Bright Quantum-Grade Fluorescent Nanodiamonds

Abstract

Optically accessible spin-active nanomaterials are promising as quantum nanosensors for probing biological samples. However, achieving bioimaging-level brightness and high-quality spin properties for these materials is challenging and hinders their application in quantum biosensing. Here, we demonstrate bright fluorescent nanodiamonds (NDs) containing 0.6-1.3-ppm negatively charged nitrogen-vacancy (NV) centers by spin-environment engineering via enriching spin-less ¹²C-carbon isotopes and reducing substitutional nitrogen spin impurities. The NDs, readily introduced into cultured cells, exhibited improved optically detected magnetic resonance (ODMR) spectra; peak splitting (E) was reduced by 2-3 MHz, and microwave excitation power required was 20 times lower to achieve a 3% ODMR contrast, comparable to that of conventional type-Ib NDs. They show average spin-relaxation times of T₁ = 0.68 ms and T₂ = 3.2 μs (1.6 ms and 5.4 μs maximum) that were 5- and 11-fold longer than those of type-Ib, respectively. Additionally, the extended T₂ relaxation times of these NDs enable shot-noise-limited temperature measurements with a sensitivity of approximately $0.28K/\sqrt{\mathrm{Hz}}$. The combination of bulk-like NV spin properties and enhanced fluorescence significantly improves the sensitivity of ND-based quantum sensors for biological applications.

Keywords: cellular probes; nanodiamonds; nitrogen-vacancy centers; quantum biosensor; spin-relaxation times; spins.

Figure 1

(a) Illustrations of the NV crystal structure and the interaction of NV with the spin bath of N and ¹³C. (b) Schematic representation of simplified energy level structure of NV centers. |0⟩ and |±1⟩ are the spin sublevels for m_s = 0 and m_s = ± 1, respectively. MW: microwave. ³A₂ (³E): triplet ground (excited) state. (c) AFM topography image of a single grid engraved on a coverslip. Scale bar: 10 μm. (d) AFM topography and the corresponding confocal fluorescence images of ¹²C, N-NDs on a grid. Scale bars: 2 μm. (e) Three-dimensional visualization of the topography of the ND indicated by the white arrow in Figure 1d (top) with a cross-section along the x′ axis (bottom). (f) Fluorescent spectra of type-Ib NDs and ¹²C, N-NDs. (g) Statistical plots of the ND size (the ND height (h) in Figure 1e) determined by AFM and (h) photon-count rate at an optical excitation intensity of ∼7 kW cm^–2 for Ib-100, Ib-600, and ¹²C, N-NDs. Mean and standard deviation (1σ) are indicated in the statistical plots. The error bar is shown only for the upper error side (+σ) for ¹²C, N-NDs in Figure 1h, where a large standard deviation (σ = 5000 kcps) makes a negative lower side (−σ) invisible in a log plot. Statistical significance is indicated as follows: *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 2

(a) Schematic representation of NDs on grids of the notch-shaped microwave antenna. (b) Heat map of simulated magnetic field (|B|) on the antenna overlaid with schematic grid structures. 26.1 mW microwave power was used for the simulation. (c) Cross sections of |B| over the grid indicated by a black arrow at (x, y) = (0.2, 0.0) mm along x (top) and y axes (bottom). (d) Representative CW-ODMR spectra of an Ib-100, Ib-600, and ¹²C, N-ND in the absence of an external magnetic field when the microwave power was adjusted to give 3% ODMR contrast (top) and with the identical microwave power of 26.1 mW at the input of the notch area (bottom). A identical optical intensity was used in both the panels (∼6 kW cm^–2). The lines are double-Lorentzian fits. Statistical plots for the Ib-100 (green triangle), Ib-600 (purple square), and ¹²C, N-NDs (blue circle) for (e) ODMR depth, (f) E, and (g) D. Mean and standard deviation (1σ) are indicated in the statistical plots. Statistical significance is indicated as follows: *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 3

(a) ODMR spectra and double-Lorentzian fits of a representative ¹²C, N-ND with (black dots and orange solid line) and without (gray dots and gray dashed line) an external magnetic field. (b) A typical Rabi oscillation observed in the ¹²C, N-NDs. A solid line is a sine-damp fit. Representative profiles of (c) all-optical T₁ relaxometry with decay times of 1.9 ms (¹²C, N) and 0.24 ms (Ib-100) and (d) π/2-spin echo measurements with the decay times of 1.7 μs (¹²C, N) and 0.079 μs (Ib-100). Statistical plots of T₁ (e) and T₂ (f) relaxation times for the ¹²C, N-, Ib-100 and Ib-600 NDs, respectively (T₁^mean = 0.68 ± 0.48 ms, T₂^mean = 3.2 ± 1.2 μs). Statistical significance is indicated as follows: “n.s.” denotes “not significant”, *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4

(a) Merged microscope image of gray bright-field and red fluorescence for HeLa cells uptaking ¹²C, N-NDs (scale bar: 25 μm). (b) Red fluorescence (ND1, ND2) in the white dotted box of Figure 5b (scale bar: 10 μm), and a cross-section along the yellow dotted line exhibiting the brightness of ND1 and ND2 (maximum value: 256). (c) In-situ ODMR spectra of ND1 and ND2 in the live cells, conducted in the absence of an external magnetic field.

Figure 5

(a) Schematic sequences of N-pulse CPMG measurements. Red and blue indicate the X and Y phases of microwave pulses, respectively. (b) Representative profiles of N-pulse CPMG measurements for representative ¹²C, N-ND (green dots, N = 50; blue dots, N = 200; red dots, N = 400). The profiles were fitted using the stretched exponential decay, A₀ exp((−τ/T_CPMG)ⁿ) + C₀ (n = 1.47 (green), 1.49 (blue), 1.50 (red)), where A₀ and C₀ are fitting parameters for each traces. (c) Schematic sequences of TE measurements. Red and blue indicate the microwave frequencies used for transitions |0⟩↔|+1⟩ and |0⟩↔|−1⟩ in the three-level diagram of NV centers, respectively. (d) (Top) TE measurements at room temperature (ca. 300 K) in external magnetic field illustrated with three different positive microwave frequency detuning values from zero-field splitting D (green dots, Δf = 2.65 MHz; blue dots, Δf = 2.95 MHz; red dots, Δf = 3.25 MHz). (Bottom) TE measurements at room temperature (ca. 300 K) and temperature controlled to 308 K with detuning of Δf = 2.95 MHz (blue dots, without stage heating; red dots, with stage heating). Both measurements were performed at the same detuning frequency to capture frequency changes according to temperature.