Purcell enhanced and tunable single-photon emission at telecom wavelengths from InAs quantum dots in circular photonic crystal resonators

TL;DR

The work demonstrates Purcell-enhanced single-photon emission at telecom wavelengths from InAs/InP quantum dots coupled to circular photonic crystal resonators, achieving both brightness and spectral tunability. Simulations with 3D FEM predict $F_p>20$ and dipole-collection efficiencies near $90\%$ for NA $=0.65$, with resonance scrubbed by adjusting the central-disk radius $R$ and grating period $\Lambda$; experimentally, C-band emission is observed with tunable resonances and bright single-photon output, including phonon-assisted excitation achieving $g^{(2)}(0)=0.019(0.049)$. The authors also realize electrically contacted resonators enabling Stark tuning over $>10$ nm in the telecom O-band, paving the way for scalable, tunable quantum light sources compatible with fiber networks. Overall, the circular photonic crystal resonator platform shows promise for integrated, electrically controllable, high-purity quantum light sources operating at telecom wavelengths.

Abstract

Embedding semiconductor quantum dots into bullseye resonators has significantly advanced the development of bright telecom quantum light sources for fiber-based quantum networks. To further improve the device flexibility and stability, the bullseye approach should be combined with a pin diode structure to enable Stark tuning, deterministic charging, and enhanced coherence. In this work, we fabricate and characterize photonic structures incorporating hole gratings that efficiently support charge carrier transport while maintaining excellent optical performance. We report bright, Purcell-enhanced single-photon emission in the telecom C-band under above-band and phonon-assisted excitation. Additionally, we present electrically contacted resonators, demonstrating wide range tuneability of quantum dot transitions in the telecom O-band. These results mark significant steps toward scalable and tunable quantum light sources for real-world quantum photonic applications.

Figures (9)

• Figure 1: Circular photonic crystal resonator. (a) Pictorial illustration of the device. (b) Scanning electron microscope top view of an exemplary InP device. (c) Simulated performance of a device operating in the telecom C-band, with Purcell factor (blue) and dipole collection efficiency in NA = 0.65 (red) as a function of wavelength. (d) Cavity modes measured under high-power above-band excitation for a series of InP devices with different central disk radii. The nominal radius of the central device (blue) is set to 790 nm. Labels indicate the variation in central disk radius measured in nm.
• Figure 2: Device characterization under above-band excitation. (a) Photoluminescence spectra of two C-band quantum dots embedded in circular photonic crystal cavities, with the brightest transitions at 1529 nm and 1540 nm labeled QD1 and QD2, respectively. (b) SNSPD count rates as a function of the optical pump power for QD1, measured under above-band CW and 80 MHz pulsed excitation. (c) Second-order correlation function $g^2(\tau)$ of QD1 measured under CW excitation at half saturation power. Fitting the HBT data with a three-level model (solid line) yields $g^2(0) = 0.107(0.005)$. (d) Radiative lifetimes of QD1 ($\sim$800 ps), QD2 ($\sim$750 ps), and of an exemplary QD transition in the bare InP slab ($\sim$1.62 ns). The gray area indicates the instrument response function, while the solid lines represent the exponential fit of the decay traces. (e) Comparison of the radiative lifetimes of QD1 and QD2 with those of 10 transitions in the bare InP slab (purple). The dashed line and shaded region indicate the mean and standard deviation, respectively. (f) Second-order correlation function $g^2(\tau)$ of QD2 measured under CW excitation at half saturation power. The HBT data are fitted with a four-level model (solid line), which yields $g^2(0) = 0.281(0.006)$.
• Figure 3: Device characterization under phonon-assisted excitation. (a) Photoluminescence spectrum of QD2 under phonon-assisted excitation. The small peak at 1539 nm corresponds to residual leakage of the excitation laser through the bandpass filters. (b) Second-order correlation function $g^2(\tau)$ of QD2 measured under phonon-assisted excitation. The HBT data is fitted with a three-level function (solid blue line), convolved with a Gaussian function with 70 ps FWHM to account for the detector time resolution. The dashed black line represents the deconvolved fit function. (c) Zoom-in of the HBT fit around zero delay. The deconvolved fit function (black dashed line) yields $g^2(0) = 0.019(0.049)$.
• Figure 4: Electrically contacted resonators. (a) Scanning electron microscope image of an exemplary device fabricated on a suspended GaAs slab. (b) I-V characteristic of the diode measured at room temperature and cryogenic temperature. (c) Microscope image of telecom wavelength emission from QDs coupled to resonators when a forward bias of 1.6 V is applied to the diode. The image is recorded with a 1200 nm long-pass filter in front of the camera to suppress the short wavelength emission. (d) Photoluminescence spectra recorded on an electrically contacted device under above-band excitation as a function of applied voltage to the diode. A wavelength shift exceeding 10 nm is observed for two isolated QD transitions in the telecom O-band.
• Figure S1: Simulations. (a) Top view of the simulation domain, showing the electric field profile $|E^2|$ of the cavity mode. (b) Simulated Purcell factor as a function of wavelength for five InP devices with varying central disk radius. The nominal radius of the central device (blue) is set to 790 nm. Labels indicate the variation in central disk radius measured in nm.