Development of smart photocrosslinkable Epoxy-based inks for 4D printing

Introduction: 4D printing, leveraging shape memory polymers (SMPs), enables the creation of adaptive constructs that can change their shape in response to external stimuli. Among various SMPs, epoxy-based materials, such as diglycidyl ether derivatives of bisphenol A (BDGE), have shown significant utility in 4D printing. However, conventional mold casting remains the predominant method for fabricating BDGE-based smart constructs, which constrains design flexibility. In this work, we developed BDGE-PEGDA smart inks with a two-step crosslinking process – (i) photocrosslinking of the PEGDA network, and (ii) thermal crosslinking of the BDGE SMP. The inclusion of PEGDA, not only facilitates the fabrication of intricate 4D geometries through restructuring steps but also allows for fine-tuning mechanical properties and thermal responsiveness. Moreover, BDGE-PEGDA inks are compatible with various fabrication methods including mold casting, photolithography, as well as 3D printing, offering enhanced versatility.

Materials and

Methods: The bulk ink was prepared by mixing two different inks, namely (i) shape morphing ink and (ii) photocrosslinkable ink. The shape morphing component included BDGE (M, 10 mmol) and poly(propylene glycol) bis(2-aminopropyl) ether (C, 275-400 mmol), while the photocrosslinkable component was composed of polyethylene glycol diacrylate (P, 0-25 wt%) and Irgacure 2959 (1 wt%). Post-mixing, the ink was extruded into the reusable PDMS replica mold and then exposed to UV light (365 nm) for 1-2 mins to initiate photocrosslinking of the PEGDA network, followed by thermal curing of the SMP network at 65 °C overnight (Figure 1A). Alternatively, gels after photocrosslinking steps could be retrieved and restructured on another PDMS mold to form complex 3D shapes and then cured thermally at 65 °C overnight. The shape memory behavior of 4D constructs was evaluated at 45 °C. Thermal (Tg value) and tensile properties of the samples were evaluated using differential scanning calorimetry and uniaxial mechanical tester, respectively. For biological evaluations, L929 mouse fibroblast cells were seeded on the scaffolds and their morphology was visualized post-staining with Texas red-phalloidin and DAPI using confocal microscopy.

Results, Conclusions, and Discussions: The developed BDGE-PEGDA 4D ink offers a unique advantage of photocrosslinking to form 4D gels, which allows their restructuring into complex shapes, which are otherwise unachievable using inks without PEGDA (Fig. 1B-C). The thermal recovery study shows that BDGE-PEGDA flower-shaped constructs recovered from temporary flat to complex permanent shape, and vice versa in the case of BDGE (Fig. 1D). On further in-depth characterization, we found that the rate of thermal recovery of the 4D constructs improved significantly with an increase in PEGDA concentration (Fig. 1E). Such variations were attributed to the changes in the mechanical properties of the samples instead of changes in their thermal values. Tg values of all the samples were within 22 - 23.5 °C (Fig. 1F); however, with the addition of PEGDA, samples demonstrated more plastic-like behavior (Fig. 1G). Besides, the samples were found to be biocompatible, supporting adhesion and growth of L929 cells, particularly those with 15% PEGDA and at least 3 mmol poly(propylene glycol) bis(2-aminopropyl) ether concentration (Fig. 1H).

Our study demonstrates that integrating a photocrosslinkable polymer, such as PEGDA, into conventional epoxy-based SMPs like BDGE, significantly improves design flexibility. This combination facilitates the creation of complex 4D geometries through various fabrication techniques such as mold casting and photolithography. Our forthcoming efforts will focus on optimizing the combination of PEGDA and BDGE for advanced additive manufacturing techniques, particularly stereolithography. This advancement holds promise for the creation of precise, functional 4D structures for applications in biomedicine, drug delivery, and biorobotics.

Acknowledgements (Optional): The authors would like to thank NSF EBMS program grants #2110842 and #1856321 for financial support.