Control and Coherence of the Optical Transition of Single Nitrogen Vacancy Centers in Diamond

We demonstrate coherent control of the optical transition of single nitrogen-vacancy defect centers in diamond. On applying short resonant laser pulses, we observe optical Rabi oscillations with a half period as short as 1 ns, an order of magnitude shorter than the spontaneous emission time. By studying the decay of Rabi oscillations, we find that the decoherence is dominated by laser-induced spectral jumps. By using a low-power probe pulse as a detuning sensor and applying postselection, we demonstrate that spectral diffusion can be overcome in this system to generate coherent photons.

Figure 1

(a) Confocal fluorescence image of individual $\mathrm{N}\mathrm{\text{−}}V$ centers. (b) Resonant excitation leads to exponential decay of fluorescence due to photoionization. Inset: Ionization rate $\Gamma$ as function of power of the resonant excitation laser. (c) Series of resonance scans. Green repump pulses applied in between scans to recharge the $\mathrm{N}\mathrm{\text{−}}V$ center induce spectral jumps. (d) Sum of individual excitation scans yields frequency distribution of spectral jumps for NV1 and NV2. For comparison, for NV2 individual scans have been shifted to coincide in center frequency and then summed, to obtain the intrinsic linewidth. Data is normalized to the area under the curve.

Figure 2

(a) Rabi oscillations for $P=38\mu \mathrm{W}$, $P=19\mu \mathrm{W}$, and $P=4.7\mu \mathrm{W}$. The lowest data set indicates the shape of the excitation pulse. (b) Rabi frequency plotted versus square root of excitation power. (c) Rabi oscillations as function of detuning for ${\Omega }_0=2\pi \times 272\mathrm{MHz}$. (d),(e) Comparison of measured Rabi oscillations on resonance (d) and 300 MHz detuned (e), and numerical data obtained from numerically integrating the optical Bloch equations.

Figure 3

(a) Damping constant $\tau$ of Rabi oscillations as function of Rabi frequency. The dashed line indicates limit of $\tau$ for $T_2$ set by the single-scan linewidth $\Delta =48\mathrm{MHz}$. Coherence is limited by spectral jumps (note difference between NV1 and NV2). Postselection can remove deteriorating effects of spectral jumps and enhance $\tau$ to the value expected from the linewidth. (b) An additional laser field at $\lambda =640\mathrm{nm}$ is applied: This detuned laser has no influence on coherence.

Figure 4

(a) Detuning-resolved Rabi oscillations: detuning from resonance is derived from fluorescence counts during a 20 ms weak resonant probe pulse. Subsequently measured Rabi oscillations are plotted as function of this count rate. Total data is obtained from $3\times {10}^7$ excitation cycles. (b) Data from plot (a) is summed within three ranges. On-resonance (high counts during probe pulse) oscillations show lower Rabi frequency and increased coherence. Transients have been scaled by the indicated factor for better comparison.