Table of Contents
Fetching ...

Low-power integrated optical parametric amplification via second-harmonic resonance

Devin J. Dean, Taewon Park, Hubert S. Stokowski, Luke Qi, Sam Robison, Alexander Y. Hwang, Jason Herrmann, Martin M. Fejer, Amir H. Safavi-Naeini

TL;DR

The paper tackles the challenge of delivering practical, low-power, broadband, and low-noise optical parametric amplification on a chip. It introduces a second-harmonic-resonant architecture in thin-film lithium niobate that resonantly builds the SH pump while a broadband traveling-wave OPA performs signal amplification, enabling >17 dB gain with <200 mW input power and a 110 nm near-quantum-limited noise bandwidth. The approach achieves high SHG efficiency (up to 95%), pump recirculation, and robust pump–signal multiplexing via dichroic couplers, yielding on-chip gains exceeding 12 dB across wide spectral regions and a near-quantum-limited NF across 1520–1630 nm. The results hold promise for scalable, integrated OPAs in both quantum and classical photonics and can be extended to other wavelength bands using quadratically nonlinear materials.

Abstract

Optical amplifiers are fundamental to modern photonics, enabling long-distance communications, precision sensing, and quantum information processing. Erbium-doped amplifiers dominate telecommunications but are restricted to specific wavelength bands, while semiconductor amplifiers offer broader coverage but suffer from high noise and nonlinear distortions. Optical parametric amplifiers (OPAs) promise broadband, quantum-limited amplification across arbitrary wavelengths. However, their miniaturization and deployment has been hampered by watt-level power requirements. Here we demonstrate an integrated OPA on thin-film lithium niobate that achieves >17 dB gain with <200 mW input power -- an order of magnitude improvement over previous demonstrations. Our second-harmonic-resonant design enhances both pump generation efficiency (95% conversion) and pump power utilization through recirculation, without sacrificing bandwidth. The resonant architecture increases the effective pump power by nearly an order of magnitude compared to conventional single-pass designs, while also multiplexing the signal and pump. We demonstrate flat near-quantum-limited noise performance over 110 nm. Our low-power architecture enables practical on-chip OPAs for next generation quantum and classical photonics.

Low-power integrated optical parametric amplification via second-harmonic resonance

TL;DR

The paper tackles the challenge of delivering practical, low-power, broadband, and low-noise optical parametric amplification on a chip. It introduces a second-harmonic-resonant architecture in thin-film lithium niobate that resonantly builds the SH pump while a broadband traveling-wave OPA performs signal amplification, enabling >17 dB gain with <200 mW input power and a 110 nm near-quantum-limited noise bandwidth. The approach achieves high SHG efficiency (up to 95%), pump recirculation, and robust pump–signal multiplexing via dichroic couplers, yielding on-chip gains exceeding 12 dB across wide spectral regions and a near-quantum-limited NF across 1520–1630 nm. The results hold promise for scalable, integrated OPAs in both quantum and classical photonics and can be extended to other wavelength bands using quadratically nonlinear materials.

Abstract

Optical amplifiers are fundamental to modern photonics, enabling long-distance communications, precision sensing, and quantum information processing. Erbium-doped amplifiers dominate telecommunications but are restricted to specific wavelength bands, while semiconductor amplifiers offer broader coverage but suffer from high noise and nonlinear distortions. Optical parametric amplifiers (OPAs) promise broadband, quantum-limited amplification across arbitrary wavelengths. However, their miniaturization and deployment has been hampered by watt-level power requirements. Here we demonstrate an integrated OPA on thin-film lithium niobate that achieves >17 dB gain with <200 mW input power -- an order of magnitude improvement over previous demonstrations. Our second-harmonic-resonant design enhances both pump generation efficiency (95% conversion) and pump power utilization through recirculation, without sacrificing bandwidth. The resonant architecture increases the effective pump power by nearly an order of magnitude compared to conventional single-pass designs, while also multiplexing the signal and pump. We demonstrate flat near-quantum-limited noise performance over 110 nm. Our low-power architecture enables practical on-chip OPAs for next generation quantum and classical photonics.

Paper Structure

This paper contains 23 sections, 31 equations, 13 figures, 1 table.

Figures (13)

  • Figure 1: (a) Important metrics for an on-chip OPA applications. Representative values are listed - exact metric requirements vary with application. (b) Energy flow in second-harmonic-resonant OPA. Energy flows from the fundamental pump to the broadband amplified signal via the resonant second harmonic pump. (c) Image of fabricated photonic chip atop copper holder. (d) Simulated OPA gain as a function of input power, for parameters measured in this work ($\gamma = 0.3$, $\eta = 2000\frac{\%}{\text{Wcm}^2}$, $L = 6$ mm) and also for improved parameters ($\gamma = 0.3$, $\eta = 4000\frac{\%}{\text{Wcm}^2}$, $L = 1.2$ cm). The top of each curve represents phase-sensitive gain, while bottom represents phase-insensitive gain (Equation (\ref{['eq_G_OPA']})). Shaded green represents desired integrated OPAs with low power and high gain. (e) Chip implementation of second-harmonic-resonant OPA. All couplers shown are dichroic couplers that couple nearly all light around the fundamental frequency while almost no light at the second harmonic.
  • Figure 2: Resonant Second Harmonic Generation. (a) SHG transfer function, showing hundreds of SH resonances across 10s of nm. (b) Transmission of FH pump (blue) and leakage of resonant second harmonic (orange, scaled to match FH pump), for two different on-chip input powers. (c) Conversion efficiency from input to second harmonic, measured by depletion of the FH pump. Dots are measured data while lines are model for $\eta_0 = 2000\frac{\%}{\text{Wcm}^2}$ and $\gamma \approx 0.3$. (d) Second harmonic pump power to OPA, for both resonant and nonresonant cases and the same parameters given above. Dotted line shows when the second harmonic power in the OPA is equal to the input fundamental power to the system.
  • Figure 3: (a) OPA Measurement Setup. Fundamental pump laser is amplified by an erbium-doped fiber amplifier (EDFA), sent through a fiber polarization controller (FPC), and coupled by lensed fiber onto the chip, where it generates the resonant second harmonic pump. A nondegenerate signal laser is input on the other side of the chip for phase-insensitive amplification measurements. For phase-sensitive measurements, a tap of the fundamental pump laser is input as degenerate seed instead. Output light is collected by multimode fiber and measured by an optical spectrum analyzer (OSA). (b) Lower Plot: OPA on-off gain spectrum for on-chip FH pump power of 225 mW. The gain at each signal wavelength is measured while the FH pump wavelength is tuned. Upper Plot: Gain spectrum for fixed pump wavelength at 1575.59 nm on a single SH resonance. Points are extracted from one horizontal slice of the lower plot, and the line is to guide the eye. (c) On-chip net gain as a function of on-chip FH pump power, for signal wavelengths around 1590 nm. Green markers represent phase-insensitive amplification measurements and blue markers represent phase-sensitive ones. Error bars on phase-insensitive amplification measurements show the minimum and maximum gain to accurately show the gain ripple due to chip-facet reflections as discussed in section \ref{['sec_methods_OPA']}. Curves are calculated based on the SHG performance of section \ref{['secSHG']}.
  • Figure 4: OPA Noise Figure Measurements. (a) Signal gain for one FH pump wavelength around 1575.31 nm, alongside the spontaneous parametric fluorescence spectrum with 2 nm resolution setting on the optical spectrum analyzer. (b) Phase-insensitive noise figure, as a function of signal wavelength. Solid purple line represents the expected noise figure based on measured losses and the SPF spectrum of (a). Dotted line represents the quantum-limit of noise figure based on the SPF spectrum of (a). (c) Degenerate amplification as phase drifts in time. (d) Phase-sensitive noise figure based on SPF level and amplification. Points are measured datasets from SPF and gain for three adjacent SH pump modes, as in (c). Curve represents expected noise figure based on measured losses. Dotted line represents the 0 dB quantum limit.
  • Figure 5: (a) Gain versus pump power for chip-scale OPAs in the literature (see Table \ref{['litreview_table']}). Green circles represent phase-insensitive amplification measurements and blue squares phase-sensitive amplification. Darker points correspond to data from this work. Curves are simulations using the same parameters as Section \ref{['secSHG']}. (b) Gain rate (at max reported power) vs loss rate of chip-scale OPAs. Dashed lines are lines of constant nonlinearity-to-loss ratio. Those references that did not include loss rate information could not be plotted.
  • ...and 8 more figures