Near transform-limited single photons from rapid-thermal annealed quantum dots

TL;DR

This work addresses whether rapid thermal annealing (RTA) can tune the emission wavelength of InAs/GaAs quantum dots without sacrificing quantum optical performance. It employs resonance fluorescence on a single QD in a p-i-n diode, post-grown RTA at high temperature to blue-shift emission, and comprehensive measurements of $T_1$, $T_2$, linewidth, and $g^{(2)}(\tau)$ at cryogenic temperature. The key finding is that RTA preserves high optical quality, yielding near transform-limited photons with $T_2$ close to the Fourier limit and strong antibunching, indicating robust single-photon emission. These results establish RTA as a viable tuning tool for QD emission that maintains coherence and may suppress non-radiative losses, facilitating integration into photonic devices for quantum communication.

Abstract

Single-photon emitters are essential components for quantum communication systems, enabling applications such as secure quantum key distribution and the long-term vision of a quantum internet. Among various candidates, self-assembled InAs/GaAs quantum dots (QDs) remain highly promising due to their ability to emit coherent and indistinguishable photons, as well as their compatibility with photonic integration. In this work, we investigate the impact of post-growth rapid thermal annealing (RTA) on the quantum optical properties of single self-assembled QDs embedded in a p-i-n diode structure. The annealing process induces a controlled blueshift of the emission wavelength by promoting Ga in-diffusion and intermixing. Using resonance fluorescence measurements at cryogenic temperatures (4.2 K), we investigate the single-photon statistics, the emission linewidths, and coherence time $T_2$ of the emitted photons. Our results show that, despite the high annealing temperature of $760^\circ$C, the process does not degrade the optical quality of the quantum dots strongly. Instead, we observe single-photon emission with near transform-limited linewidths, where the dephasing time $T_2$ is only a factor 1.5 above the Fourier-limit $T_2=2T_1$. These findings demonstrate that rapid thermal annealing (RTA) serves as an effective tuning method that preserves the key single-photon emission properties and may help reduce undesirable effects such as non-radiative Auger recombination in quantum photonic applications.

Figures (5)

• Figure 1: Color-coded photoluminescence (PL) intensity map of the single quantum dot (QD1) as function of the gate voltage and emission wavelength for an excitation laser intensity of $1.78\,\mathrm{\mu W/\mu m^2}$. The color scale represents the PL intensity in counts per second. Two emission lines corresponding to the neutral exciton and the charged trion transition are visible with their voltage-dependent shift by the quantum-confined Stark effect.
• Figure 2: a) Resonance fluorescence measurement of the exciton ($X^0$), with its fine structure splitting, and trion ($X^-$) for different laser frequencies and gate voltages. b) The trion transition becomes visible for gate voltages above $50$ mV (indicated by the dashed line), where the quantum dot (QD) is charged with a single electron by tunneling from the electron reservoir. Below this voltage, the QD remains uncharged, and at higher frequencies, the exciton transition can be observed.
• Figure 3: a) Average exciton occupation probability, obtained from integrated photon counts per second for increasing laser excitation intensity. At saturation ($\Omega \approx \Gamma/\sqrt{2}$, with $\Gamma=1/T_1$ and occupation probability equals 0.25) an RF signal of about 1 Mcounts per second is observed. The excitation intensity is also scaled in units of the saturation power $\Omega^2=\Gamma^2/2$. The solid orange line is a fit to the data. b) Power broadening of the exciton transition for selected intensities between 100 $\mathrm{\mu W/\mu m^2}$ and 400 $\mathrm{\mu W/\mu m^2}$. The measured data is shown as dots, while the fits are shown as solid lines. c) Linewidth of the resonant fluorescence from a single quantum dot as a function of excitation intensity. The increase in linewidth at higher intensities is due to power broadening caused by stronger coupling between the excitation field and the dot.
• Figure 4: a) Linewidth measurement on two different rapid-thermally annealed quantum dots, QD1 and QD2, receptively. The spectra were obtained for a fixed gate voltage of $V_g= -39.7\,$mV (QD1) and $V_g= 19.5\,$mV (QD2), while a narrow-band single-mode laser (< 1 MHz) is tuned in frequency steps of 10 MHz across the resonances of the exciton transition. A low excitation intensity of $0.5\, \text{nW}/\mu \text{m}^2$ was used to reduce power broadening and the spectra were fitted with a Lorentzian line-shape. b) Time-resolved QD1 fluorescence measurement under resonant pulsed excitation, where an excitation pulse of $1\,$ns was obtained by using an electro-optical modulator. The fit is a convolution of the system response ($270\,$ps) and an exponential decay that yields a $T_1$-time of $670 \pm 30\,$ps. The inset shows the data on a semi-logarithmic scale, demonstrating the single-exponential decay.
• Figure 5: Normalized second-order correlation measurements $g^{(2)}(\tau)$ of QD1 $g^{(2)}(\tau)$ (blue dots) for three different excitation intensities in a) to c). The orange line represents a fit to the data, convolved with the instrument response function (IRF) as purple line. All correlation measurements exhibit pronounced antibunching at zero time delay. The deconvoluted data, shown as solid red lines, yield a second-order correlation function $g^{(2)}(\tau)$ without the instrument response function. The solid green lines represent fits to the data without considering the instrument response function, assuming an ideal single-photon emitter with the parameter A fixed to unity (see text).