Acoustic Levitator

Sound waves alone can pin a droplet of water — or a living fly — in mid-air, without any physical contact and without any special properties of the material being held.

Creates a pattern of pressure nodes and antinodes in air between opposing ultrasonic transducers (or between a transducer and a reflector). Small objects — liquid droplets, powders, polystyrene beads, biological tissue samples, even live insects — are pushed toward the pressure nodes where radiation forces balance gravity. More advanced phased-array implementations can translate, rotate, and sort objects along 3D paths without any physical contact, making the technique valuable for contamination-free handling of pharmaceutical compounds and for mixing small droplets for crystallography experiments.

Two facing transducer arrays (or one array plus a reflector) emit 40 kHz ultrasound. Because the sound speed in air is ~343 m/s at room temperature, the wavelength is about 8.6 mm. When the two waves interfere, they form a standing wave with pressure nodes spaced at half-wavelength intervals (~4.3 mm apart). An object smaller than roughly half a wavelength (~4.3 mm) experiences a net acoustic radiation force — described by the Gor'kov potential — that pushes it toward the nearest pressure node. The force arises because the object is slightly more compressible and denser than air, so it scatters the sound field asymmetrically, and the resulting unbalanced momentum transfer traps it. Phased arrays can reshape this field in milliseconds by adjusting the relative phase of each element, translating objects along any 3D path at speeds up to ~25 cm/s.

In plain terms

Drop a ping-pong ball into a fountain and watch it hover in the water jet — the moving water pushes up exactly as hard as gravity pulls down. An acoustic levitator does the same thing with invisible sound waves instead of water: regions of high and low pressure take turns pushing the object from all sides, locking it in the quietest spot between the waves.

Also called: ultrasonic levitator, acoustic trap, standing-wave trap, acoustic tweezer (airborne).

Why it matters

Contactless handling prevents cross-contamination and eliminates substrate interactions that can change crystallisation pathways in pharmaceutical assays — the same compound crystallises differently when it never touches a vessel wall.
Asier Marzo and colleagues showed in 2015 (Nature Communications) that a single-sided 20×20 phased array can trap, translate and rotate particles using only holographically computed phase patterns, with no mechanical moving parts.
The technique works equally well for conducting, magnetic, biological, and optically opaque materials — any object less dense than roughly a few g/cm³ at the operating frequency can be levitated.
Sound travels well through biological tissue, opening routes for ultrasonic manipulation inside microfluidic chips and, in longer-term research, inside biological samples.

Where it lives

Marzo et al. single-sided levitator — 400-element (20×20) 40 kHz array, particles translated at 25 cm/s (University of Bristol / University of Sussex, 2015–present)
Holographic Acoustic Tweezers (HAT) system — 512-emitter (2 × 256) 40 kHz phased arrays, up to 25 simultaneous trapped particles (Nature Methods, 2019)
tec5 / Liguid acoustic levitators — commercial single-axis levitators used in pharmaceutical crystallisation and spectroscopy sample preparation
ESA containerless processing — acoustic levitation of molten metals in microgravity analogue experiments for materials science

Key numbers

Operating frequency (typical): 40 kHz (Below 20 kHz enters the audible range at dangerous intensities; above 60 kHz gives shorter working distances)
Wavelength in air: ~8.6 mm at 343 m/s (Sets the node spacing (~4.3 mm) and the maximum trappable particle size)
Maximum particle diameter: <~4.3 mm (λ/2) (Larger particles experience sign-reversals in the radiation force and are ejected)
Transport speed (phased array): Up to ~25 cm/s (Marzo et al. 2015, Nature Commun. 6:8661, 16 Vpp, 400-element array)
Array drive voltage (Marzo 2015): 16 V peak-to-peak (10 mm diameter Murata transducers; pressure amplitudes ~0–4000 Pa at trap)
Positioning accuracy: ~0.05–0.4 mm (λ/21 – λ/171) (Marzo et al. 2015 supplementary data)

Surprising facts

A 40 kHz acoustic levitator can suspend a droplet of mercury (density 13.6 g/cm³) — the densest liquid on Earth — in mid-air. The sound waves do not care what the object is made of, only how compressible it is relative to air.
The pressure nodes that trap objects are spaced only 4.3 mm apart at 40 kHz. Switching one transducer element's phase by half a cycle moves a trapped particle from one node to the next in milliseconds — the acoustic equivalent of shifting a gear.
Acoustic levitation can crystallise pharmaceutical compounds in new polymorphic forms that are impossible to grow on any solid surface. The absence of a nucleation-seeding wall allows metastable crystal structures to form that would otherwise never appear.
In the Marzo 2015 experiment, the optimisation algorithm computed the phase pattern for 400 transducer elements using 9000 gradient-descent iterations — and the resulting trap was strong enough to counteract a deliberate lateral push from a human finger without releasing the particle.

Open questions

Spontaneous particle clustering and ordered self-assembly: When multiple particles are levitated in the same standing-wave node, they do not simply crowd together at random. Secondary acoustic radiation forces (Bjerknes-type scatter interactions) cause them to self-organise into ordered 2D lattices with inter-particle spacings that scale with wavelength. The exact structure — square, hexagonal, chain — depends on particle size, density, and local field geometry in ways that are not fully predicted by pairwise Gor'kov theory. Clusters have been observed to spontaneously flip configuration without any change in the drive signal.

A short history

Acoustic levitation was first demonstrated experimentally by Bücks and Müller in 1933, who suspended small particles in a resonant tube between a piezoelectric emitter and a reflector. Ludwig Bergmann expanded the theory in the 1930s–40s. Systematic study of the acoustic radiation force on spherical objects was formalised by Gorkov (Gor'kov) in 1962, providing the potential-energy framework still used today. For decades the technique was confined to the laboratory curiosity of levitating single objects in a fixed node. The field transformed in 2015 when Asier Marzo, Bruce Drinkwater, and colleagues at the University of Bristol and University of Sussex demonstrated a single-sided phased-array levitator capable of full 3D manipulation — translating, rotating, and sorting objects — without any mechanical actuator, using a computed holographic phase pattern.

Sources

Marzo et al., Nature Commun. 6:8661 (2015) — holographic acoustic elements for levitated object manipulation (https://doi.org/10.1038/ncomms9661)
Gor'kov, Soviet Physics Doklady 6, 773 (1962) — forces on a small particle in an acoustical field (https://ui.adsabs.harvard.edu/abs/1962SPhD....6..773G/abstract)
Marzo & Drinkwater, PNAS (2019) — holographic acoustic tweezers (https://doi.org/10.1073/pnas.1813047116)
Physical Review Applied 18, 034026 (2022) — particle-size effect in airborne standing-wave acoustic levitation at 40 kHz (https://link.aps.org/doi/10.1103/PhysRevApplied.18.034026)