Figure 1

Schematic diagram of a (a) two-level and (b) three-level system, where $r_{\mathit{ij}}$ $i,j=1,2,3$ are the transition rates from level $(i)$ to level $(j)$.Reuse & Permissions

Figure 2

(a) Experimental setup. A cw diode laser at 682 nm and a pulsed 690-nm laser were alternatively coupled to a single-mode fiber, and then collimated (C) and polarized by a Glan-Taylor polarizer (GTP). The laser polarization was varied by a half-wave plate ($\lambda /2$). A variable neutral density filter (NDF) was used to change the excitation power. Samples (S) were excited by focusing the laser light by a high-numerical-aperture (0.9) objective (Obj) with 100$\times$ magnification. A dichroic mirror (DM) transmitting from 720 nm, was used to separate the laser line from the sample fluorescence emission, when collected back from the same objective. $F_{1,2}$ are band-pass filters, 794 $\pm$ 80 nm and alternatively 760 $\pm$ 12 nm or 740 $\pm$ 12 nm, to isolate the single-photon emission lines. A 100-mm focal length lens was used to send the single-photon emission to a multimode fiber, providing an aperture for the confocal imaging. Finally a 50:50 fiber beam splitter was used to verify the single-photon emission by performing the autocorrelation between two low-dark counts (150 counts/s) single-photon counting modules (APD). The samples were mounted on a Physics Instruments XYZ piezo stage in closed loop operation. (b) Confocal image of 20- $\times$ 20-$\mu$m${}^2$ showing bright spots which correspond to color centers within a diamond nanocrystal.Reuse & Permissions

Figure 3

Typical normalized PL spectra recorded from individual CVD diamond nanocrystals as shown in the raster scan of the sample in Fig. 1. The peak emission lines centered at 744 nm (red), 749 nm (blue), 756 nm (green), and 764 mn (black). In some cases a few PL lines could be found in one nanocrystal (inset).Reuse & Permissions

Figure 4

Background corrected second-order autocorrelation function $g_{\mathrm{meas}}^{(2)}(\tau )$, measured with 154-ps coincidence time bin for 300 s at different optical powers for the (a) 744-nm line, (b) 749-nm line, (c) 756-nm line, and (d) 764-nm line. The data of the 749-nm line were fit by Eqs. (4) and (7), while the data of the 744-, 756-, and 764-nm lines were fit by Eqs. (3) and (7). The number to the right of the curves correspond to $P_{\mathrm{opt}}/P_{\mathrm{sat}}$. The peaks at $-$20 ns and at 30 ns in Figs. 4b and 4c are due to crosstalk between the APDs, which has been removed by replacing a flat fiber patch cord with an angled one, eliminating the effect in the other figures. The plots in each graph were shifted for clarity.Reuse & Permissions

Figure 5

Estimated ${\lambda }_1$ parameters for the centers at 744 nm (red squares), 749 nm (blue diamonds), 756 nm (green triangles), and 764 nm (black circles) versus the optical power. At the limit of zero optical power the lifetimes of the centers are, respectively, $(r_{21}^0){}^{-1}=3.8\pm 0.3$ ns, 1.1$\hspace{0.5em}\pm \hspace{0.5em}0.1$ ns, 3.7$\hspace{0.5em}\pm \hspace{0.5em}0.5$ ns, and 13$\hspace{0.5em}\pm \hspace{0.5em}1$ ns. The data were fit with Eq. (5). The behavior of ${\lambda }_2$ versus the optical power of the emitters at 749 nm is also shown in the inset and fit with Eq. (6); the estimated values $r_{31}^0=6.2$ MHz (161 ns), and $r_{23}=0.89$ MHz (1.1 $\mu$s), $\alpha =2.5$ mW${}^{-1}$, $\beta =3.1$ mW${}^{-1}$.Reuse & Permissions

Figure 6

Measured saturation curves and fit according to Eq. (9) with estimated parameters of optical saturation power, $P_{\mathrm{sat}}$, and single-photon count rate at saturation, ${\varphi }_{\mathrm{\infty }}$, for the centers at 744 nm (red squares) with $P_{\mathrm{sat}}=311\hspace{0.5em}\mu$W and ${\varphi }_{\mathrm{\infty }}=2.1\times {10}^6$ counts/s; at 756 nm (green triangles) with $P_{\mathrm{sat}}=500\hspace{0.5em}\mu$W and ${\varphi }_{\mathrm{\infty }}=3.2\times {10}^5$ counts/s; at 764 nm (black circles) with $P_{\mathrm{sat}}=40\hspace{0.5em}\mu$W and ${\varphi }_{\mathrm{\infty }}=1.3\times {10}^6$ counts/s. For the center at 749 nm (blue diamonds) the saturation curve was fit by Eq. (10) with the estimated values $r_{21},r_{12},r_{31},r_{23}$, and $\eta {\eta }_{\mathrm{QE}}$ used as fit parameter. From the fit a relative fluorescence quantum yield of ${\eta }_{\mathrm{QE}}^{3\mathrm{level}}/{\eta }_{\mathrm{QE}}^{2\mathrm{level}}=0.24\pm 0.08$ is obtained, assuming the collection efficiency constant.Reuse & Permissions

Figure 7

Direct measurement of the lifetime for the centers at 744 nm (red squares), 749 nm (blue diamonds), and 764 nm (black circles), using a pulsed laser at 20-MHz repetition rate with 200-ps pulse width. The data were fit with a single exponential. The deduced lifetimes of the centers are, respectively, 4.12 $\pm \hspace{0.5em}0.02$ ns, 1.41 $\pm \hspace{0.5em}0.02$ ns, and 14.2 $\pm \hspace{0.5em}0.2$ ns. (Inset) Antibunching measurement recorded from a single emitter at 749 nm under pulsed laser excitation at 40 MHz and average power of 70 $\mu$W.Reuse & Permissions