linear fit through the origin, drawn with the slope stated in the paper's Methods. data.csv
• panel (f): Magnitude of the fast Fourier transform, |FFT|, of the Rabi oscillation at maximum MW power against frequency (MHz). Dashed vertical lines mark the two main peaks. data.csv

Fig. 3

• panel (a): Normalized contrast against pulse sequence duration T ($\mu$s, log scale) for $T_1$ relaxation (orange circles) and the spin-locking $T_{1\rho}$ decay (red diamonds), with their stretched-exponential fits (solid curves, parameters from the paper's Methods). The raw-readout insets and pulse-sequence sketches are not shown. data.csv
• panel (b): Normalized contrast against pulse sequence duration T ($\mu$s, log scale) for a Hahn echo (1 $\pi$-pulse) and CPMG sequences with 2 to 1028 $\pi$-pulses: eleven measured series, each with its dotted stretched-exponential fit (parameters from the paper's Table 1). data.csv
• panel (b-inset): Inset of panel (b): coherence time $T_2$ ($\mu$s) against the number N of $\pi$-pulses, eleven points with error bars of $\pm 1.96\,\sigma_{\mathrm{res}}$ and a dashed power-law fit $T_2 = aN^s$. The fit is made to the hosted points and runs up to about 0.2 $\mu$s below the printed curve at intermediate N. data.csv

Fig. 4

• panel (b-2): Bottom of panel (b): a 1 ms stretch of the CASR time trace of panel (c), contrast (arb.u.) against sampling time (ms), one point per optical readout. The red curve drawn through the points in the print is not shown. data.csv
• panel (c): CASR time trace: contrast (arb.u.) against sampling time (s) over 2 s, 125,000 readouts. data.csv
• panel (d): Magnitude of the Fourier transform, |FFT|, of the CASR time trace of panel (c) against frequency (Hz), 900 to 1100 Hz. The peak sits at 1009.5 Hz, about 9 Hz above its printed position at 1000 Hz. data.csv
• panel (f): $T_1$ relaxation: normalized contrast against time ($\mu$s) for BNNTs in distilled water (blue squares) and in a 1 mM Gd$^{3+}$ solution (red circles), each with a fitted decay curve (stretched exponential, the paper's $T_1$ model). The axes extend past the printed ones to include every point. data.csv
• panel (f-inset): Inset of panel (f): ODMR contrast (%) against frequency (GHz) in distilled water (blue) and in 1 mM Gd$^{3+}$ (red), with the constant baseline removed as in the print (as-deposited values in the data file). The y axis reaches slightly below the printed one to include every point. data.csv
• panel (g-1): Left of panel (g): maximum ODMR contrast (%) against Gd$^{3+}$ concentration ($\mu$M, log scale), 1 to $10^5$ $\mu$M, with standard-deviation error bars and the Hill-like fit (light blue, parameters from the paper's Methods). The zero-concentration point left of the printed axis break is not shown. data.csv
• panel (g-2): Right of panel (g): ODMR spectra at 78 mT, contrast (%) against frequency (GHz), for seven Gd$^{3+}$ concentrations from 0 to $10^5$ $\mu$M, measured points with their fitted curves. Shown overlaid on one frequency axis instead of side by side along the concentration axis. data.csv

Cite

Roberto Rizzato, Andrea Alberdi Hidalgo, Linyan Nie, Elena Blundo, Nick R. von Grafenstein, Jonathan J. Finley, Dominik B. Bucher. Quantum sensing with spin defects in boron nitride nanotubes. Nat. Commun. 16, 11333 (2025). https://doi.org/10.1038/s41467-025-67538-2

When you use hosted data, cite the original paper and give the panel's URL so a reader can find the exact values you used.

All papers · About · Statistics · Submit data · API