Negative charge enhancement of near-surface nitrogen vacancy centers by multicolor excitation

Nitrogen vacancy (NV) centers in diamond have been identified over the past few years as promising systems for a variety of applications, ranging from quantum information science to magnetic sensing. This relies on the unique optical and spin properties of the negatively charged NV. Many of these applications require shallow NV centers, i.e., NVs that are close (a few nm) to the diamond surface. In recent years there has been increasing interest in understanding the spin and charge dynamics of NV centers under various illumination conditions, specifically under infrared (IR) excitation, which has been demonstrated to have significant impact on the NV centers' emission and charge state. Nevertheless, a full understanding of all experimental data is still lacking, with further complications arising from potential differences between the photodynamics of bulk and shallow NVs. Here we suggest a generalized quantitative model for NV center spin- and charge-state dynamics under both green and IR excitation. We experimentally extract the relevant transition rates, providing a comprehensive model which reconciles all existing experimental results in the literature, except for highly nonlinear regimes. Moreover, we identify key differences between the photodynamics of bulk and shallow NVs, and use them to significantly enhance the initialization fidelity of shallow NVs to the useful negatively charged state.

Figure 1

IR excitation impact on steady-state fluorescence of shallow and bulk NVs. (a) Shallow ${\mathrm{NV}}^{-}$ steady-state fluorescence under simultaneous green and IR excitations as a function of both green and IR powers normalized by the fluorescence without the IR excitation, on a logarithmic scale. Regions of enhancement and suppression (above black dotted line) of the fluorescence can be observed. (b) Bulk ${\mathrm{NV}}^{-}$ steady-state fluorescence under simultaneous green and IR excitations normalized by the fluorescence without the IR excitation, as a function of both green (logarithmic scale) and IR (linear scale) powers. Only suppression of the fluorescence is observed.

Figure 2

IR excitation impact on ${\mathrm{NV}}^{-}$ fluorescence as a function of time. Blue dots, experimental data; red curve, fit using numerical simulation according to the model presented in Fig. 3 using the rates extracted in Fig. 4. (a) Time-resolved ${\mathrm{NV}}^{-}$ fluorescence for $159\mu \mathrm{W}$ green laser and 38 mW IR laser. A sharp suppression (A, inset, and Appendix pp5) followed by a slow increase ends with an increase of steady-state fluorescence (B). (b) Time-resolved fluorescence for $241\mu \mathrm{W}$ green laser and 38 mW IR laser. A sharp suppression (A and inset) followed by a slow increase ends with an overall decrease of steady-state fluorescence (B).

Figure 3

Energy level diagram and relevant transitions for the neutral and negatively charged NV center. The ${\mathrm{NV}}^0$ levels are simplified to include only ground and excited states, and for the ${\mathrm{NV}}^{-}$ the $m_s=\pm 1$ states are combined, as are the singlet levels, since their detailed inclusion does not modify the analysis presented in this work. $K_e^{-}$ and $K_e^0$, green excitation rates (for ${\mathrm{NV}}^{-}$ and ${\mathrm{NV}}^0$, respectively), depicted by solid green arrows. $K_f^{-}$ and $K_f^0$, fluorescence rates, depicted by solid red arrows; $K_s$, nonradiative transition rate, depicted by blue arrow; $K_i$ and $K_r$, ionization and recombination rates for both green and IR excitations ($K_{\alpha }=K_{\alpha {}_G}+K_{{\alpha }_{IR}}$), depicted by dashed purple arrows.

Figure 4

Dependence of fast-quench PL on green laser power for surface NVs in the CVD sample. QSS PL normalized by SS PL as a function of green laser power for (a) ${\mathrm{NV}}^{-}$ PL and (b) ${\mathrm{NV}}^0$ PL, with fixed IR power. Black dots: experimental data. Blue curves: fits derived from Eqs. (2) and (3) (see Appendix pp2).

Figure 5

IR effect on ${\mathrm{NV}}^{-}$ steady-state population for surface (solid lines) and bulk (dashed lines) NVs. IR excitation power was optimized for strongest effect (25 mW for surface NVs, 35 mW for bulk NVs). Blue, green excitation only; red, simultaneous excitation with green and IR. Increase of 25% and 3% in ${\mathrm{NV}}^{-}$ population is calculated for surface and bulk NVs, respectively, compared to initialization with green only for the presented power range.

Figure 6

Energy level diagram used in the numerical optimization algorithm for Fig. 2.

Figure 7

High resolution of the PL as a function of time. IR laser is added to a green excitation after $1.3\mu \mathrm{s}$. An approximately constant PL can be seen at $\approx 1.5\mu \sec$ for more than 100 ns.

Figure 8

Fast-dynamics timescale as a function of green laser power. (a) Time evolution for different green powers and constant IR power. (b) Dip time extracted from (a) as a function of green power. Constant decrease rate is shown.

Figure 9

Green laser saturation power dependency on the ionization and recombination cross sections. Saturation curves for different ionization and recombination cross sections are fitted to simulated fluorescence. ${\mathrm{C}}_i$ and ${\mathrm{C}}_r$ denote $\frac{K_{i_G}}{\mathrm{mW}}$ and $\frac{K_{r_G}}{\mathrm{mW}}$, respectively.

Figure 10

Dependence of fast-quench PL on green laser power for bulk NVs in the HPHT sample. QSS PL normalized by SS PL as a function of green laser power for (a) ${\mathrm{NV}}^{-}$ PL and (b) ${\mathrm{NV}}^0$ PL, with fixed IR power. Black dots, experimental data; blue curves, fits derived from Eqs. (D3) and (D4).