Stark Tuning and Charge State Control in Individual Telecom C-Band Quantum Dots

TL;DR

This work demonstrates Stark tuning and charge-state control of individual InAs/InP quantum dots emitting in the telecom C-band using an n++-i-n+ diode structure grown by MOVPE with droplet epitaxy. The device achieves >2.4 nm spectral tuning via the quantum confined Stark effect and enables deterministic loading of neutral and charged excitons, plus control of the fine structure splitting for potential entangled-photon generation. High-purity single-photon emission is confirmed with $g^{(2)}(0) = 0.04$ after deconvolution, indicating strong antibunching at telecom wavelengths. By suppressing p-dopant diffusion and stabilizing the charge environment, this platform offers a scalable route for cavity QED, spin-photon interfaces, and integration with fiber networks at the C-band.

Abstract

Telecom wavelength quantum dots (QDs) are emerging as a promising solution for generating deterministic single photons compatible with existing fiber optic infrastructure. Emission in the low loss C band minimizes transmission losses, making them ideal for long distance quantum communication. In this work, we present a demonstration of both Stark tuning and charge state control of individual InAs/InP QDs operating within the telecom C-band. These QDs are grown by droplet epitaxy and embedded in an InP based n++-i-n+ heterostructure fabricated using MOVPE. The gated architecture enables the tuning of emission energy via the quantum confined Stark effect, with a tuning range exceeding 2.4 nm. It also allows for control over the QD charge occupancy, enabling access to multiple discrete excitonic states. Electrical tuning of the fine structure splitting is further demonstrated, opening a route to entangled photon pair generation at telecom wavelengths. The single photon character is confirmed via second order correlation measurements. These advances enable QDs to be tuned into resonance with other systems, such as cavity modes and emitters, marking a critical step toward scalable, fiber compatible quantum photonic devices.

Figures (4)

• Figure 1: (a) Schematic of the sample structure. We begin with a 300 nm n-doped InP layer $\bigl(2.0\times 10^{18}\,\mathrm{cm}^{-3}\bigr)$. Next, a 70 nm Al$_{0.48}$In$_{0.52}$As barrier is grown to reduce current and block carriers. The intrinsic region then comprises a 35 nm undoped InP spacer, a 1 nm droplet-epitaxy InAs quantum dot layer (for telecom-wavelength emission), and a 15 nm undoped InP capping layer. Another 70 nm Al$_{0.48}$In$_{0.52}$As barrier provides electrical isolation. Finally, a 35 nm n$^+$ InP layer $\bigl(2.0\times 10^{18}\,\mathrm{cm}^{-3}\bigr)$ and a 35 nm n$^ {++}$ InP layer $\bigl(1.0\times 10^{19}\,\mathrm{cm}^{-3}\bigr)$ is deposited. (b) Schematic conduction band profile above 0.2 V Gate voltage. (c) Microscope image of fabricated mesa structure (0.14 $\mathrm{mm}^{2}$), showing the top and bottom contacts.
• Figure 2: (a) PL for 80 $\mu$W excitation as a function of applied bias voltage for a QD, showing voltage dependent tuning of QD lines in C-band (b) Representative normalized PL spectra from (a) for (i) high background -0.2 V bias voltage and (ii) low background 0.9 V gate voltage. (c) Average background count level with above band excitation of the sample at 80 $\mu$W power at different gate voltages (d) $|I| V$ characteristics of the measured device at 4.2 K (with 30 $\mu$W of above-band (852 nm) excitation power). (e) HBT measurement confirming high purity single-photon emission from the QD at a gate voltage of 1.18 V and emission wavelength of 1530.3 nm. The normalized second-order correlation function $g^{(2)}(\tau)$ exhibits a clear anti-bunching dip at $\tau = 0$. Fitting yields $g^{(2)}(0) = 0.04 \pm 0.04$.
• Figure 3: (a) Micro-PL of a single selected QD as a function of gate voltage, using an above-barrier pumping scheme, in the low background gate voltage regime. (b) Polarization dependent PL of the (i) $X^{0}$, (ii) XX and the $X^{-}$ (iii) emission lines at 1.1 V gate voltage. The emission lines in (i-ii) exhibit polarization-dependent shifts arising from their intrinsic fine-structure splitting, their converse polarization dependence confirm a direct radiative cascade. (c) PL intensity as a function of excitation power for the $X^{0}$, $XX$ and $X^{-}$ states. The dashed lines represent the power-dependent fit before dot saturation. (d) PL intensity spectra under weak above-band pumping (7% of saturation) of the QD for different applied gate voltages. For specific voltage regions, 2 and 4, only the $X^{0}$ or the $X^{-}$ are present respectively.
• Figure 4: (a)(i) PL Intensity of the $X^{0}$ state and (ii) Measurements of the fine structure splitting as a function of the applied gate voltage.