Quantum Transduction: Research Directions

Quantum transduction converts a signal between physical carriers while retaining the quantum information required by a task. For microwave–optical links, the central research question is whether conversion can preserve useful nonclassical correlations at adequate rate without damaging the processor. Classical signal conversion, qubit readout and remote entanglement are distinct milestones.

Required background. Quantum transduction develops the channel and noise conventions.

Efficiency and added noise must share a convention

An ideal passive converter has an input–output relation

$$a_{\mathrm{out}}=\sqrt{\eta}\,a_{\mathrm{in}}+\sqrt{1-\eta}\,e,$$

where the environment mode $e$ preserves commutation relations. Thermal occupation adds noise; active gain changes the channel and requires additional noise operators. State whether reported noise is input-referred or output-referred, integrated over which bandwidth, and measured at which pump power.

Internal conversion, fiber-to-fiber efficiency and qubit-to-detected-photon probability are different metrics. Optical filtering, extraction loss, duty cycle and detector noise belong in an end-to-end account.

Approaches and selected evidence

Electro-optic, optomechanical and atom-mediated schemes trade coupling strength, pump heating, bandwidth and thermal occupation. Lauk and colleagues review these mechanisms. Mirhosseini and colleagues demonstrated conversion of a transmon excitation into an optical photon using a mechanical intermediary; this is a specific quantum-source conversion result, not a deterministic high-fidelity network channel. Perspective, experiment.

Established: conversion mechanisms and bounded quantum experiments. Active: simultaneous efficiency, low noise, bandwidth and compatibility with a working processor. Conjectural: useful remote fault-tolerant gates inferred from a classical conversion measurement alone.

Open questions

Can pumps be isolated without losing the conversion rate? How does heating evolve during repeated operation? Which heralded protocol tolerates low efficiency but rejects conversion noise? Can bidirectional conversion be calibrated consistently enough to preserve phase and indistinguishability between remote devices?

Graduate project: map the useful-channel region

Implement a Gaussian loss/noise channel model and propagate one arm of an entangled input through it. Compute a valid entanglement witness or covariance-based criterion for the declared Gaussian states, then include extraction and filtering losses. Deliver an efficiency–noise diagram and compare it with experimentally reported parameter ranges. A laptop linear-algebra calculation suffices.

Check vacuum and pure-loss limits analytically. Failure tests change noise-reference conventions, add asymmetric bandwidth and include pump duty cycle. Do not extend a Gaussian witness to arbitrary non-Gaussian states without proof. Research begins when a protocol or device change moves a measured operating region across an application-specific boundary, such as positive certified entanglement rate after all losses.

Entry points and review horizon

Use the perspective to choose a mechanism and the experiment to understand a complete conversion chain. Literature checked on 11 October 2026; selected evidence is from 2020 and does not claim current efficiency records. New results should update the full channel budget, not only one headline metric.

References

• N. Lauk et al., “Perspectives on quantum transduction,” Quantum Science and Technology 5, 020501 (2020). Preprint.
• M. Mirhosseini, A. Sipahigil, M. Kalaee and O. Painter, “Superconducting qubit to optical photon transduction,” Nature 588, 599–603 (2020). Article.