Efficient photon-pair generation in layer-poled lithium niobate nanophotonic waveguides

Abstract

Integrated photon-pair sources are crucial for scalable photonic quantum systems. Thin-film lithium niobate is a promising platform for on-chip photon-pair generation through spontaneous parametric down-conversion (SPDC). However, the device implementation faces practical challenges. Periodically poled lithium niobate (PPLN), despite enabling flexible quasi-phase matching, suffers from poor fabrication reliability and device repeatability, while conventional modal phase matching (MPM) methods yield limited efficiencies due to inadequate mode overlaps. Here, we introduce a layer-poled lithium niobate (LPLN) nanophotonic waveguide for efficient photon-pair generation. It leverages layer-wise polarity inversion through electrical poling to break spatial symmetry and significantly enhance nonlinear interactions for MPM, achieving a notable normalized second-harmonic generation (SHG) conversion efficiency of 4615% W^-1cm^-2. Through a cascaded SHG and SPDC process, we demonstrate photon-pair generation with a normalized brightness of 3.1 × 10⁶ Hz nm^-1 mW^-2 in a 3.3 mm long LPLN waveguide, surpassing existing on-chip sources under similar operating configurations. Crucially, our LPLN waveguides offer enhanced fabrication reliability and reduced sensitivity to geometric variations and temperature fluctuations compared to PPLN devices. We expect LPLN to become a promising solution for on-chip nonlinear wavelength conversion and non-classical light generation, with immediate applications in quantum communication, networking, and on-chip photonic quantum information processing.

Fig. 1. Layer-poled lithium niobate (LPLN) nanophotonic waveguide for efficient photon-pair generation.

a Schematic of LPLN waveguide cross-section in x-cut TFLN. Dark and light pinks indicate inverse domain polarities. b Mode profiles (E_z component) of TE₀₀ mode at 1550 nm and TE₀₁ mode at 775 nm for MPM. c Theoretical comparison of normalized SHG conversion efficiency among different nonlinear TFLN waveguide schemes, including LPLN (red), PPLN (yellow), MPM between TE₀₀ at FH and TE₂₀ at SH (purple), and MPM between TE₀₀ at FH and TE₀₁ at SH without poling (green). The blue line is normalized SHG efficiency versus nonlinear coupling parameter, a measure of mode overlap considering χ⁽²⁾ polarity distribution, with MPM condition. d A false-colored scanning electron micrograph of a LPLN waveguide cross-section, showing the electrical poling induced inverse polarities. e A top-view optical micrograph of a fabricated LPLN waveguide. f Top-view laser-scanning SHG imaging of a LPLN waveguide, where the unpoled waveguide is bright but the poled waveguide becomes dark due to the destructive interference of SH signals from the inversely polarized LN layers. g Schematic of cascaded SHG-SPDC processes for photon-pair generation. h Coincidence spectrum measured from 1486 nm to 1625 nm, covering telecom S, C, and L bands. i, Joint spectral intensity of the photon pairs. The dark cross is due to the residual pump noise

Fig. 2. Numerical analysis and classical measurements of nonlinear LPLN nanophotonic waveguides.

a Numerically calculated nonlinear coupling parameter (Γ) as a function of poling depth. The maximum Γ at a poling depth of ~290 nm (red dot) is more than two orders of magnitude higher than that of an unpoled LN waveguide (green dot). The inset illustrates that Γ is essentially determined by the overlap integral of the electric fields (E_z component) of the two involved modes and the χ⁽²⁾ polarity distribution in a TFLN waveguide. The gray dashed line indicates the 3 dB bandwidth (BW) of Γ, which corresponds to a poling depth range of 113 nm. b Measured SHG power as a function of pump power in a 2.5 mm long LPLN waveguide. A linear fitting reveals an on-chip conversion efficiency of 4615 ± 82% W⁻¹cm⁻². Inset: normalized SHG spectra of a LPLN waveguide (green) and a reference, unpoled waveguide (brown), showing a >20 dB difference. Both waveguides have the same width, are fabricated on the same chip, and are tested under the same conditions. c Simulated phase-matching wavelength shift (at FH) as a function of the TFLN thickness variation (nominal thickness = 600 nm) in a LPLN waveguide (blue) and a similar PPLN waveguide (red). Their thickness sensitivity is fitted to be 2.3 and −10.5 (in unit of nm wavelength shift per nm thickness change), respectively. d Phase-matching wavelength shift (at FH) as a function of waveguide width variation (nominal width = 1100 nm) for LPLN (blue) and PPLN (red) waveguides, showing a sensitivity of −0.32 and −1.5, respectively. Here, the data for the LPLN waveguide is based on the measurement, and data for PPLN is from simulation. Inset: Measured SHG spectra (x-axis: pump wavelength in nm) of LPLN waveguides with widths of 1200 nm (magenta), 1100 nm (green), and 1000 nm (yellow). e Measured phase-matching wavelength shift (at FH) as a function of temperature for LPLN (blue) and PPLN (red) waveguides, showing a fitted temperature sensitivity of 0.18 nm/^∘C and 0.77 nm/^∘C, respectively

Fig. 3. Non-classical characterization of photon-pair generation in a LPLN nanophotonic waveguide using a cascaded SHG-SPDC scheme.

a Experimental setup for (i) photon-pair preparation, (ii) broadband photon spectral characterization, (iii) pair-generation rate, CAR, and heralded correlation measurements, and (iv) two-photon interference measurement. BPF: bandpass filter; PC: polarization controller; WDM: wavelength division multiplexer; FBG: fiber Bragg grating; TBPF: tunable BPF; SNSPD: superconducting nanowire single-photon detector; DWDM: dense wavelength division multiplexer; PS: phase shifter. b Joint spectral intensity constructed by correlation measurement over 32 DWDM channels (ITU frequency Ch14-Ch29 and Ch48-Ch33), revealing strong frequency correlation through coincidences exclusively along the diagonal elements in the 16 × 16 matrix. c Measured (blue dots) and quadratically fitted (blue line) on-chip photon-pair generation rate and coincidence-to-accidental ratio (red) as a function of on-chip pump power. A raw coincidence histogram measured for 15 seconds is shown in the inset. d Heralded second-order correlation function measured at various time delays at a pump power of 0.5 mW. The correlation is 0.008 ± 0.002 at zero time delay, indicating the measurements are in the single-photon regime. e Two-photon interference measured at a pump power of 0.5 mW, and the measured visibility yields 98.0%. Raw coincidence histograms at constructive and destructive interference are shown in the insets. Measurements in c-e are all done between Ch21 and Ch41