A scalable near-visible integrated photon-pair source for satellite quantum science

Abstract

Quantum state distribution over vast distances is essential for global-scale quantum networks and fundamental test of quantum physics at space scale. While satellite platforms have demonstrated thousand-kilometer entanglement distribution, quantum key distribution and quantum teleportation with ground, future constellations and deep-space missions demand photon sources that are robust, compact, and power-efficient. Integrated photonics offers a scalable solution, yet a critical spectral gap persists. Although telecom-band integrated photon-pair sources are well established, near-visible photons offer distinct advantages for satellite-to-ground links by mitigating diffraction loss and maximizing the collection efficiency of optical telescopes. Scalable integrated sources in this regime have remained elusive due to the fundamental challenge of achieving anomalous dispersion in materials transparent at visible wavelengths. Here we bridge this gap by demonstrating an integrated near-visible photon-pair source based on a wide-bandgap, ultralow-loss, silicon nitride (Si$_3$N$_4$) microresonator. By engineering the dispersion of higher-order waveguide modes, we overcome the intrinsic normal dispersion limit to achieve efficient phase matching. The device exhibits a spectral brightness of 4.87$\times$10$^7$ pairs/s/mW$^2$/GHz and a narrow photon linewidth of 357 MHz. We report high-purity heralded single-photon generation with a heralding rate up to 2.3 MHz and a second-order correlation function as low as 0.0041. Furthermore, we observe energy-time entanglement with 98.4% interference visibility, violating the CHSH limit even at flux exceeding 40.6 million pairs/s. Combined with the proven radiation hardness of Si$_3$N$_4$, this source constitutes a flight-ready hardware foundation for daylight quantum communications and protocols requiring on-orbit multiphoton interference.

Figures (7)

• Figure 1: Scalable manufacturing and satellite deployment of near-visible integrated photon-pair sources.a, Photograph of a 150-mm-diameter Si$_3$N$_4$ wafer fabricated using CMOS-compatible foundry processes, yielding hundreds of photonic chips. b, Optical micrograph of a single chip hosting an array of microresonators, demonstrating high-density integration. c, Magnified view of the specific microresonator utilized for photon-pair generation in this work. d, The compact packaged module occupies a volume of only a few cubic centimeters, meeting the strict SWaP requirements for space payloads. e, Artistic rendering of the packaged source deployed on a quantum satellite. f, Performance benchmarking of state-of-the-art integrated photon-pair sources Chen:2024Li:25Steiner:21Kumar:19Guo:17aZeng:24Zhao:22Ma:20Rahmouni:24Yasui:25. Spectral brightness (calibrated at 1 mW pump power) is plotted against photon linewidth. Blue and red circles denote sources operating in the telecommunication and near-visible bands, respectively. Inset: Comparison of the bandgaps and transparency windows of common integrated photonic materials. The wide bandgap of Si$_3$N$_4$ (5.0 eV) makes it a unique candidate for high-performance operation in the near-visible band.
• Figure 1: Characterization of PGR and CAR.a, Experimental definition of the coincidence count rate $n_\mathrm{cc}$ and accidental coincidence count rate $n_\mathrm{acc}$. A representative two-photon correlation histogram (recorded at $P=1.52$ mW) is presented to illustrate the extraction procedure in the presence of background noise. The $n_\mathrm{cc}$ is determined by integrating the histogram bins within a coincidence window centered at zero delay, while the $n_\mathrm{acc}$ is calculated using an identical window at a large delay offset. The coincidence window width is adaptable, allowing for optimization across different experimental regimes. b, Measured single-photon count rates for signal ($n_\mathrm{s}$) and idler ($n_\mathrm{i}$) modes as a function of pump power $P$. Data points are fitted with polynomial functions (solid curves). The linear components (dashed lines) represent the background noise floor. c, Net coincidence count rate $n_\mathrm{net}$ as a function of $P$, exhibiting the expected quadratic scaling confirmed by a polynomial fit. d, Measured CAR as a function of $P$. Error bars represent one standard deviation derived from Poissonian counting statistics.
• Figure 2: Design and dispersion engineering of the near-visible photon-pair source.a, Principle of cavity-enhanced SFWM. Two pump photons at $f(\mu_\mathrm{p})$ annihilate to generate a signal–idler pair at $f(\mu_\mathrm{s})$ and $f(\mu_\mathrm{i})$, strictly aligned with the microresonator resonance grid. b, Diagram illustrating the phase-matching mechanism. Nonlinear SPM and XPM induce frequency detunings ($-\Delta$ and $-2\Delta$ for the pump and signal/idler resonances) that distort the resonance grid; efficient SFWM requires compensating for these shifts via anomalous GVD. c, Schematic of the Si$_3$N$_4$ microresonator featuring an adiabatically tapered waveguide geometry, accompanied by simulated optical modes. The waveguide width is expanded to 2200 nm in the loop to minimize propagation loss, and tapered to 1450 nm in the coupling region to ensure TE$_{20}$ excitation with near-unity coupling ideality. d, Simulated $D_2/2\pi$ for eigenmodes of a Si$_3$N$_4$ microresonator with 41.6 $\mu$m radius, 2200 nm width and varying thickness from 300 to 900 nm. While the TE$_{00}$ mode exhibits strong normal GVD ($D_2<0$, solid lines) in the near-visible regime (orange shaded zoom) regardless of thickness, the higher-order modes TE$_{20}$ and TE$_{30}$ (dashed lines) achieve near-zero GVD. e, FDTD simulation of the coupling region, confirming near-unity coupling ideality from the microresonator TE$_{20}$ mode to the bus TE$_{00}$ mode. The field magnitude is plotted as $|E|^{0.5}$ to enhance visualization. f, Simulated transmissions illustrating mode selectivity. The adiabatic coupler design suppresses parasitic excitation of modes other than the TE$_{20}$ by $>15$ dB.
• Figure 2: Observation of the transition from SFWM to OPO. As pump power $P$ increases, $g_\mathrm{h}^{(2)}(0)$ rises abruptly toward unity, signaling the transition from SFWM to OPO. The power where this transition occurs is defined as the oscillation threshold $P_\mathrm{th}$. a, Measured $g_\mathrm{h}^{(2)}(0)$ and heralding rate $n_1$ for the TE$_{20}$ mode (loaded $Q = 0.72 \times 10^6$) as a function of $P$ up to 2 mW. At $P_\mathrm{th} = 1.52$ mW, the heralding rate is $n_1=2.3$ MHz. b, Measured $g_\mathrm{h}^{(2)}(0)$ and heralding rate $n_1$ for the TM$_{10}$ mode (loaded $Q = 1.4 \times 10^6$) as a function of $P$ up to 1.2 mW. At $P_\mathrm{th} = 0.69$ mW, the heralding rate is $n_1=2.1$ MHz. Error bars represent one standard deviation derived from Poissonian counting statistics.
• Figure 3: Spectroscopic characterization and dispersion verification.a, Normalized microresonator transmission spectrum. Red stars mark the specific resonances selected for photon-pair generation: pump ($384.49233$ THz), signal ($383.43767$ THz), and idler ($385.54702$ THz). b, Measured $D_\mathrm{int}/2\pi$ profile. The experimental data confirm the engineered anomalous GVD of the TE$_{20}$ mode, in sharp contrast to the normal GVD exhibited by the TE$_{00}$ mode (simulated, solid curve). Inset: Magnified view of the phase-matching window. The pump mode frequency lies below the arithmetic mean of the signal and idler frequencies (dashed line), satisfying the phase-matching conditions required for SFWM. c, Resonance profiles of the signal, pump and idler modes, overlaid with their Lorentzian fits. The extracted $\kappa_0/2\pi$ and $\kappa_\mathrm{ex}/2\pi$ are marked for each mode.