Microwave-to-Optical Conversion in Quantum Systems

• Microwave-to-optical conversion is the coherent mapping of GHz microwave photons to THz optical photons while preserving key quantum properties.
• It employs diverse mechanisms such as electro-optic modulation, optomechanical coupling, and atomic nonlinearities, each addressing unique efficiency and noise challenges.
• These conversion techniques underpin hybrid quantum networks, advanced sensing, and broadband communications through innovative device architectures and material integration.

Microwave-to-Optical Conversion encompasses the set of physical and engineering methodologies that facilitate the coherent transfer of information and quantum states between the microwave (GHz) and optical (THz) domains. This capability underpins hybrid quantum networks, quantum interconnects for superconducting processors, advanced sensors, and wideband signal processing platforms. Realizing efficient, noise-minimal, and bandwidth-flexible microwave-to-optical transducers is a central challenge, as the processes must bridge vastly different photon energies, overcome disparate loss and noise mechanisms, and often contend with severe engineering and integration constraints.

1. Physical Principles and Platform Taxonomy

The essential task in microwave-to-optical conversion is the coherent, bidirectional mapping of microwave photons (typically 1–10 GHz) into optical photons (often near-infrared, 1550 nm) with the preservation of amplitude, phase, and quantum statistics. This is accomplished by engineering physical systems that admit strong, controllable coupling between electromagnetic fields at these different frequencies, often mediated by one or more intermediate quantum excitations.

The main transduction mechanisms are:

1. Electro-Optic (EO) Modulation: Utilizes the Pockels effect in χ^2 materials (e.g., LiNbO₃, GaP) to directly modulate an optical resonance’s frequency with a microwave field, often within whispering-gallery-mode (WGM) or nanophotonic cavities (Soltani et al., 2017, Rueda et al., 2016, Javerzac-Galy et al., 2015, McKenna et al., 2020).
2. Opto(magneto)mechanical Coupling: Couples both microwave and optical fields to a mechanical (phononic) mode via radiation pressure or piezoelectric (PE) effects, allowing for beam-splitter–type interactions (Andrews et al., 2013, Forsch et al., 2018, Stockill et al., 2021, Zhao et al., 2024, Shao et al., 2019, Hönl et al., 2021).
3. Atomic (Rydberg or Λ-type) Nonlinearities: Harnesses high-lying Rydberg states in room-temperature or cold atomic ensembles to facilitate multi-wave mixing (six-wave, four-wave, sum-frequency) among microwave and optical fields (Borówka et al., 2023, Han et al., 2017, Vogt et al., 2018, Smith et al., 2023, Petrosyan et al., 2019, Gard et al., 2017, Petrosyan et al., 2019).
4. Magnon-Based Magneto-Optic and Cavity Optomagnonic Approaches: Utilizes magnon (spin wave) modes in ferrimagnetic materials (such as YIG spheres) strongly coupled to both microwave fields and optical cavities (Chai et al., 2021, Wu et al., 2024).

Each class has distinct strengths, integration constraints, thermal requirements, and operational bandwidths.

2. Experimental Implementations and Device Architectures

Device architectures span the atomic, solid-state photonic, electro-optomechanical, and magneto-optic domains:

• Rydberg-Atom Converters: Free-space six-wave mixing in warm vapor cells or cold atom clouds links microwave transitions between Rydberg states to optical emissions (e.g., 776 nm for ⁸⁵Rb) (Borówka et al., 2023, Vogt et al., 2018, Han et al., 2017). Cell-based implementations enable room-temperature operation, high bandwidth (up to 59 MHz), and ultra-wide dynamic range.
• Integrated Nanophotonic Transducers: On-chip optomechanical resonators in piezoelectric, low-loss materials (e.g., GaP, Si, TFLN, LN) mediate phonon-photon coupling (Stockill et al., 2021, Zhao et al., 2024, Hönl et al., 2021, Shao et al., 2019, Yang et al., 12 Sep 2025). These can support microwave drive via electrodes or inductors and readout via fiber or waveguide coupling.
• Electro-Optic Resonator Converters: EO (Pockels) effect-based devices in LiNbO₃ exploit triply-resonant conditions, monolithic integration with superconducting circuits, and doublet (split) optical resonances to enhance conversion and suppress noise (Soltani et al., 2017, Javerzac-Galy et al., 2015, Rueda et al., 2016, McKenna et al., 2020).
• Phase-Matched Traveling-Wave Platforms: Novel architectures leveraging continuous (non-cavity) phase matching within hybrid photonic-phononic waveguides in TFLN offer exceptional channel count, >250 MHz microwave and >40 nm optical bandwidths, and multi-channel operation in a single device (Yang et al., 12 Sep 2025).
• Magnetostatic and Magnonic Converters: Utilization of strong magnon-photon and magnon-microwave coupling in ferrimagnetic WGM or Fabry–Pérot cavities (YIG spheres/flakes) offers broadband frequency conversion (tuning >2.5 GHz) albeit with low absolute efficiency due to presently weak optomagnonic coupling (Chai et al., 2021, Wu et al., 2024).

3. Performance Metrics: Efficiency, Bandwidth, and Noise

Key performance benchmarks for microwave-to-optical converters are:

• Photon-Conversion Efficiency ($\eta$): Defined as the ratio of detected optical photon rate to incident microwave photon rate, corrected for port and coupling losses. State-of-the-art atomic converters achieve $\eta$ up to 3–5% in room-temperature Rydberg vapor (Borówka et al., 2023), with theoretical prospects $>$60% in optimized cold-atom or cavity configurations (Vogt et al., 2018, Han et al., 2017, Petrosyan et al., 2019). Mechanically mediated chip-based transducers (GaP, Si, TFLN, LN) report internal efficiencies from $6.8\times10^{-8}$ (GaP, pulsed) (Stockill et al., 2021) to 2.2% (TFLN, traveling wave, multi-channel) (Yang et al., 12 Sep 2025) and 2.2% (Si, continuous, $n_{\rm add}<1$) (Zhao et al., 2024).
• Bandwidth: Instantaneous conversion bandwidths span $\sim$16–59 MHz for Rydberg converters (Borówka et al., 2023), $>$20 MHz in integrated EO and optomechanical devices (McKenna et al., 2020), and $>$250 MHz for traveling-wave TFLN systems (Yang et al., 12 Sep 2025). Bandwidths are limited by mechanical, optical, or microwave Q, and in atomic systems by optical or polariton linewidths.
• Dynamic Range and Multi-Channel Capability: Warm-vapor Rydberg converters sustain dynamic ranges $>$57 dB, detecting from single-photon to $>10^6$ photon/s rates (Borówka et al., 2023). Multiplexed architectures enable parallel operation across numerous frequency bins (Smith et al., 2023, Yang et al., 12 Sep 2025).
• Noise Figures: Input-referred added noise is often benchmarked. State-of-the-art silicon optomechanical devices attain $n_{\rm add}=0.58$–0.94 at sub-percent efficiency in continuous wave operation (Zhao et al., 2024). Pulsed mechanical systems demonstrate sub-phonon added noise in GaP and GaAs (Stockill et al., 2021, Forsch et al., 2018). Rydberg-atom vapor systems reach noise-equivalent temperatures $T_{\rm NE}=3.8$ K (with filtering) and can resolve quantum photon statistics at room temperature (Borówka et al., 2023).
• Quantum Coherence: Multiple platforms preserve phase and amplitude of input fields as evidenced by $g^{(2)}(\tau)$ autocorrelation and Mach–Zehnder interferometry, establishing suitability for quantum information protocols (Borówka et al., 2023, Forsch et al., 2018).

Platform/Class	Best Reported $\eta$	Bandwidth (FWHM)	$n_{\rm add}$ or $T_{\rm NE}$	Key Reference
Rydberg vapor (hot)	3.1%	16–59 MHz (tunable)	$T_{\rm NE}=3.8$ K	(Borówka et al., 2023)
Rydberg (cold)	$\sim 0.3$–5% (exp) <br> $\sim60$% (th.)	4+ MHz <br> 100 kHz (Yb)	$n_{\rm add}\ll1$ (theory)	(Vogt et al., 2018, Petrosyan et al., 2019)
TFLN Traveling Wave	2.2% (internal)	$>250$ MHz (μw)	MHz-scale channel, multi-ch.	(Yang et al., 12 Sep 2025)
GaP OMC (pulsed)	$6.8\times 10^{-8}$ (sys)	67–370 kHz (tun.)	$n_{\rm th} = 0.55$	(Stockill et al., 2021)
Si Nanobeam OMC	2.2% (ctns), $n_{\rm add}=0.94$	88.9 kHz (ctns), $\eta B$ = $1.9$ kHz	$n_{\rm add}<1$	(Zhao et al., 2024)
YIG WGM magnon	$\sim 3.6 \times 10^{-6}$	2.5 GHz tuning range	Not quantum-limited	(Chai et al., 2021)
LiNbO₃ WGM EO SSBC	0.1%	$>1$ MHz	Near quantum-limited	(Rueda et al., 2016)

4. Noise Mechanisms and Strategies for Quantum-Limited Operation

The principal noise sources vary by platform:

• Thermal Phonon Occupancy: In optomechanical and mechanical transducers, ground-state operation ($n_{\rm th}<1$) is required for quantum-coherent conversion. This mandates cryogenic cooling (<100 mK) or radiative cooling schemes (Zhao et al., 2024, Stockill et al., 2021, Forsch et al., 2018).
• Spontaneous Emission and Fluorescence: In atomic platforms, fluorescence from Rydberg or intermediate states imposes a noise floor; detuned multi-photon processes and narrowband filtering are employed to suppress this (Borówka et al., 2023, Vogt et al., 2018, Petrosyan et al., 2019).
• Johnson and Circuit Noise: In solid-state systems, circuit impedance mismatches and electrical noise contribute. High-impedance matching and filtering are critical (Stockill et al., 2021, Yang et al., 12 Sep 2025).

Noise-equivalent temperature ($T_{\rm NE}$) and added quanta $n_{\rm add}$ are quantitative figures of merit. Strategies for single-quantum-level sensitivity include:

• Pulsed operation to limit heating (mechanical systems) (Stockill et al., 2021, Forsch et al., 2018).
• Microfabricated vapor cells or integrated hollow-core fibers for enhanced atomic field overlap and reduced background (Borówka et al., 2023).
• Narrowband optical filtering and phase-locked detection (Borówka et al., 2023, Zhao et al., 2024).

5. Applications and Integration Prospects

Microwave-to-optical conversion underpins several emerging quantum and classical technologies:

• Quantum Interconnects: Up-conversion of superconducting qubit microwave photons to telecom-band optical photons enables long-distance, fiber-based entanglement distribution for networked quantum information processing (Borówka et al., 2023, Soltani et al., 2017, Andrews et al., 2013).
• Quantum Sensing and Bolometry: Photon-counting of thermal microwave backgrounds yields high sensitivity for microwave detection, radio astronomy, and quantum sensors (Borówka et al., 2023).
• Multi-Channel and Frequency-Bin Quantum Processing: Frequency-division multiplexing (FDM) and frequency-bin qubit mapping are realized in vapor cells and traveling-wave photonic-phononic platforms, supporting massive parallelism (Smith et al., 2023, Yang et al., 12 Sep 2025).
• Microwave Photonic Links and Classical Communications: High linearity, low-noise, and wideband conversion enable photonic readout and processing of classical signals from radar, communications, or control electronics (Shao et al., 2019, Yang et al., 12 Sep 2025).

6. Outlook and Future Directions

Progress in microwave-to-optical conversion is driven by advances in material integration, system engineering, and quantum control techniques. Prominent directions include:

• Enhanced Field Confinement: Implementing vapor cells in MW waveguides or cavities, shrinking optomagnonic mode volumes, and exploiting super-inductors for impedance matching are essential for boosting single-photon conversion efficiency (Borówka et al., 2023, Wu et al., 2024, Stockill et al., 2021).
• Scalable Integration: Microfabricated atomic cells, all-on-chip nanophotonic circuits, and CMOS-compatible EO platforms promise integration with classical and quantum hardware (Yang et al., 12 Sep 2025, Zhao et al., 2024, Javerzac-Galy et al., 2015).
• Noise Minimization and Quantum Regime Protocols: Ground-state cooling, pulsed and continuous quantum-enabled operation ($n_{\rm add}<1$), and bidirectionality are converging in multiple state-of-the-art devices (Zhao et al., 2024, Borówka et al., 2023, Stockill et al., 2021).
• Multi-Channel and Multiplexed Architectures: Continuous phase-matched, traveling-wave designs in TFLN and multiplexed atomic systems are establishing the feasibility of 10–100 channel quantum interfaces within a single chip or cell (Smith et al., 2023, Yang et al., 12 Sep 2025).
• Hybrid and Co-Integrated Quantum Systems: Direct interfacing with superconducting qubits, atomic memory networks, and fiber-optic links is a focal point; coherent, bidirectional, and phase-preserving conversion is critical for quantum internet architectures (Andrews et al., 2013, Soltani et al., 2017, Borówka et al., 2023).

The field is now approaching the performance and scalability required for quantum network deployment, subject to further advances in efficiency, noise engineering, and system integration.