Photopatterned microswimmers with programmable motion without external stimuli

The study addresses how to achieve highly programmable, self-propelled microswimmers that operate without external stimuli, overcoming fabrication and programmability limitations of existing Marangoni-effect swimmers. The context is that many natural organisms exploit surface tension gradients (Marangoni effect) to move rapidly on water. Existing artificial Marangoni swimmers can autonomously move but suffer from imprecise, low-throughput fuel loading and limited integration of multiple functional parts, restricting motion programmability, especially below the millimeter scale. Furthermore, external-stimuli-driven microswimmers require bulky external energy/control sources and are hard to miniaturize due to battery constraints. The purpose here is to introduce a fabrication-compatible fuel (polyvinyl alcohol, PVA) and maskless photolithography to precisely integrate multiple functional materials, enabling programmed, time-dependent motion behaviors (e.g., direction change, disassembly) without external stimuli.

Prior work shows Marangoni propulsion allows insects like Microvelia and rove beetles to move rapidly by secreting surfactants to create surface tension gradients. Artificial Marangoni swimmers (macro to micro) have been demonstrated to move linearly or circularly for minutes without external energy, using simple fuel release systems (e.g., surfactants). However, amphiphilic fuels dissolve during fabrication washes, preventing standard microfabrication and precise fuel placement. Reported methods (manual pipetting, hand assembly) lack throughput and precision, and film machining approaches yield uncontrolled, omnidirectional fuel release that reduces efficiency and programmability. External-stimuli swimmers (light, ultrasound) can be programmable but require bulky equipment and cannot easily integrate onboard power at microscale. Hence, there is a need for fabrication-compatible fuels and methods enabling precise, scalable, multi-material microswimmers with programmable motion.

• Design principle: Marangoni microswimmers consist of a hydrophobic, floating body (polyurethane acrylate, PUA) and a fuel compartment formed by a PEGDA hydrogel containing solid-phase PVA. Upon contact with water, the hydrogel imbibes water and releases PVA, lowering local surface tension to produce propulsion from low to high surface tension regions.
• Photolithographic fabrication: Maskless photolithography patterned each compartment (fuel, body, rudder) by UV irradiation onto photocurable monomer solutions.
  • Fuel part: PVA (87–90% hydrolyzed, MW 30–70 kDa) was dissolved in DMSO and mixed with PEGDA 700 (fuel resin). UV polymerization crosslinked PEGDA while PVA remained unreacted. The uncured solution was spun off (spin-coating 90 s: 1500 rpm 30 s, 3000 rpm 60 s). Ethanol wash precipitated/solidified PVA within the hydrogel network. Additional photolithography steps for other parts were performed in lipophilic solvents (e.g., ethanol), leveraging PVA’s poor solubility in such solvents.
  • Body part: PUA monomer patterned by UV as floating hydrophobic body.
  • Multi-part integration: Sequential compartments crosslinked via residual moieties between steps to form integrated swimmers. UV exposure: 0.1 s per component at ~20 mW/cm² through a DMD-based system (Lightningcure LC8 Hg-Xe lamp; DMD by TI; IX71 microscope, 10x NA 0.3). Spacer between PDMS-coated and TMSPMA-coated glass slides set swimmer thickness (e.g., 210 µm typical for fuel disks). Feature sizes ranged from a few micrometers to millimeters. Throughput: ~2400 microswimmers per 76.2 mm × 25.1 mm slide in ~12 min, including centrifugation/spin and PVA solidification; scalable with slide size.
  • Rudder fabrication: Porous PEGDA hydrogels formed by adding anhydrous ethanol (porogen) at various PEGDA:ethanol volume ratios (10:0 to 5:5) to tune pore size and swelling kinetics.
• Characterization:
  • SEM verified PVA-loaded hydrogel morphology and cross-sections versus PVA-free controls.
  • PVA release quantification: UV–Vis spectrophotometry (NanoDrop) at λ_max ~215 nm. Calibration built from PVA standards; multiple identical fuel compartments (e.g., 224/mL) incubated to quantify low mass release. Time-resolved release profiles measured for varying PVA loading concentrations and fuel geometries (disks of known radius and thickness).
  • Swimming tests: Swimmers released on water; motion recorded (iPhone 8 or Photron FastCam MC2.1). ImageJ used to extract trajectories and speeds. Propulsion time defined from start until speed fell to average speed of fuel-less swimmers (~0.3 cm/s baseline due to environmental noise). Various body/fuel geometries produced linear, circular, and rotary motions; motion curvature tuned by body asymmetry and fuel placement.
  • Disassembly control: PVA bridges formed when two fuel compartments were patterned within ≤100 µm spacing, leaving solidified PVA connecting them. Additional PVA stacking on bridges by depositing PVA solution, spinning off to retain in the bridge indentation, and ethanol precipitation to increase thickness. A dissolution model used free-standing PVA films of varied thickness; time to form a 90 µm hole upon water exposure measured to map thickness-to-dissolution time and correlate with swimmer disassembly time (target bridge gap 90 µm).
• Materials: PEGDA (Aldrich), PVA (Aldrich), PUA (minuta), ethanol (Daejung), Irgacure 1173 photoinitiator (10 vol%), rhodamine B methacrylate dye (0.25 wt% in PUA for tracking), TMSPMA-coated glass for adhesion/compartment formation.
• Motion programming: Direction control via relative placement of fuel compartment (propulsion axis) and hydrodynamic drag distribution (body geometry). Time-dependent direction change via rudder swelling to adjust structural balance. Time-programmed disassembly via controlled-thickness PVA bridges linking swimmers or cargo.

• Fabrication and integration:
  • Maskless photolithography with PVA fuel enabled precise, multi-material microswimmers at high throughput (≈2400 per 76.2 × 25.1 mm slide in ~12 min). Sizes from micrometers to millimeters were produced. PVA remained intact during ethanol-based processing, enabling iterative fabrication steps.
• Motion modalities (without external stimuli):
  • Linear, circular, and rotary motions achieved by controlling fuel placement and body geometry. Circular motion realized by off-center fuel or asymmetric body drag; curvature tunable via body size. Rotary motion up to ~20 rotations/s observed.
  • Coordinated parallel movement of identical swimmers demonstrated.
• Fuel release and propulsion performance:
  • PVA lowers water surface tension significantly (literature: ~25% reduction at 5% PVA of MW ~70 kDa).
  • PVA release mass scaled linearly with initial loading and fuel size. For a 200 µm-diameter, 210 µm-thick disk at 0.7–3.5 wt% PVA, total released PVA ≈0.2–1.4 µg. >90% released within 10 s; release completed within ~2 min.
  • Propulsion reflected release kinetics: maximum speed at water contact, then decreasing with time. With maximum PVA concentration, 750 µm-axis rocket-like swimmers traveled >150 body lengths within 1 s at peak speed; average motility maintained for ~80 s, total distance ~1.5 m. Maximum speed and propulsion time increased approximately linearly with PVA concentration.
• Time-programmed direction change (hydrogel rudder):
  • Porous PEGDA rudders (PEGDA:ethanol 10:0 to 5:5) exhibited faster and larger swelling with higher ethanol content; swelling saturation time decreased from ~20 s to ~5 s as ethanol increased. Rudder size chosen to balance swimmer upon full swelling.
  • Programmable transition from circular to linear motion achieved; direction-change time tunable from ~3 s (more porous) to ~17 s (less porous), matching swelling saturation times. Reverse transitions (linear to circular) also demonstrated.
  • Spiral trajectories scanned areas up to ~12.57 cm² (≈25,000× swimmer top-face area) for PEGDA:ethanol 9:1.
• Time-programmed disassembly via PVA bridges:
  • When fuel compartments of two swimmers were ≤100 µm apart, a solidified PVA bridge formed, causing coupled propulsion followed by disassembly upon dissolution.
  • PVA bridge thickness increased with additional PVA stacking cycles; dissolution time of free films increased linearly with thickness, enabling predictive control of disassembly times from <1 s to ~10 s for bridged rocket-like swimmers.
  • Demonstrated scenarios: two opposing linear swimmers disassembled to convert circular-to-linear motion; identically directed swimmers disassembled while maintaining linear motion; cargo-like hydrogels bridged to a single swimmer and released after a set time.
• Application indications: Hydrophobic PUA bodies enable oil recognition/assembly at oil-water boundaries; spiral swimmers located oil-contaminated areas in preliminary tests.

The work directly addresses the challenge of programmable, autonomous microswimmer motion without external stimuli by introducing a fabrication-compatible fuel (PVA) and photopatterned multi-material integration. Precise fuel placement and body geometry control the propulsion and drag balance, enabling diverse motion modes and their time dependence. Quantitative characterization links fuel loading and geometry to PVA release kinetics and propulsion performance, providing a basis to study Marangoni swimmer kinetics systematically. The rudder concept uses tunable hydrogel swelling to time-align propulsion and drag axes, effecting controllable transitions (circular↔linear), while PVA bridges program timed disassembly events that change motion states or release cargo. These capabilities broaden utility in unstructured environments where individual swimmer control is impractical, and suggest applications in environmental sensing/cleanup (e.g., oil detection at interfaces), biomedical or microfluidic transport on aqueous media (including seawater or blood), and scalable deployment of many swimmers with predefined behaviors.

This study presents a high-throughput, maskless photolithography platform for Marangoni microswimmers using PVA as a fabrication-compatible fuel, enabling precise integration of multiple functional parts. The swimmers exhibit programmable linear, circular, and rotary motions, as well as time-dependent behaviors such as direction change via hydrogel rudders and timed disassembly via PVA bridges. Quantitative control over fuel release, speed, propulsion duration, and behavioral transitions is achieved through design of fuel loading, geometry, and hydrogel porosity. These advances provide both a versatile experimental platform to explore Marangoni propulsion kinetics and a pathway toward practical applications like environmental sensing and cargo delivery on aqueous surfaces. Future research should further elucidate collective behaviors, develop fuels with time-varying surface tension properties to mitigate crowding effects, and expand function in complex or biological fluids.

• Collective operation in confined spaces can reduce propulsion efficacy due to global surface tension lowering from many swimmers releasing fuel simultaneously; novel fuels with time-modulated surface activity may be needed.
• The overall kinetics of Marangoni swimmers, while better parameterized here, remain incompletely understood and warrant further theoretical and experimental study.
• Motion programming depends on precise fabrication and environmental conditions (e.g., water purity, airflow), and performance may vary with medium composition and external perturbations.