Fig. 1. Configuration and SEM images of the chip. (a) Schematic of the integrated chip, where the grating couplers are used to couple the light into the chip as well as to calibrate the coupling status. The PRF consists of seven cascaded Bragg gratings to filter out the pump light, and a U-shaped NbN nanowire is integrated at the end of the PRF to detect photon pairs. (b) Colorized SEM image of the Bragg gratings (blue). Scale bar corresponds to 1 μm. (c) Colorized SEM image of the U-shaped nanowire (red) on top of the waveguide (blue). Scale bar corresponds to 1 μm. (d) Colorized SEM image of the whole SNSPD. Two inductors made of wider meander nanowires (red) are added and connected to gold contact pads (yellow). Scale bar corresponds to 10 μm.

Fig. 2. Classical measurements of the integrated chip. (a) Transmission spectrum of the PRF at room temperature (300 K) and cryogenic temperature (2.2 K). The PRF shows a rejection ratio above 55 dB, which remains consistent between room and cryogenic temperatures. Inset shows a more accurate rejection ratio of the PRF obtained by scanning the SNSPD in increments of 20 pm, and the rejection ratio achieves 56.5 dB at 1544.5 nm. (b) System detection efficiency curve and dark count rate curve of the SNSPD as a function of bias current, tested at 2.2 K with 1550 nm photons.

Fig. 3. Experimental setup for verifying entangled photon pairs. Energy-time entangled photon pairs are generated in a silicon microring resonator by SFWM, which is pumped by a CW laser, and then coupled to the integrated chips after being separated by a 32 channel 100 GHz DWDM. A time tagger is used to record coincidence counts. DWDM, dense wavelength division multiplexer; PC, polarization controller; MRR, microring resonator.

Fig. 4. Time correlation measurement. (a) Transmission spectrum of the silicon microring, including signal, idler, and pump resonances. Inset is a colorized SEM image of the microring (blue), with a scale bar corresponding to 10 μm. (b) Measured coincidence counts of the correlated photon pairs, where the blue dots represent experimental data integrated for 60 s, and the gray line is the fitting curve.

Fig. 5. Energy-time entanglement measurement. (a) Two UMZIs used for energy-time entanglement analysis, each with a time difference of 400 ps. (b) Single-side count rates of signal and idler photons, respectively. (c) Experimentally measured two-photon interference fringes for Franson-type interference under two nonorthogonal bases. The blue and red dots are measured coincidences integrated for 300 s, and the blue and red lines are fitting curves. Here, the phase difference is represented by the square of the voltage applied on the UMZI. The error bars come from the Poisson distribution of photons and are equal to N, where N is the number of coincidence counts.