Very Large and Reversible Stark-Shift Tuning of Single Emitters in Layered Hexagonal Boron Nitride

Combining solid-state single-photon emitters (SPEs) with nanophotonic platforms is a key goal in integrated quantum photonics. In order to realize functionality in potentially scalable elements, suitable SPEs have to be bright, stable, and widely tunable at room temperature. In this work, we show that selected SPEs embedded in a few-layer hexagonal boron nitride ($h$-BN) meet these demands. In order to show the wide tunability of these SPEs we employ an atomic force microscope (AFM) with a conductive tip to apply an electrostatic field to individual $h$-BN emitters sandwiched between the tip and an indium-tin-oxide-coated glass slide. A very large and reversible Stark shift of $(5.5\pm 0.3)\mathrm{nm}$ at a zero-field wavelength of 670 nm is induced by applying just 20 V, which exceeds the typical resonance linewidths of nanodielectric and even nanoplasmonic resonators. Our results help to further understand the physical origin of SPEs in $h$-BN as well as for practical quantum photonic applications where wide spectral tuning and on/off resonance switching are required.

Figure 1

Schematic representation of the experiment, electrostatic field distribution and $g^{(2)}$ function. (a) An $h$-BN flake is located on an ITO-covered glass substrate. The oil-immersion objective lens below excites the $h$-BN SPE and collects its emission as an AFM tip can be used to deliver an electrostatic field causing a Stark shift of the ZPL. (b) The electrostatic field strength between the AFM tip ($h$-BN starts at $y=125\mathrm{nm}$) and ITO (starts at $y=0$) caused by the application of 20 V is represented by the contour diagram. The superimposed stream flow chart shows the field orientation. The assumed SPE position (discussed in the text) is marked by the white rectangle. Points with the highest $|\overrightarrow{\mu }\overrightarrow{E}|$ and $|\overrightarrow{E}{|}^2$ within the assumed SPE area are marked by the triangle and the circle, respectively. (c) The second-order autocorrelation of the emitters’ fluorescence [$g^{(2)}(\tau )$] shows a clear antibunching at $\tau =0$ that indicates for primarily single-photon emission.

Figure 2

Determination of the dipole orientation. (a) Simulated degree of polarization $\delta$ with respect to the dipole out-of-plane angle $\theta$. The orange circle represents the measured $\delta$. In the inset, a corresponding polarization measurement of the fluorescence light (dots) and a fit (solid line) is shown. The signal is normalized to the total intensity detected by both APDs and corrected for its different detection efficiencies. (b) Simulated Fourier image with a dipole orientation of $\theta =(59.9\pm 0.2{)}^{\circ }$ and $\varphi =(52.9\pm 0.2{)}^{\circ }$ determined by the polarization measurement shown in (a). (c) Fourier image of the SPE taken with a $\mathrm{NA}=1.4$ objective lens. The striking similarity between (b) and (c) proves that the simulation is suited to derive the experimental results very well.

Figure 3

Spectra of the Stark-shifted $h$-BN SPE fluorescence. (a) Measured spectra and corresponding fits of the unshifted emission in blue (left peak) and shifted in orange (right peak), taken with an integration time of 2.5 s and binned to 1-nm bins. This shift of $(5.9\pm 0.6)\mathrm{nm}$ is recorded with a tip to substrate distance of approximately 125 nm and a voltage of 20 V. (b) $h$-BN-ZPL spectra recorded while the electric field strength between tip and ITO layer is lowered and increased again (from left to right, in 1-V steps). (c) Blue dots are spectral shifts determined by fits of Eq. (S4) within the Supplemental Material [29] to the data shown in (b). The solid line shows Eq. (2) that is fit to the data points. (d) Spectra recorded while a voltage of 20 V is switched on and off. (e) Spectral shifts determined by fits to the data in (d) show an average shift of $(5.5\pm 0.3)\mathrm{nm}[(15.4\pm 0.8)\mathrm{meV}]$.