Optical Gating of Resonance Fluorescence from a Single Germanium Vacancy Color Center in Diamond

Scalable quantum photonic networks require coherent excitation of quantum emitters. However, many solid-state systems can undergo a transition to a dark shelving state that inhibits the resonance fluorescence. Here, we demonstrate that by a controlled gating using a weak nonresonant laser, the resonant fluorescence can be recovered and amplified for single germanium vacancies. Employing the gated resonance excitation, we achieve optically stable resonance fluorescence of germanium vacancy centers. Our results are pivotal for the deployment of diamond color centers as reliable building blocks for scalable solid-state quantum networks.

Figure 1

(a) Experimental setup. AOM, acousto-optic modulator; BS, $50:50$ nonpolarizing beam splitter; Obj, objective; $S$, sample; $F$, spectral filter; SPAD, single-photon avalanche detector. Inset, schematic of a GeV center. (b) Normalized PL spectrum of the GeV color center at 5 K, excited at 532 nm with a power of 0.4 mW ($0.06P_1$) for an exposure time of 5 s. The purple line is the fitting with four Gaussian peaks, labeled as $A$, $B$, $C$, and $D$ from high to low energy. Inset, energy structure of the GeV center with four optical transitions labeled according to the spectrum. Splitting reflects the best-fit parameters. (c) Scanning electron microscope (SEM) image of a FIB milled SIL. Scale bar, $3\mu \mathrm{m}$. (d) Room temperature second-order autocorrelation function $g^{(2)}(\tau )$ of the GeV center under nonresonant excitation (532 nm, $\sim 1\mathrm{mW}$). Fitting with single exponential decay (solid line) gives $g^{(2)}(0)=0.17\pm 0.03$.

Figure 2

(a) PLE spectra of transitions $C$ (left) and $D$ (right) when the gating laser at 532 nm is on (orange) and off (blue). The purple curve corresponds to the gating wavelength at 405 nm. Zero detuning corresponds to 602.2903 and 602.4828 nm for $C$ and $D$, respectively. (b) RF intensity of transition $C$ for gating on (orange) and gating off (blue). Right, PL intensity under solely nonresonant excitation. Background has been subtracted from the data. Bin size, 100 ms. (c) $g^{(2)}(\tau )$ of the GeV center under resonant pumping of transition $C$. $g^{(2)}(0)=0.07$. The oscillatory signal at $\tau \approx 5\mathrm{ns}$ is the Rabi oscillations induced by resonant pumping. Inset, stochastic jump of the RF. Bin size, $33\mu \mathrm{s}$. (d) Rabi oscillations of transition $C$ in pulsed measurement. Red curve is a fitting by the two-level system [22]. Inset, Rabi frequency against the square root of resonant power with a linear fit (red). Resonant power is 200 nW for (a) and (b), 300 nW for (c), and $2.4\mu \mathrm{W}$ for (d). Nonresonant power is $1.2\mu \mathrm{W}$ ($1.8\times {10}^{-4}$ $P_1$) for all.

Figure 3

2D map of normalized PLE spectra (transition $D$) by varying (a) the resonant excitation power, or (e) the nonresonant excitation power. Normalization constant, (a) $10\mathrm{kcnt}/\mathrm{s}$ and (e) $4\mathrm{kcnt}/\mathrm{s}$. Gating power in (a), $7\times {10}^{-5}{P}_1$; resonant power in (e), $0.35P_0$. (b) and (f) are the center frequency ${\nu }_0$ of each line in (a) and (e), respectively, extracted from Lorentzian fitting. The shaded region represents the standard deviation of ${\nu }_0$, (b) $\sigma \sim 50\mathrm{MHz}$, and (f) $\sigma \sim 25\mathrm{MHz}$. (c) and (g) are the Lorentzian linewidth $\mathrm{\Delta }\nu$ of each line in (a) and (e), respectively. (d) Resonant-power dependence of RF, measured by setting the resonant laser at zero detuning. (h) Gating-power dependence of RF (blue), evaluated by subtracting the background from the maximum count rate of each line in (e). Background count rate (red) is measured at a far detuning of $\sim 10\mathrm{GHz}$.

Figure 4

Gating and shelving dynamics. (a) Time-resolved PL by modulating the resonant beam with constant nonresonant power of $7\times {10}^{-7}{P}_1$. (b) Physical model. $G$, $E$, $D$, and $M$: ground, excited, dark, and metastable state; $k_{GD}$, $k_{DG}$, and $k_{ED}$: population transfer rates from $G$ to $D$, $D$ to $G$, and $E$ to $D$. $\mathrm{\Omega }$: resonant Rabi frequency; ${\mathrm{\Gamma }}_{\mathrm{sp}}=1/T_1=280\mathrm{MHz}$: spontaneous decay rate, determined by lifetime measurement [22]. Gray arrows depict the possible physical processes underlying $k_{ED}$. (c) and (f) are the time-resolved PL by modulating the nonresonant beam with (c) constant resonant power ($0.9P_0$) or (f) constant nonresonant power ($7\times {10}^{-5}{P}_1$). Black curves are the fittings by using Eq. (3). (d) and (g) are the dynamical rates extracted from the fittings in (c) and (f), respectively. Dashed blue horizontal lines in (d) depict $k_{ED}$, representing its trivial nonresonant-power dependence in this experiment. Solid straight lines in (d) are the fittings with $k_{GD}^{\mathrm{on}}=3.5\times {10}^6\times P^{0.96}$ (red) and $k_{DG}^{\mathrm{on}}=2.1\times {10}^7\times P^{1.07}$ (purple), where $P$ denotes the nonresonant power in the unit of $P_1$. (e) and (h) are the on-period steady-state population of dark state ${\rho }_D^{\infty }$ and two-level system ${\rho }_E^{\infty }$, evaluated by using the rates in (d) and (g), respectively. In (a), (c), and (f), raw data (orange dots) are vertically shifted for clarity, with the zero-intensity level indicated by the gray horizontal lines. Top panel: modulation protocol.