High-Temperature Activation of Single-Photon Emitters in monolayer WS2

TL;DR

This work demonstrates that in situ high-temperature annealing of hBN-encapsulated WS2 on suspended micro-heaters can create spectrally isolated defect-bound excitons that function as single-photon emitters at cryogenic temperatures. The emergent line X_L appears around ~1100 K, ~75–86 meV below the neutral exciton with a sub-0.2 meV linewidth, lifetime ~0.9 ns, and antibunching $g^{(2)}(0)\approx0.4$, confirming true single-photon emission. Photoluminescence excitation shows absorption resonances at the A and B excitons, indicating a WS2-origin defect state; temperature calibration employs Raman shifts and a $E_g(T)$ model, validating measurements up to ~1200 K. The results establish high-temperature in situ annealing as a controllable route to defect-bound excitons in van der Waals materials, with encapsulation stabilizing the charge environment and enabling robust quantum light sources for 2D systems.

Abstract

Controlled activation of defect-bound excitonic states in two-dimensional semiconductors provides a route to isolated quantum emitters and a sensitive probe of defect physics. Here we demonstrate that \textit{in situ} high-temperature annealing of hBN-encapsulated monolayer WS$_2$ on a suspended microheater leads to the emergence of spectrally isolated single-photon emitters at cryogenic temperatures. Annealing at temperatures around 1100 K produces a sharp emission line, $X_L$, red-shifted by approximately 80 meV from the neutral exciton and exhibiting a linewidth below 200 $μ$eV. Photoluminescence excitation spectroscopy and power-dependent measurements show that $X_L$ originates from annealing-induced defects in the WS$_2$ monolayer, while second-order photon correlation measurements reveal clear antibunching with $g^{(2)}(0)<0.5$. These results establish high-temperature \textit{in situ} annealing as a controlled means to access defect-bound excitonic states and single-photon emission in van der Waals materials.

Figures (3)

• Figure 1: Optical characterization of hBN-encapsulated monolayer WS2 on a micro-heater membrane. (a) Schematic of the micro-heater chip showing the suspended SiC membrane between metallic contacts and the hBN/WS2/hBN stack transferred on top. (b) Optical image of a representative encapsulated flake on the membrane. (c) Low-temperature PL spectrum at 3.6K displaying sharp neutral-exciton, trion, and dark-state related resonances with linewidths of a few meV ($\lambda_{\mathrm{laser}} = 514.5~\mathrm{nm}$ (CW), power $P_{\mathrm{laser}} = 1~\mu\mathrm{W}$). (d) PLE spectrum of the highlighted low-energy region exhibiting well-resolved A- and B-exciton absorption resonances and excited excitonic states ($P_{\mathrm{laser}} = 4~\mu\mathrm{W}$).
• Figure 2: Thermal-annealing-induced single-photon emission in encapsulated WS2. (a) PL spectra at different temperatures, showing the red-shift of neutral exciton energy with temperature used for temperature calibration. (b, c) PL spectrum after annealing around 1100 K of two samples measured ($\lambda_{\mathrm{laser}} = 514.5$ nm (CW), $P_{\mathrm{laser}} = 1~\mu\mathrm{W},\; T = 3.65~\mathrm{K}$) in panel (b) and ($\lambda_{\mathrm{laser}} = 503~\mathrm{nm}$ (pulsed), $P_{\mathrm{laser}} = 1~\mu\mathrm{W},\; T = 3.72~\mathrm{K}$) in panel (c).
• Figure 3: Optical signatures of the SPE $X_\mathrm{L}$. (a) PLE spectrum recorded at the $X_\mathrm{L}$ emission energy, showing strong resonances at the A and B exciton absorption energies. (b) Power dependence of the $X_\mathrm{L}$ intensity and linewidth, exhibiting saturation of the intensity and resolution-limited linewidth at low power. ($\lambda_{\mathrm{laser}} = 503 ~\mathrm{nm}$ (pulsed)). (c) Time-resolved PL traces of $X_\mathrm{L}$ and trion emission, highlighting the longer lifetime of the localized state. ($\lambda_{\mathrm{laser}} = 506~\mathrm{nm}$ (pulsed), power $P_{\mathrm{laser}} = 2.5~\mu\mathrm{W}$). (d) Second-order correlation function $g^{(2)}(\tau)$ under pulsed excitation, displaying pronounced antibunching with $g^{(2)}(0) < 0.5$. ($\lambda_{\mathrm{laser}} = 506~\mathrm{nm}$ (pulsed), power $P_{\mathrm{laser}} = 1~\mu\mathrm{W}$).