Multimaterial cryogenic printing of three-dimensional soft hydrogel machines

Abstract

Hydrogel-based soft machines are promising in diverse applications, such as biomedical electronics and soft robotics. However, current fabrication techniques generally struggle to construct multimaterial three-dimensional hydrogel architectures for soft machines and robots, owing to the inherent hydrogel softness from the low-density polymer network nature. Herein, we present a multimaterial cryogenic printing (MCP) technique that can fabricate sophisticated soft hydrogel machines with accurate yet complex architectures and robust multimaterial interfaces. Our MCP technique harnesses a universal all-in-cryogenic solvent phase transition strategy, involving instant ink solidification followed by in-situ synchronous solvent melting and cross-linking. We, therefore, can facilely fabricate various multimaterial 3D hydrogel structures with high aspect ratio complex geometries (overhanging, thin-walled, and hollow) in high fidelity. Using this approach, we design and manufacture all-printed all-hydrogel soft machines with versatile functions, such as self-sensing biomimetic heart valves with leaflet-status perception and untethered multimode turbine robots capable of in-tube blockage removal and transportation.

Fig. 1. Multimaterial cryogenic printing technique for soft hydrogel machines.

a Schematic diagram of the mismatch between desired structure designs and soft formless hydrogels, owing to weak and inconsistent interfacial mechanics across low-density polymer networks. b Schematic diagram of proposed multimaterial cryogenic printing technique by all-in-cryogenic solvent phase transition, involving cryogenic printing and cryogenic cross-linking procedures. c Schematic diagram of an untethered multimode turbine robot actuated by a rotating magnetic field: sweeping with soft-hard composite blades and dragging by generating a trapping vortex. d Schematic diagram of a turbine robot capable of transporting cargo through an underwater Y-shape tube by steering its rotational axis to switch between sweeping and dragging modes. M in-plane magnetization, B magnetic field.

Fig. 2. Characterizations of multimaterial cryogenic printing technique.

a–e Cryogenic printing: a Schematic diagram of printing high-resolution 3D structures enabled by instant solidification against capillarity- and gravity-driven instability. b In-situ observations of instant ink-solvent solidification and SEM cross-sectional image of printed filamentary hydrogels. Scale bars, 200 μm (optical images), 5 μm (SEM image). c The kinetic predictions on printed linewidth compared to experimental measurements under varying platform moving speeds and extrusion pressures. Pred. prediction. d The kinetic prediction errors on the layer thickness of thin walls compared to experimental measurements, by using multiple hydrogel inks like PEDOT:PSS-PVA, PVA, and SA. e The generalized resolution improvement on multiple hydrogel choices and printing substrates by using cryogenic printing. RTP room-temperature printing, PI polyimide, PDMS polydimethylsiloxane, PET polyethylene terephthalate. The comparisons use the same 32 G nozzles for printing hydrogel inks. f–j Cryogenic cross-linking: f Schematic diagram of cryogenic cross-linking reactions at the synchronously melting ice-water interfaces. g Differential scanning calorimetry (DSC) curves of typical hydrogel inks and cross-linking baths. The cross-linking process is designed to react within the temperature window between the melting temperature of frozen inks and the solidification temperature of baths. h Raman spectra of frozen PVA structures after immersing in cross-linking baths show the reaction process at −5 °C. i Uniaxial tensile tests of the printed multimaterial hydrogel samples and corresponding composed single-material samples. Typical snapshots of the heterogeneous samples are inserted within this panel. @, heterogeneous sample. j Printing performance comparisons on mechanical tunability (the ratio of maximum and minimum Young’s modulus, Y_max/Y_min) and extreme aspect ratio (the ratio of structural height to resolution) with existing approaches. DIW direct-ink-writing. Error bars indicate one standard deviation from the mean over three samples.

Fig. 3. Characterizations of printed structures.

a–c Single-material structures: a Sierpinski pyramids. b Y-shape tubes. c Hollow cubes with a spherical cavity. d–f Multimaterial structures: d Voxelated digital cube. e Cylindrical in-tube stent. f Primitive lattice. g–i Material interface in a printed hydrogel gyroid lattice: g SEM images of material interfaces between PEDOT:PSS-PVA and PVA show a tightly bonding morphology. h The soft printed hydrogel withstands twisting deformation. i The printed hydrogel recovers to its original shape after twisting, illustrating mechanical robustness. j–l Shape fidelity in a printed hydrogel primitive lattice: j X-ray computed tomography (CT) scanning slices demonstrate a uniform structure array with thin-walled ( < 1 mm), overhanging, and hollow features: Horizontal (X-Y plane) and vertical (X-Z plane). k, l The printed structure overlays well with the designed model, and quantitative error analysis shows agreement between the two models (428 μm error range corresponds to 95.4% of all points). To distinguish different materials, the printed hydrogels are composed of pigments: PVA (red) and SA (blue). Scale bars, 5 mm (a–f), 20 µm (g), 1 mm (j), 5 mm (k).

Fig. 4. Self-sensing biomimetic heart valve.

a Schematic diagram of a biomimetic hydrogel heart valve: Flow-response leaflets and induced PEDOT:PSS-PVA chamber signal sensing. b Photograph of printed hydrogel heart valve and surface profile analysis. Its profile error is less than 6% with a maximum surface inclination of 43.7°. c Photograph of typical operation statuses of the hydrogel heart valve and corresponding loading and unloading resistance responses, exhibiting a great linearity R² = 0.99 under simulating fluid pressure (−11.7 to 12.8 kPa). d Loading and unloading resistance responses under periodic square-wave flow show timely, repetitive, and stable signals following the leaflet displacement. This self-sensing valve can withstand extreme hydrodynamic pressure ( > 140 mmHg), covering the normal range of native aortic blood pressure. Scale bars, 5 mm.

Fig. 5. Untethered multimode magnetic turbine robot.

a Photograph of continuously printing the turbine robot into customized sizes (typically ~12.5 mm diameter and ~5.4 mm height). b The simulation analysis of a typical flow field around the rotating swimming turbine robot in an underwater pipe with a diameter of 20 mm. The trapping vortex behind the robot is marked in gray color, where the vector direction of the Y-axis component velocity is opposed to the Y direction. v, swimming speed; f, rotating frequency. c The relative velocity component in the Y direction of trapped objects to the robot under varying swimming speeds and rotating frequencies. d The removal process of sticky blockages in an underwater straight tube by using the turbine robot to trap floating obstacles and avoid blockage migration. Typical processes include: Swimming to the target location, rotating sweeping, flipping (mode switching), vortex trapping, and dragging for removal. The propulsion and rotation directions of the robot are marked in white and brown, respectively. e The multimode turbine robot transports a capsular-like cargo through a Y-shape tube by steering its rotational axis. The angle between the front and rear paths is about 125.6°. Scale bars, 5 mm.