3D-Printed Hydrogels as Photothermal Actuators

Abstract

Thermoresponsive hydrogels were 3D-printed with embedded gold nanorods (GNRs), which enable shape change through photothermal heating. GNRs were functionalized with bovine serum albumin and mixed with a photosensitizer and poly(N-isopropylacrylamide) (PNIPAAm) macromer, forming an ink for 3D printing by direct ink writing. A macromer-based approach was chosen to provide good microstructural homogeneity and optical transparency of the unloaded hydrogel in its swollen state. The ink was printed into an acetylated gelatin hydrogel support matrix to prevent the spreading of the low-viscosity ink and provide mechanical stability during printing and concurrent photocrosslinking. Acetylated gelatin hydrogel was introduced because it allows for melting and removal of the support structure below the transition temperature of the crosslinked PNIPAAm structure. Convective and photothermal heating were compared, which both triggered the phase transition of PNIPAAm and induced reversible shrinkage of the hydrogel-GNR composite for a range of GNR loadings. During reswelling after photothermal heating, some structures formed an internally buckled state, where minor mechanical agitation recovered the unbuckled structure. The BSA-GNRs did not leach out of the structure during multiple cycles of shrinkage and reswelling. This work demonstrates the promise of 3D-printed, photoresponsive structures as hydrogel actuators.

Keywords: 3D printing; bioplotting; cycloaddition; gelatin; gold nanorods; photothermal heating; poly(N-isopropylacrylamide).

Figure 1

Schematic overview of material design and experimental approach for fabrication and actuation.

Figure 2

(a) Extinction spectra of aqueous CTAB-GNRs and BSA-GNRs normalized to peak extinction of 1 with an inset TEM image of BSA-GNRs; (b) DIW within a sacrificial support matrix of acetylated gelatin under UV exposure; (c) confocal fluorescence microscopy image of an unloaded structure dyed with fluorescein; (d) bright-field microscopy image of a GNR-loaded structure (OD 10), where the black dot in the center is an air bubble.

Figure 3

Shrinkage and reswelling behavior during one cycle of convective heating followed by three cycles of photothermal heating for OD 2.5, 5, 10, and 20 samples loaded with GNRs. (a) Measurements of the area during cycling in the top panel and the area of the shrunken structure divided by the area of the swollen structure in the bottom panel highlight minor differences in the shrinkage behaviors. To assess the error in the measurements, 10 measurements were taken of the area of an OD 10 sample in the swollen and shrunken state during the second cycle of photothermal heating, from which respective standard deviations of 0.019 and 0.0039 cm² were calculated. These errors propagated into an error of the ratio of the shrunken/swollen areas of 0.45% for the second cycle of photothermal heating for the OD 10 sample. These errors are representative of the other cycles and samples. Photographs of the (b) swollen state and shrunken states immediately after (c) convective heating and (d) photothermal heating with common scale bar (2 mm) shown in the left panel of (c). (e) Images from a thermal imaging camera immediately after photothermal heating, corresponding to (d). Note: Panels (b–d) were corrected to remove tick marks from a ruler on the edge of the photographs. An uncorrected version of these panels is presented in Figure S7.

Figure 4

(a–c) Photographs of an OD 20 sample, which underwent inward buckling during reswelling after successive cycles of photothermal heating. For each cycle, the left image shows an image of a buckled state, and the right image is after complete unbuckling and reswelling. The images are snapshots taken from Movie S6.