Movable surface acoustic wave tweezers: a versatile toolbox for micromanipulation

The study addresses the limitation of conventional SAW tweezers whose acoustic fields are spatially fixed by the interdigital transducer (IDT) geometry and permanent bonding to microchannels, restricting dynamic translation/rotation and field switching. The research question is whether a multilayer, movable SAW platform can decouple the transducer from the channel to enable dynamic control (translation and rotation) of acoustic traps over a large working area while maintaining biocompatible, label-free manipulation. The work is motivated by the need for dexterous micromanipulation comparable to 3D acoustical tweezers, but within a SAW framework that is compact and compatible with lab-on-chip systems. The proposed system aims to provide flexible, movable acoustic fields capable of precise manipulation (translation, in-plane and out-of-plane rotation, clustering) of diverse targets (particles, bubbles, droplets, cells, microorganisms), improving degrees of freedom and working range and enabling selective, localized operations.

Prior SAW advances have pursued tunable, chirped, or reconfigurable fields using methods such as phase shifts and frequency modulation to achieve rapid sorting and manipulation. Examples include tunable slanted-finger SAW tweezers (dynamic acoustic field location/period control), wavenumber-spiral acoustic tweezers (reshaped pressure distributions), nanosecond pulse SAW tweezers (altered field regions inside PDMS channels), and bisymmetric coherent acoustic tweezers (complex wave-node arrays). Despite these innovations, SAW devices remain constrained by fixed transducer geometries and bonded channels, limiting field motion and reconfiguration. In contrast, 3D acoustical tweezers can easily move transducers and offer dexterity for arbitrary patterning. The paper situates its contribution as bridging this gap by introducing movable SAW fields without relying solely on complex signal modulation or permanent bonding.

System architecture: A movable SAW tweezer platform integrates XY and Z displacement tables, a customized 3D-printed clamp, and a multilayer SAW device. The lithium niobate (LiNbO3) piezoelectric substrate with IDTs is mounted on the XY table; an independently encapsulated PDMS microchannel is mounted on the Z table via the clamp. A thin liquid coupling layer connects the channel bottom to the transducer surface, enabling acoustic transmission while allowing relative motion between the transducer and the channel. Working principle: IDTs generate SAWs that propagate on the LiNbO3 surface and leak into the coupling layer at the Rayleigh angle, converting to longitudinal waves that pass through the coupling layer and channel bottom to form a standing acoustic field in the channel. Orthogonal IDT pairs produce a superposed standing field with isolated nodal traps. By translating/rotating the transducer relative to the channel, the nodal pattern and thus the acoustic traps move accordingly, dragging trapped particles along predefined trajectories. Demonstrated control includes trajectory drawing and complex path following. Multilayer acoustics: The device includes four layers: LiNbO3 substrate, liquid coupling layer, channel bottom (either PDMS film or borosilicate glass sheet), and the fluid in the microchannel. Due to sound-speed mismatches, wavelengths differ in each layer (λs on substrate; λc in coupling layer; λb in channel bottom; λa in channel). Acoustic attenuation differs by material: attenuation length on the substrate surface scales with SAW wavelength; in fluids, attenuation length scales with the square of the longitudinal wavelength. Design therefore balances impedance matching and attenuation. Channel bottom design: Two materials evaluated—PDMS (impedance close to water, favorable for resonance and bonding) and borosilicate glass (stiff, robust, easier handling but larger impedance mismatch). Thickness trade-offs exist: thick layers increase attenuation; overly thin layers reduce robustness and complicate bonding. For a rigid-superstrate–liquid–LiNbO3 stack, a superstrate thickness of ~0.35–0.7λ can maximize channel power density; in this work, glass thickness of 130–160 µm is used (within that regime). PDMS thin films are also used where resonance and coupling are desired. Coupling layer: A thin liquid film (deionized water or fluorinated oil) provides reversible coupling, enabling detachability and reusability of chips. Simulations (COMSOL) explore pressure distributions for different coupling-layer heights (e.g., 100 µm, 200 µm) and SAW modes (symmetric/antisymmetric), showing spatial-period changes along the Z-axis and the formation of isolated nodes within the channel. Experimental platform and operation: The relative motion of IDTs and channel is commanded via displacement tables to realize: (1) linear translation of acoustic traps; (2) in-plane rotation (XY) about arbitrary centers; (3) out-of-plane rotation (tilting) of targets. A visualization with a fluorinated oil film on the substrate qualitatively shows the SAW interference pattern. Particle manipulation is demonstrated by translating a 31.1 µm particle to draw trajectories; input powers and SAW wavelengths (e.g., λs = 200 µm or 400 µm) are varied. Velocity–power relations for different particle sizes and coupling media are recorded (graphs provided for water vs fluorinated oil coupling).

• The multilayer, movable SAW tweezer system decouples the transducer from the microchannel, enabling dynamic translation and rotation of the acoustic field over millimeter-scale ranges while maintaining micron-level precision of target motion.
• Demonstrated precise manipulations include translation, in-plane rotation about arbitrary points, out-of-plane rotation, and cluster formation of diverse targets (particles, bubbles, droplets, cells, microorganisms).
• Direct field motion: Acoustic traps move concomitantly with transducer motion, allowing intuitive path control; a 31.1 µm particle was guided to trace complex trajectories (including a Mona Lisa dot pattern) at an input power of 9.8 W.
• Localized field reconfiguration via microtools: Micro air bubbles can act as secondary sound sources to locally reshape the SAW field, enabling indirect manipulation of biotargets and spatially selective collection of cancer cells, with minimal impact on unintended samples.
• Multilayer acoustics validated: Simulations show that wavelength and pressure-field spatial periods vary across layers and with coupling-layer thickness (e.g., 100 vs 200 µm) and mode symmetry (symmetric/antisymmetric), aligning with observed node patterns visualized using a fluorinated oil film.
• Performance dependencies characterized: Velocity–power characteristics were measured for different particle sizes (50–250 µm) under varying conditions (e.g., λs = 200 µm or 400 µm; coupling layers of deionized water vs fluorinated oil; transducer–channel spacing ds = 1.0 or 3.2 mm), indicating tunable manipulation performance via wavelength, spacing, coupling medium, and power.

By allowing relative motion between the IDT-bearing substrate and the microchannel, the system overcomes the intrinsic limitation of fixed-field SAW devices, delivering dexterous, large-range, and precise manipulation akin to movable 3D acoustical tweezers while retaining SAW advantages (compactness, label-free, biocompatibility). The observed ability to translate traps, execute arbitrary-center in-plane rotations, and perform controlled out-of-plane rotations addresses the need for versatile operations in biological and microrobotic applications. Furthermore, leveraging microbubbles as local acoustic scatterers introduces a means for on-demand, spatially selective field reconfiguration that is not constrained by transducer geometry or input-signal-only approaches, enabling targeted actions such as selective cancer cell collection. The multilayer design and coupling strategy also facilitate chip reusability and flexible interfacing, suggesting a pathway toward practical benchtop micromanipulation platforms that bridge innovations in SAW acoustofluidics with biomedical workflows.

The work introduces movable SAW tweezers with a multilayer, decoupled architecture that enables dynamic, precise manipulation of microscale targets via transducer–channel relative motion. Key contributions include: (1) large working-range translation with micron-level precision; (2) arbitrary-center in-plane rotation and controllable out-of-plane rotation; (3) localized wavefield reconfiguration using microbubbles as secondary sources for selective biomanipulation; and (4) a reusable, detachable superstrate-coupling configuration that enhances flexibility. The platform expands the degrees of freedom of SAW manipulation and offers a practical interface for complex micromanipulation in biomedical microrobotics. Future work could optimize acoustic coupling and attenuation via material/thickness tuning, quantify biocompatibility across cell types under varied powers/durations, integrate automated multi-axis control for closed-loop operations, and extend microtool strategies (e.g., engineered scatterers) for programmable, task-specific field shaping.

• Acoustic impedance mismatches among layers (substrate–coupling–channel bottom–fluid) can produce reflections and modify wavelengths, complicating precise field prediction and potentially reducing efficiency.
• Attenuation trade-offs: Thicker channel bottoms increase attenuation; very thin films improve coupling but reduce robustness and complicate bonding/handling. Material choice (PDMS vs glass) impacts both impedance matching and mechanical durability.
• Coupling-layer thickness and material (water vs fluorinated oil) strongly affect field distribution and manipulation performance, requiring careful calibration and potentially limiting universality across setups.
• Quantitative performance metrics (e.g., maximum velocities, forces, long-term cell viability across powers) are not fully detailed in the provided excerpt, and external conditions (ds spacing, λs, medium) may constrain operational windows.