Mitigating the Transition of SiV$^-$ in Diamond to an Optically Dark State

TL;DR

This work tackles the problem of SiV$^{-}$ centers transitioning to the optically dark SiV$^{2-}$ state under resonant excitation. It introduces a hybrid optical-electrical stabilization scheme using interdigitated surface electrodes to apply static electric fields, and reveals that the resonant laser, in conjunction with a nearby defect environment, can reverse this transition by generating free holes that re-establish SiV$^{-}$. The key findings show a ≥3× increase in steady-state SiV$^{-}$ photoluminescence under resonant drive for most emitters, with activation of centers near the positively biased electrode and notable emitter-to-emitter variability linked to local surroundings. This approach provides a scalable, Stark-tunable route to deterministic charge-state control in group-IV color centers, advancing their use in quantum networks and information-processing technologies.

Abstract

Negatively charged silicon vacancy centers in diamond (SiV$^-$) are promising for quantum photonic technologies. However, when subject to resonant optical excitation, they can inadvertently transfer into a zero-spin optically dark state. We show that this unwanted change of charge state can be quickly reversed by the resonant laser itself in combination with static electric fields. By defining interdigitated metallic contacts on the diamond surface, we increase the steady-state SiV$^-$ photoluminescence under resonant excitation by a factor $\ge3$ for most emitters, making it practically constant for certain individual emitters. We electrically activate single \sivs near the positively biased electrode, which are entirely dark without applying local electric fields. Using time-resolved 3-color experiments, we show that the resonant laser not only excites the SiV$^-$, but also creates free holes that convert SiV$^{2-}$ to SiV$^-$ on a timescale of milliseconds. Through analysis of several individual emitters, our results show that the degree of electrical charge state controllability differs between individual emitters, indicating that their local environment plays a key role. Our proposed electric-field-based stabilization scheme enhances deterministic charge state control in group-IV color centers and improves its understanding, offering a scalable path toward quantum applications such as entanglement generation and quantum key distribution.

Figures (3)

• Figure 1: A single SiV$^{-}$ in pure bulk diamond under resonant excitation converts into a dark state.a Photoluminescence excitation spectra as a function of resonant excitation power. The emitter PLE presents a Lorentzian line that saturates and broadens with increasing power. For stabilizing the emitter in the negative state, we use 300µW of a 2.41eV-photon energy laser. b Firstly, we apply an off-resonant stabilization laser for 5ms with a short intermediate resonant probe, followed by a 2ms pause. Then, we perform a 38ms resonant readout. During the readout, the counts decay exponentially to almost zero intensity. c Fitted single exponential decay time for the decay from the bright SiV$^{-}$ charge state to the dark state presented in panel b. The decay time is linear in the resonant laser power.
• Figure 2: Charge state stabilization of a single SiV into its negative state. a We use the same optical pulse sequence as in Fig. \ref{['fig:Fig1']} and, as indicated in the top panel, additionally apply an electric voltage. The resonant laser is tuned to the photoluminescence intensity maximum, measured at 20V. The emitter's intensity is negligible for negative voltages and seems highest between $0$ and +150V. b The SiV$^{-}$ photoluminescence intensity decays within less than 1ms at 0V while it disappears in approximately 5ms at 10V and is almost perfectly stable at 50V. Double arrows indicate the intensity differences shown in the following panel for the example of 10V. c The black circles indicate the SiV$^{-}$ photoluminesccence countrate difference between combined resonant/off-resonant and purely resonant excitation. Purple triangles and orange squares are the background-subtracted countrates during the beginning and end of the resonant excitation pulse, respectively. The difference between the initial (purple) and the final count rate (orange) is almost zero above 40V, which indicates that the charge state is stable at these voltages. All intensities decrease at the highest voltages.
• Figure 3: Time-resolved SiV$^-$ photoluminescence and charge state recovery of a single SiV$^{-}$ induced by a near-resonant laser. A green laser pulse initializes the charge state into SiV$^-$. After a delay $\tau_1 = 2ms$, two resonant pulses separated by variable delay $\tau_2$ excite the SiV$^-$. a Without the near-resonant laser, the SiV$^-$ photoluminescence decreases during resonant excitation and remains unchanged after the delay $\tau_2$. Red-shaded regions indicate 1ms integration windows used to calculate intensities $I_{\text{1,init}}$, $I_{\text{1,final}}$, $I_{\text{2,init}}$, and $I_{\text{2,final}}$. b With the near-resonant laser, the intensity increases progressively with increasing $\tau_2$, indicating charge state recovery to the SiV$^-$ charge state. While the near-resonant laser exhibits minimal direct excitation of the SiV$^-$ ground state (evidenced by low photoluminescence during $\tau_2$), it likely excites other defects or creates free carriers that facilitate charge state conversion back to SiV$^-$. c Normalized recovery during $\tau_2$, calculated as the intensity recovered relative to the intensity lost during the first resonant pulse. Without the near-resonant laser, the normalized recovery remains near zero, but approaches unity within 10ms of exposure to the near-resonant laser. d Intensity dynamics at $\tau_2 = 10ms$ for varying bias voltages. Orange diamonds show intensity loss during the first resonant pulse, large black circles show recovery without the near-resonant laser (negligible), and small red circles show recovery with the near-resonant laser. The near-resonant laser restores most of the lost intensity, with complete recovery achieved between 20 and 120V.