Broadband Spatio-Spectral Mode Conversion via Four-Wave Mixing

Technical Summary: Broadband Spatio-Spectral Mode Conversion via Four-Wave Mixing

Problem Statement
Quantum networks require the integration of diverse qubit platforms, many of which operate in the visible spectrum (e.g., color centers in diamond), with low-loss transmission channels in the infrared telecom band. Current solutions for Quantum Frequency Conversion (QFC) typically treat spectral conversion and spatial out-coupling as separate functions, often requiring complex structured pump illumination, Spatial Light Modulators (SLMs), or separate coupling stages. Furthermore, integrating these processes into scalable, heterogeneous quantum repeater nodes remains challenging due to the need for multiple interfaces and fabrication steps. The authors identify a need for a unified, scalable architecture that bridges visible emitters to the infrared spectrum while minimizing system-level losses and fabrication complexity.

Methodology
The paper proposes a framework for scalable, broadband, free-space phase-matched four-wave mixing (FFWM) within a diamond ring resonator. The core innovation lies in combining spectral conversion and spatial extraction into a single nonlinear optical process.

• Physical Mechanism: The system utilizes a $\chi^{(3)}$ Bragg-scattering four-wave mixing (BS-FWM) interaction. Two pump photons (one clockwise, one counter-clockwise) interact with a signal photon from a visible color center (e.g., SnV) to generate an idler photon in the infrared.
• Phase Matching Strategy: The authors engineer the in-plane angular momenta of the interacting waves such that the induced nonlinear polarization $P^{(3)}$ drives an idler mode with azimuthal number $m_{idl} = 0$. This specific mode has zero in-plane momentum, causing it to radiate normally (out-of-plane) from the ring resonator. This eliminates the need for separate output couplers or SLMs; the nonlinear interaction itself acts as the output coupler.
• Device Architecture: The device is a diamond-on-insulator ring resonator coupled to a silicon nitride bus waveguide. A back reflector breaks emission symmetry, directing the idler photons into free space for collection by an objective lens. The geometry is fixed, supporting multiple discrete resonant pump combinations without thermal tuning.
• Modeling Framework: The authors develop an end-to-end theoretical model to calculate single-photon coupling efficiency ($\eta$). This model integrates:
  • Non-radiative decay and Zero-Phonon Line (ZPL) branching ratios of the emitter.
  • Purcell-enhanced loading of the signal mode.
  • Pump-dependent signal-idler coupling rates.
  • Engineered radiation rates of the low-Q idler mode.
  • Free-space collection efficiency (including S-waveplate transformation and Gaussian beam overlap).
  • The system is analyzed using a non-Hermitian effective Hamiltonian and verified against Qutip Monte Carlo simulations.

Key Contributions

1. Unified Spatio-Spectral Interface: The design converts a cavity-bound visible photon directly into a deliberately radiative infrared mode within a single nonlinear process, removing the need for structured pump illumination or separate spatial out-coupling elements.
2. Broadband Operation: A fixed ring geometry supports multiple discrete pump combinations that satisfy phase-matching conditions across a spectral window of approximately 165 nm centered near the telecom O-band.
3. Fabrication-Ready Pathway: The work outlines a pathway using diamond-on-insulator technology and heterogeneous integration (diamond chiplets), compatible with current fabrication capabilities.
4. Comprehensive Efficiency Analysis: The paper presents a detailed system-level model that quantifies the probability of retrieving a telecom photon from a single color-center decay, explicitly accounting for emitter non-radiative losses and the trade-offs between pump power, cavity quality factor ($Q$), and the $\beta$ factor (radiative ZPL emission probability).

Results

• Spectral Range: Simulations indicate that a single ring can support in-plane phase-matched solutions spanning over 500 nm in the infrared, with the out-of-plane phase-matching condition (constructive interference) fulfilled for a subset of approximately 165 nm in the O-band.
• Spatial Efficiency: The far-field emission profile of the $m=0$ idler mode is radially polarized. After transformation by an S-waveplate, the mode achieves a maximum spatial power overlap of 24% with a Gaussian beam (assuming 0.82 numerical aperture). With coated optics, the spatial efficiency $\eta_{spatial}$ is calculated at 0.21.
• System Efficiency: For a SnV center (with a bulk radiative factor $r_{ZPL} \approx 0.48$) and a cavity quality factor $Q_{cav} = 10^5$, the system achieves a saturated idler retrieval efficiency ($\eta_{idler}$) of 85.3% with a pump budget of 15.2 W. The total end-to-end efficiency from the color center to an optical fiber is estimated at $\eta \approx 0.18$.
• Optimization Limits: The analysis reveals that efficiency is limited by the $\beta$ factor. As pump power increases to drive conversion, the effective quality factor of the signal mode decreases, eventually causing lossy decay to dominate. The maximum achievable efficiency is strictly bounded by the emitter's intrinsic radiative properties and the cavity's ability to enhance the ZPL emission.
• Noise Suppression: Competing nonlinear noise processes are shown to be negligible due to phase-mismatching and frequency detuning relative to the cavity modes.

Significance and Claims
The authors claim this work represents a step forward in analyzing and reducing system-scale losses in spin-photon interfaces. By integrating the spatial and spectral interfaces into one nonlinear process, the proposed architecture simplifies the coupling of emitters to quantum networks, potentially reducing the number of fabrication steps and photonic interfaces required.

The paper emphasizes that the design does not rely on lossless modes but rather on the strategic transfer of losses: the "loss" of the idler mode (radiation out of the cavity) is engineered to be the desired output channel. This approach allows for modular unit cells that natively bridge visible color centers to the infrared, offering a scalable alternative to bulky fiber platforms or SLM-dependent systems. The authors conclude that with attainable quality factors ($Q \sim 10^5$) and reasonable pump powers (achievable with pulsed quasi-CW lasers), this method offers a viable path toward efficient, scalable quantum networking building blocks.