Fig. 1. (a) Cross-sectional illustration of the waveguide structure which consists of a polymer waveguide on top of an LNOI substrate. (b) Schematic of the directional coupler with w1=3.4  μm, w2=0.92  μm, gap=0.35  μm, and L=162  μm. (c) Simulated effective refractive indices of the TE0,TE1,TM0,TM1, and TM2 modes at 780 nm and the TE0 mode at 1560 nm as a function of the waveguide width w. The four insets show the corresponding electric field (|E|) profiles. (d) Simulated propagation loss of the TE0 mode and TM2 mode as a function of the waveguide width w at the respective wavelengths with specific thickness. The BIC point for the TM2 mode is indicated. (e), (f) Electric field (|E|) profiles of the TE0 mode at SHG pump wavelength (e) and the TM2 mode at second-harmonic wavelength (f). (g) Normalized spectrum of second-harmonic generation from a fabricated device.

Fig. 2. Experimental setup to characterize the SHG conversion and photon-pair generation. (a) Second-harmonic generation measurement. EDFA, erbium-doped fiber amplifier; FPC, fiber polarization controller; PD, photodetector. (b) Correlated photon-pair measurement. LPF, long-pass filter; SNSPD, superconducting nanowire single-photon detector; TCSPC, time-correlated single-photon counter. (c) Time-energy entanglement measurement. (d) Heralded single-photon detection.

Fig. 3. (a) One sample data series of single count rates CA and CB of two detector channels A and B in 1 min time. (b) Coincidence counts per minute. The coincidence time window τw is 3 ps. (c) Second-order correlation function g(2)(τ) of generated photon pairs. The circles are the normalized g(2) function calculated from measured coincidence count rates with Eq. (2), and the solid line is the fitted g(2) function from Eq. (3).

Fig. 4. (a) On-chip pair generation rate (PGR, units of MHz) versus pump power (units of mW) in the waveguide. The error bars are one standard deviation in each direction. (b) Time-bin coincidence histogram detected for a 0.42 mW on-chip pump power. (c) Coincidence-to-accidental ratio (CAR) versus PGR. The data in this figure are measured in Fig. 2(b).

Fig. 5. (a)–(c) Coincidence histograms over 1 min for different interferometric phases with the corresponding phases noted in (d). (d) The left axis shows the normalized coincidence counts versus the phase. The visibility of the raw data yields 94%. The right axis shows the single counts measured at the corresponding phase. The data in this figure are measured in Fig. 2(c).