Observation of the Dynamic Jahn-Teller Effect in the Excited States of Nitrogen-Vacancy Centers in Diamond

The optical transition linewidth and emission polarization of single nitrogen-vacancy (NV) centers are measured from 5 K to room temperature. Interexcited state population relaxation is shown to broaden the zero-phonon line and both the relaxation and linewidth are found to follow a $T^5$ dependence for $T<100\mathrm{K}$. This dependence indicates that the dynamic Jahn-Teller effect is the dominant dephasing mechanism for the NV optical transitions at low temperatures.

Figure 1

(a) Electronic energy level diagram for the NV center in the presence of a strain field. (b) PL intensity of the $E_x$ and $E_y$ peaks as a function of half wave plate (HWP) angle for NV3. (c) ZPL spectra of a single NV excited with $Y$ polarization. Black: Collecting both polarizations. Red: Collecting $Y$ polarization. Blue: Collecting $X$ polarization.

Figure 2

(a) Temperature dependence of the polarization visibility $V=(I_X-I_Y)/(I_X+I_Y)$ for 5 NVs. The solid black curve corresponds to a fit of NV1-2 data to the three-level model depicted in (b). For the spectrally resolved NV3-5, the determination of the strain axis is discussed in the text. For NV1-2, an “analyzer” half wave plate followed by a fixed polarizer was inserted before the spectrometer to measure both components of the spectrally unresolved $X/Y$ ZPL peaks. The stain axis was determined by measuring the $X/Y$ intensities obtained from rotating the “analyzer” wave plate for specific angles of the “pump” wave plate. Strain splittings for NV1-5 were 8, 9, 44, 54, and 81 GHz, respectively. (b) Three-level model used to fit the temperature dependent polarization relaxation for $Y$-polarized excitation. For $X$-polarized excitation, $a$ and $an$ are reversed.

Figure 3

(a) PLE spectra from NV7 at several temperatures. Transitions are labeled according to the excited orbital state. (b) Individual PLE scans of NV7 illustrating spectral diffusion. Intensity scale: 0 (white) to $8600{\mathrm{s}}^{-1}$ (black). (c) Sum of scans (blue) using the procedure described in the text to remove spectral diffusion, and a Lorentzian fit (red).

Figure 4

Log-log plot showing the corrected PLE linewidths vs temperature from NV6 and NV7, as well as PL linewidths from NV3-5 (PL). The solid line is the $T^5$ fit described in the text. The strain splitting for NV7 was 7.2 GHz and was undetermined for NV6.