A quantum network consisting of computational nodes connected by high-fidelity communication channels could expand information-processing capabilities beyond those of classical networks. Superconducting qubits hold promise for scalable quantum computation at microwave frequencies, but must operate in an isolated cryogenic environment; meanwhile, quantum communication over long distances has been demonstrated with optical photons. A fast, quantum-coherent interface between the two would be a key element of a large-scale quantum network or distributed quantum computer.In this thesis, we describe the theoretical basis as well as the practical design and development of a device incorporating a silicon optomechanical nanobeam and an aluminum-nitride-based electromechanical transducer. We find that this class of device has the potential to approach ideal quantum microwave-optical transduction. Finally, we experimentally demonstrate classical, continuous-wave operation of such a device with internal conversion efficiencies above 1%. This device also has a larger bandwidth than previous efficient microwave-optical transducers, allowing us to operate in the time domain with 20 ns pulses.
A quantum network consisting of computational nodes connected by high-fidelity communication channels could expand information-processing capabilities beyond those of classical networks. Superconducting qubits hold promise for scalable quantum computation at microwave frequencies, but must operate in an isolated cryogenic environment; meanwhile, quantum communication over long distances has been demonstrated with optical photons. A fast, quantum-coherent interface between the two would be a key element of a large-scale quantum network or distributed quantum computer.In this thesis, we describe the theoretical basis as well as the practical design and development of a device incorporating a silicon optomechanical nanobeam and an aluminum-nitride-based electromechanical transducer. We find that this class of device has the potential to approach ideal quantum microwave-optical transduction. Finally, we experimentally demonstrate classical, continuous-wave operation of such a device with internal conversion efficiencies above 1%. This device also has a larger bandwidth than previous efficient microwave-optical transducers, allowing us to operate in the time domain with 20 ns pulses.
Superconducting circuits are important technological tools that have found various applications, notably hypothetical future superconducting quantum computers. Scaling such systems further is, however, proving challenging because of thermal heat load and space constraints in cryostats. One way of improving the scalability of such systems would be to resort to optical fibers for controlling the circuits in cryostats, as well as mediating long-distance interaction between them while preserving quantum coherence. Such a strategy relies on building a coherent interface between GHz frequencies and optical frequencies. This interface, referred to as "microwave-optical transducer", would need to exhibit high photon conversion efficiencies and very low added noise, ideally less than a quantum, in a cryogenic environment. Since the prospects of weaving networks of optically connected superconducting quantum processors have now attracted the attention of the scientific community for more than a decade, very efficient transducers have already been proposed and realized. However, despite their impressive performances against selected metrics, often equivalent to optimizing the on-chip conversion efficiency alone, none of the approaches pursued so far really solved the qubit interconnection problems in ways that could be applied to large-scale systems in the near future. The aim of this thesis was to develop transduction approaches that would ultimately enable optical driving of superconducting circuits and quantum state transfer between microwave and optical frequencies, while at the same time ensuring good scalability, both in terms of wafer manufacturability and integration in cryogenic environments. Integrated photonic circuits based on silicon nitride appeared as a promising platform, as silicon nitride is a CMOS-certified material, has a transparency window compatible with optical fibers and lacks two photon absorption. Realizing the coupling between microwave and optical frequencies in a way similar to what is done in optomechanics is desirable, because the field of optomechanics has seen the establishement of techniques that enable control of the low frequency degree of freedom at the quantum mechanical level, including ground state cooling, control of the susceptibility by dynamical backaction, quantum state transfer and entanglement. The heterogeneous integration of nonlinear crystalline thin films, such as lithium niobate and lithium tantalate, onto silicon nitride photonic circuitry endows it with an electro-optic activity that emulates the optomechanical interaction, while offering an inherent access to the microwave degree of freedom. However, material nonlinearities, including photorefraction, charge screening, and hysteresis, introduce complications that motivate exploring alternative approaches. Here, alongside an electro-optic implementation, a transducer that leverages the interaction of bulk acoustic resonances excited by micro-electromechanical systems with whispering-gallery-like modes of microresonators is presented. Coherent transduction between microwave and optical fields is demonstrated, and optomechanical dynamical backaction is observed. The transducers fast bandwidth, compatibility with multiplexing and ability to down-convert microwaves suggest interesting prospects, particularly upon further improvements of the optomechanical coupling.
The development of scalable quantum networks requires coherent interfaces capable of converting microwave photons used in superconducting quantum processors into optical photons suitable for long-distance fiber transmission. This review surveys recent progress in microwave-to-optical quantum transduction across optomechanical, electro-optic, and magneto-optic platforms, with emphasis on conversion efficiency, bandwidth, added noise, and operating temperature. In addition to standard metrics, we propose the internal efficiency eta_in and the magnon decay rate kappa_m/2pi as normalized parameters that enable fairer comparison across heterogeneous implementations. Optomechanical systems achieve internal phonon-to-photon efficiencies of 93% with sub-quantum added noise of 0.25 quanta at millikelvin temperatures. Electro-optic devices based on LiNbO3 and AlN have advanced from room-temperature efficiencies below 1% to millikelvin systems with internal efficiencies approaching 99.5%, added noise as low as 0.16 quanta at 60 mK, and bandwidths extending to several tens of megahertz. Magneto-optic (optomagnonic) platforms exhibit the lowest efficiencies (typically 10−10 to 10−8), but offer intrinsic non-reciprocity and broadband magnonic operation, with emerging approaches based on topological heterostructures and magnon squeezing predicting enhancements up to 10−4. Optomechanical systems appear promising for high-fidelity quantum state transfer, electro-optic transducers for high-bandwidth coherent links, and magneto-optic devices for non-reciprocal network components. We discuss the fundamental trade-off between efficiency and added noise across all three platforms, and argue that heterogeneous microwave-optical transduction is emerging as a key enabling technology for distributed quantum computing and large-scale quantum networks.
A quantum network that distributes and processes entanglement would enable powerful new computers and sensors. Optical photons with a frequency of a few hundred terahertz are perhaps the only way to distribute quantum information over long distances. Superconducting qubits on the other hand, which are one of the most promising approaches for realizing large-scale quantum machines, operate naturally on microwave photons that have roughly 40,000 times less energy. To network these quantum machines across appreciable distances, we must bridge this frequency gap and learn how to generate entanglement across widely disparate parts of the electromagnetic spectrum. Here we implement and demonstrate a transducer device that can generate entanglement between optical and microwave photons, and use it to show that by detecting an optical photon we add a single photon to the microwave field. We achieve this by using a gigahertz nanomechanical resonance as an intermediary, and efficiently coupling it to optical and microwave channels through strong optomechanical and piezoelectric interactions. We show continuous operation of the transducer with 5% frequency conversion efficiency, and pulsed microwave photon generation at a heralding rate of 15 hertz. Optical absorption in the device generates thermal noise of less than two microwave photons. Joint measurements on optical photons from a pair of transducers would realize entanglement generation between distant microwave-frequency quantum nodes. Improvements of the system efficiency and device performance, necessary to realize a high rate of entanglement generation in such networks are within reach.
Abstract Optical quantum networks can connect distant quantum processors to enable secure quantum communication and distributed quantum computing. Superconducting qubits are a leading technology for quantum information processing but cannot couple to long-distance optical networks without an efficient, coherent, and low noise interface between microwave and optical photons. Here, we demonstrate a microwave-to-optical transducer using an ensemble of erbium ions that is simultaneously coupled to a superconducting microwave resonator and a nanophotonic optical resonator. The coherent atomic transitions of the ions mediate the frequency conversion from microwave photons to optical photons and using photon counting we observed device conversion efficiency approaching 10 −7 . With pulsed operation at a low duty cycle, the device maintained a spin temperature below 100 mK and microwave resonator heating of less than 0.15 quanta.
The first quantum revolution has brought us the classical Internet and information technology. Today, as technology advances rapidly, the second quantum revolution quietly arrives, with a crucial moment for quantum technology to establish large-scale quantum networks. However, solid-state quantum bits (such as superconducting and semiconductor qubits) typically operate in the microwave frequency range, making it challenging to transmit signals over long distances. Therefore, there is an urgent need to develop quantum transducer chips capable of converting microwaves into optical photons in the communication band, since the thermal noise of optical photons at room temperature is negligible, rendering them an ideal information carrier for large-scale spatial communication. Such devices are important for connecting different physical platforms and efficiently transmitting quantum information. This paper focuses on the fast-developing field of optomechanical quantum transducers, which has flourished over the past decade, yielding numerous advanced achievements. We categorize transducers based on various mechanical resonators and discuss their principles of operation and their achievements. Based on existing research on optomechanical transducers, we compare the parameters of several mechanical resonators and analyze their advantages and limitations, as well as provide prospects for the future development of quantum transducers.
Optical quantum networks provide a natural route for connecting distant superconducting quantum processors, enabling distributed quantum computation, sensing, and communication. Piezo-optomechanical transducers are among the leading candidates for scalable microwave-to-optical quantum interfaces. However, prior one-dimensional piezo-optomechanical transducers remain limited by optical-absorption-induced heating and the resulting thermal noise. Two-dimensional optomechanical crystals offer substantially improved thermalization from better thermal anchoring, but their structural complexity has so far hindered the realization of a fully integrated two-dimensional transducer. Here, we overcome this challenge with a new design strategy based on band structure engineering. We fabricate the devices and experimentally characterize the response, measuring an electromechanical damping rate of 4.6 kHz at room temperature and electromechanical coupling rate of 0.17 MHz by wire-bonding to a multi-mode microwave resonator at 10 mK. Bidirectional transduction is performed with a calibrated internal efficiency of 0.85\% in the continuous-wave operation, alongside pulsed photon-phonon pair generation near its quantum ground state. Our results represent a significant step toward high-efficiency and low-noise transducers for entangling remote superconducting qubits.