Ultrabroadband sound control with deep-subwavelength plasmacoustic metalayers

The study addresses the long-standing challenge of controlling audible sound over broad bandwidths with compact, subwavelength structures. Passive acoustic solutions are constrained by causality-based sum rules that impose trade-offs between device thickness and bandwidth, making low-frequency control inherently bulky. Narrowband resonant metamaterials achieve strong effects but sacrifice bandwidth. Active approaches with electrodynamic or piezoelectric transducers improve reconfigurability but remain limited by the inertia of mechanical elements and stability/energy constraints away from resonance. The authors propose leveraging the non-inertial dynamics of ultrathin layers of air plasma to realize active plasmacoustic metalayers capable of broadband (>2 decades) sound manipulation with thicknesses down to λ/1000, aiming for perfect absorption and tunable reflection without relying on inertial resonances.

Prior work on acoustic metamaterials and metasurfaces has enabled subwavelength control but typically via narrowband resonances, limiting bandwidth for applications such as perfect absorption, asymmetric transmission, and wave steering. Passive absorbers (porous media, resonators) are inefficient below 1 kHz or exhibit band-limited performance. Active resonators and electro-/piezo-actuated systems can broaden operation and add tunability but are fundamentally constrained by transducer inertia and stability when driven far from resonance. Corona discharge transducers have been studied for sound generation and active absorption but not by exploiting their intrinsic monopolar (heat) and dipolar (force) source physics for broadband impedance control. This work builds on Townsend’s discharge model and prior plasma electroacoustic actuators to establish a broadband, non-inertial active approach.

Concept and modeling: The plasmacoustic metalayer comprises two electrodes separated by a deep-subwavelength air gap: a thin-wire emitter biased at high positive DC voltage and a grounded grid collector. A controlled positive corona discharge forms around the emitter, producing drifting ions that impart a body force on the neutral air (dipolar acoustic source) and inelastic processes that release heat (monopolar source). Using Townsend’s model for the discharge I = C U (U − U0), with U = UDC + UAC sin(ωt), the linearized AC components of force and heat are frequency-independent: F = (C d/μ)(2UDC − U0) UAC and H = C (3UDC^2 − 2UDC U0) UAC, where d is inter-electrode spacing and μ the effective ion mobility. These sources are inserted into 1D linear acoustic equations (continuity, momentum, energy) to yield closed-form expressions for pressure and particle velocity contributions from monopolar and dipolar sources in a duct with termination impedance Z at distance l. Active control law: A microphone placed at position −x0 in front of the metalayer senses the total pressure. The control objective is to impose a target acoustic impedance Ztg (e.g., Zc for perfect absorption) at −x0 by driving UAC via a feedback controller implementing a transfer function θ(ω) = UAC/p(−x0, ω). Using superposition of external (incident + reflected) fields and plasmacoustic sources, and the known relation between pressure and velocity through Zac in the passive duct, the authors derive θ(ω) (detailed expressions in Methods), which is then approximated as a stable, proper rational function θ(s) for real-time implementation. Experimental setup: A prototype actuator (active area 50×50 mm²) employs a 0.1 mm nichrome wire emitter arranged in five parallel runs (10 mm spacing) and a perforated stainless-steel grid collector (2% flow resistance), separated by d = 6 mm. The actuator is backed by a rigid enclosure located l = 15 mm from the metalayer center (total thickness ~31 mm with enclosure). A stable corona is produced at UDC = 8 kV; discharge parameters (C, U0, μ) are estimated from measured I–V curves and literature (humidity 50–55%). The system is tested in a square impedance tube (50×50 mm², length 1.1 m) under normal incidence with swept-sine excitation (20–2000 Hz, 1 Pa incident amplitude). A control microphone is placed 10 mm in front of the metalayer; two additional microphones along the duct enable impedance and absorption estimation per ISO 10534-2. The control θ(s) is implemented on a Speedgoat IO-334 real-time platform at 50 kHz; its output (scaled) is combined with the DC bias and fed to a TREK 615-10 high-voltage amplifier (×1000) driving the electrodes. Generalized reflection control: The target impedance is extended to Ztg = (β Zc + j Zc tan(k(l + x + Δl)))/(Zc + j β Zc tan(k(l + x + Δl))), enabling independent control of reflection magnitude via β and reflection phase (time delay) via Δl, effectively creating a virtual termination with adjustable impedance and virtual thickness behind the metalayer.

• Ultrabroadband perfect absorption: With active control targeting Ztg = Zc, measured absorption is a > 0.98 between 40 and 1940 Hz and a > 0.94 from 20 Hz to 2 kHz in an impedance tube, spanning two decades with a device thickness of ~31 mm. This addresses wavelengths up to ~17 m (at 20 Hz), corresponding to deep-subwavelength operation (down to ~λ/1000).
• Passive vs active performance: In the passive (no-bias/control) state, the structure is acoustically transparent and absorption is low (a ≈ 0.1, nearly flat over 20–2000 Hz), dictated by the rigid termination. Comparable-thickness porous material and resonator models underperform at low frequencies compared to the active plasmacoustic absorber.
• Tunable reflection (magnitude): By setting β to achieve target reflectances, the system demonstrates broadband reflection magnitudes of approximately 90%, 70%, and 40% while keeping the reflection phase near zero across the band.
• Tunable reflection (phase/time delay): For fixed reflection magnitude (≈40%), adjusting Δl produces linear reflection phase roll-off corresponding to constant added time delays of 0.22 ms, 0.56 ms, and 0.85 ms, equivalent to virtual elongations of 3.7 cm, 9.7 cm, and 14.7 cm, respectively.
• Non-inertial, broadband actuation: The frequency-independent linearized source terms (monopolar heat, dipolar force) and absence of mechanical inertia enable a broadband control law, yielding bandwidth/size ratios at least three orders of magnitude larger than conventional solutions, and allowing seamless, transparent passive behavior when not actuated.

The findings validate that controlling plasma-induced monopolar (heat) and dipolar (force) sources within a deep-subwavelength air gap enables direct, broadband manipulation of the acoustic impedance without introducing mechanical interfaces. By matching the surface impedance to that of air, the system absorbs nearly all incident sound over two decades, overcoming passive causality bounds and the bandwidth limits of resonant or inertial active transducers. The ability to program reflection magnitude and phase, including introducing virtual delays, establishes the metalayer as a versatile platform for programmable acoustic metasurfaces. Its transparency in the passive regime and simple geometry make it suitable for scalable implementations and integration into complex acoustic systems. The approach opens avenues for non-reciprocal, non-Hermitian, and time-varying acoustic systems with unprecedented bandwidth and for applications in noise control, room acoustics, audio engineering, imaging, and complex source design.

The work introduces and experimentally demonstrates plasmacoustic metalayers—ultrathin, actively controlled plasma interfaces—that achieve ultrabroadband, deep-subwavelength sound control. Using frequency-independent monopolar and dipolar source mechanisms from corona discharge, the authors realize near-perfect absorption from 20 Hz to 2 kHz with a ~31 mm-thick device and demonstrate programmable reflection magnitude and phase (time delays) via impedance targeting. This non-resonant, non-inertial actuation circumvents the bandwidth-thickness trade-off of passive structures and the inertia limits of traditional active transducers. Future research directions include time-varying control laws for breaking time-invariance, implementing non-reciprocal and topological acoustic functionalities, scaling to large metasurfaces and arrays, adapting to oblique incidence and complex boundary conditions, and robust control strategies accounting for environmental variability of plasma discharge properties.

• High-frequency limitations: The analytical model assumes collocated monopolar and dipolar sources and an effectively infinitesimal layer; at higher frequencies these assumptions break down. Control system delays and finite-order rational approximations of the transfer function also degrade performance at high frequency.
• Low-frequency measurement limits: Experimental setup signal-to-noise constraints prevented validation below 20 Hz, although the approach is, in principle, extendable to lower frequencies.
• Environmental sensitivity: Plasma discharge parameters (e.g., ion mobility, critical voltage) depend on humidity, temperature, and pressure, potentially affecting stability and necessitating re-estimation or adaptive identification for robust operation.
• One-dimensional/plane-wave assumption: Experiments and modeling are in a duct under plane-wave conditions; extension to full 3D environments may require additional sensing/control strategies and model refinements.
• Stability and safety constraints: Active high-voltage operation (8 kV bias, ~2.5 W power to maintain ionization) imposes practical constraints on implementation and requires careful control design to ensure stability over broad bands.