Printability and Swelling Behaviors of the Hydrogel Formula
To meet the desired printability and swelling behaviors, the AAm-Alg hydrogel formula was carefully designed, as shown in Fig. 1. AAm has been extensively studied as a UV-curable hydrogel solution. To minimize the toxicity of printed samples, the hydrogel formula was improved by using LAP and Tartrazine, which have superior biocompatibility. And PEGDA (Mw = 1000) was added as a crosslinker to form rigid chains connecting the flexible chains of polyacrylamide (PAAm), enhancing the mechanical properties of the polymer network 40,41. Among the variety of choices, a formable formula (AAm: PEGDA: LAP: Tartrazine: DI water =1: 0.03: 0.03: 0.015: 4) was selected.
A sufficient fluidity of the UV-curable formula is crucial to reduce the difficulty of the recoating process in DLP printing, thereby improving both the printing efficiency and quality. A hydrogel solution with low viscosity enhances printability, including continuity and thickness control 42,43. However, to create a dual-network hydrogel with an adjustable modulus, Alg was added into the formula. The addition of alginate increased the viscosity and brought distinct non-Newtonian fluid behavior to the hydrogel solution, as shown in Fig. 2a. At a low shear rate of 10 s-1, the viscosity of the hydrogel solution (0-6 wt% Alg/AAm) increased from 3.58 mPa·s to 168 mPa·s (Fig. 2a),. Besides, the rheological properties of this formula are temperature-sensitive. In a temperature scan ranging from 25 to 60℃, the 4 wt% solution viscosity decreased from 61.4 mPa·s to 26.4 mPa·s @50 s-1 (Fig. S1).
In addition, the swelling behaviors of hydrogels are closely related to Alg. Cured PAAm-Alg hydrogels exhibit hydrophilicity due to the abundant hydrophilic functional groups in the polymer chains. Therefore, those cured samples swelled at immersion in solutions, resulting in undesirable shape changes 44. Excessive deformation is not conducive to printing scaffolds with high resolution.
Advanced dual-network hydrogel formula design helps the swelling control. The deformation extent of 0-6 wt% (Alg/AAm) UV-curable hydrogel samples in DI water, 0.005 M, 0.01 M, 0.02 M, and 1 M Fe3+ environments was recorded for 7 days. At the same Alg content, the samples treated in different ion baths exhibit similar deformations (Fig. S2). However, the swelling deformations ratio of the samples decreased with a higher content of alginate in the same ion bath (Fig. 2b). Notably, the addition of alginate created a double network hydrogel that limited swelling even in DI water
without ionic crosslinking. After 1 M Fe3+ bath treatment, the swelling deformation of 4wt% and 6wt% samples (~130%, and ~120%, respectively) were comparable and acceptable. Ultimately, balancing the viscosity and swelling requirements of the formula, a UV-curable hydrogel solution containing 4wt% alginate was selected in this study. To visualize the swelling deformation of hydrogel samples, a 10 mm hydrogel ruler (4wt% Alg/AAm) was printed. In Figure 2c, the state of the hydrogel ruler was shown directly after printing, either with 1 M Fe3+ bath treatment, or after immersion in DI water. The ratio of their length to width was nearly equal (2.67, 2.53, 2.67 respectively). It could be confirmed that, after swelling, the manufactured hydrogel samples undergo consistent deformation on a macroscopic scale.
It is worth mentioning that the cost-effective UV-curable hydrogel (priced at approximately $10 for 100 g solution) exclusively incorporates readily accessible commercial-grade raw materials, thereby diminishing the entry threshold for 3D bioprinting.
Optimization of DLP Parameters
To achieve high-resolution printing, the optimal printing parameters for the UV-curable hydrogel solution were investigated. Based on the Beer-Lambert law, the curing behavior of photocurable inks can be described by Jacobs equation 43,45:
where Cd is the depth of cure at a given exposure E, Dp is the transmission depth of UV-curable solutions, and Ec is the critical exposure intensity. Printing hollow structures with a transverse channel using different energy densities inputted on the top layer was performed (Fig. S3). The thickness of the top layer film of the channel was measured to characterize the cured depth of the hydrogel solution, and the relationship between the cured depth and the energy density was fitted
(Fig. 3a, Dp = 44.4 μm, ln(Ec) = 4.97 mJ/cm2). According to this model, the ideal exposure time for 10-40 μm thick film is 4-6 s.
Utilizing these optimized process parameters, micro lines with a width ranging from 10-50 μm (Fig. 3b) were printed, achieving the upper limit resolution of the DLP printer. The application of a
higher precision UV source could further improve the resolution. Furthermore, hydrogel samples with complex 2D/3D patterns were printed, including the Xiamen University emblem (Fig. 3c) and a 3D octopus (Fig. 3d), which reproduced the details of the circular design and characters with fidelity.
To further validate the printability of the material, a flexible sample containing double-helix internal channels was printed with a diameter of 1 mm (Fig. 3e). The two channels were infused with different liquids, which demonstrated smooth flow without any obstruction (Video S1). These features endow this hydrogel formula with the capability to manufacture integrated flexible and complex 3D microfluidic chips.
Mechanical Properties of the Hydrogels with Adjustable Modulus
Alginate in the hydrogel formula formed an adjustable ion crosslinking network in the environment of multivalent cations. Different multivalent ions exhibited varying effects on crosslinking alginate, such as Na+, Ca2+, Sr2+, Ba2+, Al3+, or Fe3+ 37,46. Besides, the concentration of the cations was also related to the crosslinking density of the polymer chains 47. By controlling the Fe3+ concentration in the ion bath, the crosslinking density of the PAAm-Alg double network hydrogel was manipulated, which allowed Young's modulus adjustment to hydrogel samples. After Fe3+ ion bath treatment, the color of the samples intensifies with the rising Fe3+ concentrations.
The introduction of Alg into the hydrogel formula led to a subtle increase of modulus in untreated samples pre- and post-swelling (Fig. 4b), due to the physical double network hydrogels of PAAm-Alg 48. Following the Fe3+ ion bath treatment, the hydrogel modulus exhibited a proportional increase with the ascending ion concentration, ranging from 15.8-345 kPa (Fig. 4a). Such a large modulus adjustable range allows the printed hydrogel scaffold to simulate the modulus of almost all human soft tissues. As shown in Fig. 4c and Fig. S4, the elastic deformation portion of the stress-strain curves in Fig. 4a was used to calculate Young’s modulus of hydrogel samples, revealing its relationship with the Fe3+ concentration in the ion bath. The modulus increased rapidly and proportionally with the ion concentration, which indicated a rise in crosslinking density. However, with further increasing the Fe3+ ion concentration (>0.1 M), the modulus value gradually flatten. This deceleration was attributed to a reduction in ion crosslinking sites within the alginate segments. Cyclic stretching tests on hydrogel samples (0.1 M Fe3+) showed that the stress-strain curves of the samples had good repeatability under 15%-30% strain (Fig. 4d).
Patterned Tissue Induced by Hydrogel Scaffolds.
Various soft tissues within the human body exhibit inherent orientations, such as the organization of cardiac tissue and the intricate patterns in vascular networks. Utilizing patterned hydrogel scaffolds offers a promising method for inducing well-organized or patterned tissue growth. The cell viabilities of hydrogels crosslinked with 0 M and 1 M Fe3+ were 1.01 and 0.98, as shown in Fig. S5, compared to the control group, indicating no obvious toxicity.
Cardiac tissue scaffolds with periodic H-shaped grooves (width of 100 μm, depth of 200 μm, gap of 110 μm) were manufactured. Upon immersion in DI water, the printed H-shaped grooves experienced controlled expansion. To facilitate communication between induced cardiac tissues within each groove, additional vertical grooves were incorporated into the H-shaped grooves (Figs. 5a,c). This design allowed the seeding of cells onto the hydrogel scaffold, inducing the formation
of a cohesive and organized tissue. Cardiac tissue is characterized by its softness, with a modulus of ~10 kPa 5-10. Therefore, the cardiac tissue scaffold was prepared without Fe3+ bath treatment to maintain a minimum modulus of ~16 kPa. A mixed suspension of self-fluorescent cardiomyocyte and fibroblast cell (2:1) was seeded onto the flexible scaffolds. Due to the gravity, a considerable number of cells were settled down and distributed within the grooves. Cardiomyocytes gradually migrated and aggregated, forming organized and continuous tissue (Fig. 5a). As shown in Fig. 5d, cardiomyocytes transitioned from an initial spherical shape to an organized tissue structure induced by H-shaped grooves. The tissue in different lines was connected through the horizontal connections. Therefore, an organized mesh-like tissue was formed through topographical constraint. Comparatively, cardiac tissue cultured on a flat substrate (Fig. 5e) failed to form an organized structure and tended to grow in clusters (Fig. 5b,f). Those cells underwent arbitrary clustering, displaying uncontrollable boundaries between tissue clusters, which is more apparent in the stained images (Fig. S6). The organized cardiac tissue could be driven to contract under an electric field generated by the electrical stimulation device. Under a pulse field with frequencies ranging from 0.5 to 1.5 Hz, the cardiac tissue exhibited synchronous contractions that followed the applied frequency (Video S2). However, at 2 Hz, although the tissue demonstrated contractions, it failed to keep pace with the frequency of the electric field.
Compared with cardiac tissue, blood vessels possess a higher modulus and exhibit a more intricate macroscopic structure. Vasculature-like hydrogel scaffolds were printed and treated with 1 M Fe3+ to obtain Young's modulus of ~345 MPa, mimicking that of native blood vessels as shown in Fig. 6a. The vascular tissue scaffold was designed with semi-open channels of varying diameters. After swelling, the diameter raised from the designed 1.2, 0.3, and 0.1 mm to 1.52, 0.47, and 0.16 mm, respectively. HCAECs were seeded onto the scaffold, forming a semi-open vascular network. Tissues grown on scaffolds exhibited apparent vascular morphological features. Upon staining HCAECs across the entire scaffold, it was evident that HCAECs could be well-distributed on the scaffold and microchannels (Fig. 6b-d). Within those 0.16 mm grooves (Fig. 6e), HCAECs exhibited distinctive vascular pattern. Conversely, endothelial cells cultured on a substrate devoid of grooves did not exhibit any noticeable pattern (Fig. 6f).