Sound, as strange as it may sound, is quietly becoming one of the most promising information carriers in modern technology. While photons dominate long-distance communication and electrons run the world of computing, acoustic waves—vibrations rippling through solid materials at microwave frequencies—offer a compact, stable, and remarkably energy-efficient middle ground. A research team led by Linbo Shao at Virginia Tech, working with colleagues at the University of Virginia, has now reported a major step forward for this emerging field: a low-loss phononic integrated circuit platform built from thin-film silicon nitride patterned on a lithium niobate substrate. The work, published in Communications Engineering, demonstrates that gigahertz-frequency sound can be routed, split, stored, and even turned into a self-sustaining oscillator on a chip with losses low enough to matter for both classical and quantum applications.
To appreciate why this matters, it helps to understand what a phononic integrated circuit, or PnIC, actually is. Just as photonic integrated circuits guide light through nanoscale channels, phononic circuits guide mechanical vibrations—phonons—through engineered waveguides. At gigahertz frequencies, these vibrations have wavelengths measured in mere hundreds of nanometers, roughly five orders of magnitude shorter than electromagnetic waves of the same frequency. That enormous shrinkage factor means acoustic devices can pack resonators, filters, and delay lines into a footprint far smaller than any radio-frequency electromagnetic equivalent. This is precisely why acoustic filters already sit inside nearly every smartphone, converting crowded radio signals into clean channels. The catch has been that most existing acoustic platforms force engineers to choose: they can have low-loss waveguides, or high-quality resonators, or efficient transducers, but rarely all of these on one chip.
The Virginia Tech and University of Virginia collaboration set out to break that trade-off. Their ideal platform, as described in the paper, should simultaneously support four building blocks: low-loss acoustic waveguides, high-quality-factor resonators, high-performance modulators, and efficient electromechanical transducers. The material pairing they chose is chemically and mechanically elegant. Lithium niobate is a ferroelectric crystal famous for its strong piezoelectric properties, meaning it converts electrical signals into mechanical motion—and back again—with exceptional efficiency. It also possesses large electro-optic and acoustic velocities that make it a favorite in modern integrated photonics. Silicon nitride, by contrast, is one of the lowest-loss optical materials known, forms ultrasmooth films, and is compatible with standard semiconductor foundry processes. By patterning thin-film silicon nitride directly on lithium niobate, the team created a hybrid structure in which each material does what it does best: lithium niobate drives and senses the vibrations, while the carefully shaped silicon nitride confines and directs them.
The headline performance metric of the new platform is the propagation frequency-quality factor product of its acoustic waveguides, which reached 4.22 × 10^13. In the world of resonant and wave-based systems, the quality factor, or Q, measures how many oscillation cycles a wave or resonator completes before its energy dissipates. Multiplying Q by frequency yields a figure of merit that is largely independent of scale, allowing fair comparison across platforms. A frequency-Q product above 10^13 places these acoustic waveguides in an elite class of coherent information carriers, comparable to the best optical waveguides in relative terms. Practically, it means that a gigahertz sound wave traveling through the team’s circuits retains its energy over distances long enough to build complex, multi-element circuits—delays, filters, interferometers—without the signal fading into heat. For quantum applications, where a single phonon must survive long enough to be manipulated and read out, such low loss is not a luxury but a prerequisite.
Waveguides alone do not make a circuit, however; engineers also need ways to split and combine acoustic signals with minimal waste. The team demonstrated directional couplers—devices in which two waveguides run close enough that vibrational energy gradually transfers from one to the other through evanescent coupling, the same physics that lets a vibrating tuning fork set a neighboring fork ringing. Their couplers achieved controllable power transfer at an insertion loss of just 3.5 dB. Insertion loss quantifies how much signal is lost simply by passing through a component, and every decibel counts when signals are weak or when photons and phonons must be conserved at the quantum level. A 3.5 dB figure means that roughly 55 percent of the acoustic power survives the coupling process, a level the authors identify as a key enabler for routing phonons between resonators, delay lines, and transducers on the same chip.
Storage is the third pillar, and here the platform delivered ring resonators with a loaded quality factor of up to 17,925. An acoustic ring resonator is a closed loop of waveguide in which sound circulates, resonating at frequencies whose wavelengths fit an integer number of times around the ring. The loaded Q includes the coupling losses to the outside world, not just the intrinsic material loss, making it the number that matters in real systems. A loaded Q near 18,000 at gigahertz frequencies corresponds to phonons persisting for microseconds—a long time on the scale of microwave electronics. Long-lived acoustic resonances are the heart of the spectral processing that makes radio-frequency acoustic filters so selective, and they are equally the heart of proposed quantum memories, where mechanical modes would buffer quantum states between optical photons or superconducting qubits.
The team then assembled these building blocks into a showcase application: a 1-gigahertz phononic oscillator built around one of their ring resonators. An oscillator is the metronome of electronics—a device that produces a perfectly periodic signal used to time everything from radio transmitters to radar systems to quantum logic gates. The best electronic oscillators achieve their stability by referencing a high-Q mechanical or electromagnetic resonance, and the new phononic oscillator leverages the platform’s 17,925 loaded Q to spectacular effect. It reached a phase noise of −159.0 dBc/Hz at a 100-kilohertz offset from the carrier. Phase noise describes the rapid, random fluctuations in an oscillator’s timing; lower is better, and the decibel scale here means the noise power is 159 billion times smaller than the carrier power within a one-hertz bandwidth at that offset. That figure rivals the performance of much bulkier, power-hungry microwave oscillators, in a device etched onto a chip.
The implications ripple outward across several fields at once. For microwave acoustics, the platform offers a path toward integrated radio-frequency processors that are smaller, cleaner, and more stable than today’s surface-acoustic-wave and bulk-acoustic-wave components, which typically rely on suspended or bulky geometries. For quantum phononics, low-loss gigahertz phonons are a sought-after resource: they can couple to superconducting qubits through piezoelectric transduction, interface with solid-state defects such as those in diamond or silicon carbide, and bridge to optical photons in hybrid quantum networks. Lithium niobate’s piezoelectricity makes that electromechanical link natural, while silicon nitride’s optical excellence makes the photonic link equally promising. The authors point to integrated hybrid systems combining phonons, photons, superconducting qubits, and solid-state defects as the long-term destination—and this platform is designed from the outset to host all of them.
The research was conducted with support from several major United States funding programs, including the Air Force Office of Scientific Research, the Defense Advanced Research Projects Agency’s OPTIM, SynQuaNon, and GRYPHON programs, and fabrication was performed at the Center for Nanophase Materials Sciences, a Department of Energy Office of Science user facility. The work also reflects a growing trend in which Virginia’s university ecosystem—through the Commonwealth Cybersecurity Initiative—invests in the intersection of quantum devices and secure communications. The paper, received in March 2026 and accepted in September 2026, is open access, allowing engineers and physicists worldwide to build on its results immediately.
What makes this demonstration genuinely exciting is its completeness. Many platform papers showcase a single impressive component; this one delivers waveguides, couplers, resonators, and a working oscillator, all on the same material system, all at gigahertz frequencies, and all with numbers that stand up against the best specialized devices in each category. The phonon, long treated as the awkward cousin of the photon, now has an integrated circuit platform worthy of its ambitions. If the trajectory of silicon photonics is any guide, the coming years could see these sound-based circuits shrink radar front ends, sharpen quantum sensors, and serve as the mechanical backbone of hybrid quantum machines—proof that sometimes the future of computing hums at a frequency we cannot hear, in a material we cannot see, on a chip smaller than a fingernail.
Subject of Research: Low-loss gigahertz phononic integrated circuits on a silicon nitride–lithium niobate platform
Article Title: Low-loss phononic integrated circuits based on a silicon nitride-lithium niobate platform
Article References: Low-loss phononic integrated circuits based on a silicon nitride-lithium niobate platform. (n.d.). https://doi.org/10.1038/s44172-026-00795-1
Image Credits: AI Generated
DOI: 10.1038/s44172-026-00795-1
Keywords: phononics, integrated circuits, silicon nitride, lithium niobate, acoustic waveguides, ring resonators, quality factor, oscillators, phase noise, quantum phononics, microwave acoustics, piezoelectricity
News Source: Denise Maddox. (October 8, 2026). Sound Chips Go the Distance: Silicon Nitride Meets Lithium Niobate for Low-Loss Phononic Circuits. Scienmag.
