Frequency-Stabilized Source of Single Photons from a Solid-State Qubit

Single quantum dots are solid-state emitters that mimic two-level atoms but with a highly enhanced spontaneous emission rate. A single quantum dot is the basis for a potentially excellent single-photon source. One outstanding problem is that there is considerable noise in the emission frequency, making it very difficult to couple the quantum dot to another quantum system. We solve this problem here with a dynamic feedback technique that locks the quantum-dot emission frequency to a reference. The incoherent scattering (resonance fluorescence) represents the single-photon output, whereas the coherent scattering (Rayleigh scattering) is used for the feedback control. The fluctuations in emission frequency are reduced to 20 MHz, just approximately $5\%$ of the quantum-dot optical linewidth, even over several hours. By eliminating the $1/f$-like noise, the relative fluctuations in quantum-dot noise power are reduced to approximately ${10}^{-5}$ at low frequency. Under these conditions, the antibunching dip in the resonance fluorescence is described extremely well by the two-level atom result. The technique represents a way of removing charge noise from a quantum device.

Popular Summary

Photons are quantum particles that form light. Single photons carry quantum information in their polarizations or the phases of their wave packets, a feature that can be exploited in quantum communication and computation. In such applications, a single-photon source, a device that emits photons one by one, is a prerequisite. One of the most promising platforms for single-photon sources is based on semiconductor quantum dots. One major unsolved problem is, however, that the “color” (or wavelength) of the photons emitted by a quantum dot is not locked to a precise value; rather, it wanders around randomly. In this experimental paper, we demonstrate a way of creating, from a single quantum dot, single photons that all have the same color.

The randomness in the wavelength of the photons emitted from a quantum dot originates from the charge noise—the fluctuations of the charges in the semiconductor. The cure would then be to remove that charge noise. To this end, we have developed a quantum-classical hybrid system, where a powerful and dynamic feedback link connects a single quantum dot to a classical system of a constant-wavelength laser. Our scheme involves measuring the optical absorption of the quantum dot in a way that is very sensitive to any fluctuation in the wavelength. The fluctuation is canceled by applying exactly the opposite effect to a gate positioned above the quantum dot. With this system, we have succeeded in generating a nearly perfect stream of single-colored single photons.

Our work not only enables a stable single-photon source with potential for on-chip integration, but its highly effective removal of the charge noise in semiconductor quantum dots may also lead to a radical improvement in semiconductor-based qubits.

Figure 1

(a) Schematic view of the experiment. The narrow-band laser is stabilized to a fixed frequency by a wave meter which in turn is stabilized to a HeNe laser. Laser light is guided through optical fibers (yellow curves) and microscope optics before it is focused onto the sample, driving the $X^0$ transition resonantly ($\mathrm{BS}=\mathrm{\text{beam}}\mathrm{\text{splitter}}$, $\mathrm{PBS}=\mathrm{\text{polarizing}}\mathrm{BS}$, and $\mathrm{Pol}.=\mathrm{\text{linear polarizer}}$). Two simultaneous measurements of $X^0$ scattering are performed: RF, detected with an APD, and absorption with a photodiode (PD) underneath the sample. The dynamic stabilization is realized with an active PID feedback loop that corrects for fluctuations in the transition energy using the gate voltage $V_g$ and the square-wave modulation of a function generator (FG). (b) RF signal of the fine-structure split $X^0$ emission of a single quantum dot at wavelength 936.5 nm, a power corresponding to a Rabi energy $\Omega$ of $0.74\mu \mathrm{eV}$ and a temperature of 4.2 K. A detuning is achieved by sweeping the gate voltage. The solid red line is a Lorentzian fit to the data with linewidth $\Gamma =1.28\mu \mathrm{eV}$ (309 MHz) and $\Gamma =1.45\mu \mathrm{eV}$ (350 MHz) and with a fine-structure splitting $\Delta =11.8\mu \mathrm{eV}$. (c) The differential transmission ($\Delta T/T$) signal on the same quantum dot with integration time 100 ms per point. The red curve is a fit to the derivative of the two Lorentzians. The signal around the zero-crossing point ($\Delta T/T=0$) is used to generate an error signal for the feedback scheme.

Figure 2

Histogram of the RF peak position (a)–(c) with and (d) without the stabilization scheme. A triangle $V_g$ is applied. The scanning rate of the laser is $8.0\mu \mathrm{eV}/\mathrm{s}$ with period 10 s. Histograms of the RF peak position for (a) up-sweeps and (b) down-sweeps recorded with feedback. (c) The histogram is shown with feedback, negative detunings from the up-sweeps and positive detunings from the down-sweeps. A histogram without feedback is shown in (d). The standard deviation $\sigma$ is reduced from (d) $0.250\mu \mathrm{eV}$ (61 MHz) without active stabilization to (c) $0.089\mu \mathrm{eV}$ (22 MHz) with active stabilization.

Figure 3

(a) Time trace of the RF of a single quantum dot (the one from Fig. 1) with $\delta =0\mu \mathrm{eV}$ recorded over several hours. The binning time is $t_{\mathrm{bin}}=100\mathrm{ms}$. The time trace is plotted with (red) and without (blue) the dynamic stabilization scheme. (b) 5-s excerpts of the unstabilized (blue) and stabilized (red) time traces, with the dashed lines representing the shot-noise limits. (c) Noise spectra of the normalized RF signal $S(t)/⟨S(t)⟩$ corresponding to the time traces of (a) after correction for external noise sources. The shot noise in the experiment is shown with the dashed lines.

Figure 4

Second-order correlation $g^2(t)$ for the stabilized RF from the $X^0$ (black points). The red curve shows a convolution of the two-level atom result with a Gaussian distribution that describes the timing jitter of the detectors. The blue curve shows the two-level atom response alone.