4D Printing with Mechanically Robust, Thermally Actuating Hydrogels

A smart valve is created by 4D printing of hydrogels that are both mechanically robust and thermally actuating. The printed hydrogels are made up of an interpenetrating network of alginate and poly(N-isopropylacrylamide). 4D structures are created by printing the “dynamic” hydrogel ink alongside other static materials. 4D printing is an exciting emerging technology for creating dynamic devices that can change their shape and/or function on-demand and over time.1-3 4D printing combines smart actuating and sensing materials with additive manufacturing techniques to offer an innovative, versatile, and convenient method for crafting custom-designed sensors,4 robotics,5 and self-assembling structures.1 Stimuli-responsive volume-change materials incorporated into multimaterial structures can be harnessed to create movement in the same way that biological muscles achieve motion in animals and nastic movements are generated in plants.6, 7 Current 4D printed examples utilize water absorption1 or thermal shape memory2 to demonstrate impressive shape change, but are slow to respond, show limited reversibility, and restricted to bending type motions of flexible structures that generate little force. We here describe relatively fast and reversible, skeletal muscle-like linear actuation in 3D printed tough hydrogel materials and their incorporation into a smart valve that can control the flow of water. Hydrogels were utilized as the active material in the 4D printed structures since these materials have been processed with printers8 and some examples of hydrogels can drastically, and reversibly, alter their volume in response to changes within their environment.9, 10 We start with an ionic covalent entanglement (ICE) hydrogel11 that can be 3D printed12 and demonstrate high toughness.12 The latter is important since we require thin sections that respond quickly to external stimuli but also need to be mechanically robust to support the internal and external mechanical loads. ICE gels are a type of an interpenetrating polymer network hydrogel that is made up of an entanglement of one polymer network crosslinked with metal cations and a second polymer network crosslinked with covalent bonds.11, 13-15 This dual network structure is similar to that in double network hydrogels16, 17 and facilitates high toughness through large-scale crack-tip energy dissipation by unloading of the strands in the tight network,18 in this case due to dissociation of the ionic crosslinks.19 The loosely crosslinked covalent network serves to bridge the damage zones created by the loss of ionic bonds and prevents catastrophic crack propagation. Here, we use a thermally responsive covalent crosslinked network of poly(N-isopropylacrylamide) (PNIPAAm) to function as both the toughening agent and to provide actuation through reversible volume transitions. PNIPAAm is a widely studied temperature-sensitive hydrogel that exhibits a large reversible volume transition at a critical temperature, TC (≈32–35 °C).20, 21 The volume change is caused by a coil–globule transition of the polymer network strands22 and results in a large decrease in the water content when the temperature is increased above Tc.20 Alginate/PNIPAAm ICE gel inks with various concentrations of NIPAAm were prepared for extrusion printing. The inks had a constant alginate concentration of 4% (w/v) and NIPAAm concentrations of 10%, 15%, or 20% (w/v) with fixed amounts of covalent crosslinker and UV initiator. Printing and curing at subambient temperatures (10 °C) was necessary to prevent phase separation of the PNIPAAm (Supporting Information). Rheological characterization showed that the inks had appropriate flow12 and gelation behaviors at 10 °C for extrusion printing (Supporting Information) with an Envisiontec 3D-Bioplotter coupled with a UV-light source that enables in situ photopolymerization. The printed hydrogels were immersed in 0.1 m calcium chloride solution to crosslink the alginate polymer network and to enhance their mechanical properties.15, 23 The fully crosslinked gels were then rinsed and further immersed in water at 20 °C and swollen to equilibrium. Swelling ratios were determined for each of the ICE gels to determine their suitability for thermally induced actuation. Increasing the PNIPAAm concentration in the ICE gel ink from 10% to 20% (w/v) made no significant difference to the swelling ratio of the printed hydrogels when swollen in water at 20 °C (Figure 1), and were of a similar magnitude to previously described printed alginate/PAAm gels.12 The degree of swelling in these ICE gels has been shown previously to be controlled by the more tightly crosslinked alginate network and is relatively insensitive to the relative fractions of the ionic and covalent networks in the range used here.12 In all cases, the swelling ratio of the alginate/PNIPAAm gels was reduced significantly to 0.52–1.5 when swollen in water at 60 °C with a greater contraction occurring in the gel with the higher PNIPAAm content. The chosen temperature was well above the PNIPAAm thermal transition so as to rapidly heat the gel and generate fast actuation. Volumetric contraction ratios on heating from 20 to 60 °C increased from 6.1 to 9.3 to 16.7 for gels made from inks with NIPAAm contents of 10%, 15%, and 20% (w/v). The alginate portion of the ICE gel resists the contraction of the thermally responsive PNIPAAm phase, so the contraction ratio decreased as the alginate fraction increased. The gels dramatically changed in size when heated as shown in Figure 1a,b. Increasing PNIPAAm content in the ICE gels resulted in larger unconstrained length strains at 60 °C based on their equilibrium size at 20 °C of 41 ± 1%, 44 ± 1%, and 49 ± 1%, for the gels prepared from 10%, 15%, and 20% NIPAAm inks, respectively. The reversible nature of the thermally induced actuation was observed when the alginate/PNIPAAm gels swelled to their initial equilibrium conditions when cooled from 60 to 20 °C. The mechanical performance of printed alginate/PNIPAAm ICE gels was characterized in tension to evaluate their potential for use in load-bearing applications. Mechanical tensile tests were performed at temperatures both above and below the PNIPAAm swelling transition. As expected, the tensile strength and modulus of all ICE hydrogels increased considerably in the less-swollen state at 60 °C in comparison with the highly swollen state at 20 °C (Figure 1d,e and Table S2, Supporting Information). The fully swollen gels had similar mechanical properties to printed alginate/PAAm hydrogels described previously.12 As the NIPAAm concentration in the printing solutions was increased from 10% to 20% (w/v) the tensile strength and modulus decreased considerably when tested at 20 °C. Since the swelling ratio of the fully swollen gels at 20 °C were similar, the differences in mechanical properties can be attributed to the different ratios of the ionically crosslinked alginate network to the covalently crosslinked PNIPAAm network in the ICE gels. Recent work has shown that the moduli and tensile strengths of ICE gels are increased with higher mass ratios of ionic network to covalent network, since the former is more tightly crosslinked than the latter.13, 18 In contrast, it is seen in Figure 1d that at 60 °C the tensile strength increased with decreasing ratio of alginate to PNIPAAm. This strength increase can be attributed to the significantly reduced swelling ratio of the hybrid gels with higher PNIPAAm contents. It is well-known that the magnitude of strength and modulus are inversely proportional to the swelling ratio.24 The performance of the alginate/PNIPAAm ICE gels as thermally stimulated actuators was next investigated to assess their suitability for use in 4D printing applications. Although the free swelling of hydrogels has been extensively studied, few investigations have considered their swelling when subjected to an external load.20, 25 We have developed a method to observe the constrained swelling of the alginate/PNIPAAm ICE gels with a Universal Tensile Tester equipped with a water sleeve. Hydrogel samples were clamped to the mechanical tensile tester inside the sleeve which was filled with water at 20 °C. To ensure that the hydrogels were under tension each specimen was extended to a stroke of 1 mm (4% strain). To begin the actuation testing the water was drained from the sleeve and immediately replaced with water heated to 60 °C. The increase in temperature caused the hydrogels to deswell, however the hydrogels could not contract because the length was fixed. Consequently, a force was produced as a result of the constrained deswelling. Stress was determined from the measured force and the initial cross-sectional area of the sample (at 20 °C) and is shown as a function of time for one sample in Figure 2a. The generated stress reached a constant value after approximately 15 min and provided a measure of the blocked stress. Increasing PNIPAAm content in the ICE gels resulted in higher blocked stresses of 10.2 ± 0.6, 17.3 ± 0.4, and 20.9 ± 0.6 kPa (n = 3) for the gels prepared from 10%, 15%, and 20% NIPAAm inks, respectively. The magnitude of the blocked stress is determined from the extent of volume contraction on heating and the elastic modulus of the gel in the heated state. The much larger deswelling of the gels with higher PNIPAAm contents accounts for their higher observed blocked stresses. Interestingly, the blocked stress values were considerably smaller than expected. The blocked stress (σb) can be calculated from the free strains (εf) and final elastic modulus (ET) as . Using the unconstrained free contractions and the elastic moduli obtained from tensile tests at 60 °C, the predicted blocked stress are 590, 770, and 930 kPa for the ICE gels prepared from inks containing 10%, 15%, and 20% NIPAAm, respectively. These values are more than 50 times larger than the measured blocked stresses and the differences are likely due to the known dissociation of ionic crosslinks in the alginate network under load.11, 13 When the ICE gels were constrained during deswelling the internal forces caused some of the ionic crosslinks to fail in a stress relaxation process.26, 27 This loss of crosslinks will also cause a decrease in elastic modulus, as is confirmed by the lower elastic modulus at 60 °C (58–100 kPa) determined from the stress–stroke curves (Figure 2b) than as obtained from tensile tests (1400–1900 kPa) of previously unstrained gels (Table S2, Supporting Information). Similarly, large stress relaxation has been observed in alginate networks with ionic crosslinks.26-28 The stress–stroke curves were obtained from the second stage of the actuation test (Figure 2b). Here, the stress generated from the constrained deswelling of the hydrogels was relaxed to zero by steadily reducing the separation between clamps while recording force and crosshead displacement. The samples were then immediately restretched to the initial strain and this unloading/reloading process was repeated three times. Typical stress–stroke curves for each of the three hydrogels are shown in Figure 2b. The free strain was determined from the strain when the stress was equal to zero. As expected, free strains increased with increasing PNIPAAm content in the gels: 12 ± 2%, 19 ± 2%, and 35 ± 3%, n = 3 for the gels prepared from 10%, 15%, and 20% NIPAAm inks, respectively. These free strains were significantly smaller than as measured in the unconstrained contraction tests where contractions were in the range of 40%–49%. The free strain obtained from the stress–stroke curves are determined by the ratio of blocked stress to elastic modulus, and the larger decrease in blocked stress in comparison with the elastic modulus accounts for the observed smaller free strains. Finally, it was noted that the slope of the reloading stress–stroke curve was steeper than the unloading curve which resulted in larger free strain and blocked stress values in the second and third stress–strain cycles (Figure S3, Supporting Information). This feature can also be associated with the labile nature of the ionic crosslinks in the alginate network. It is known that partial reformation of ionic crosslinks occurs when the ICE gels are unloaded.11, 13 The reloaded ICE gel has a higher ionic crosslink density than during the previous unloaded resulting in a steeper slope in the reloading stress–stroke curve than in the previous unloading. The final step of the actuation testing was the reswelling of the hydrogels in 20 °C water. The stroke was maintained at 1 mm for 120 min and the water sleeve was drained and quickly refilled with 20 °C water. The hydrogels absorbed water which resulted in a relaxation of the stress. Hydrogels printed with a 10% (w/v) NIPAAm concentration showed a stress relaxation of 13.3 ± 0.7 kPa over 25 ± 1 min. The hydrogels printed with a 20% (w/v) NIPAAm concentration relaxed 27.0 ± 0.3 kPa of stress over a larger response time of 78 ± 2 min. These stress relaxations through cooling/swelling were larger than the corresponding stress generated during heating/contraction of the same gel. The loss of ionic crosslinks under load at 60 °C allows for a greater swelling at 20 °C than occurs in the pristine gel. Typical of PNIPAAm hydrogels the swelling rate of cooling hydrogels in 20 °C water was slower than the deswelling rate of the heating hydrogels in 60 °C water.22 A summary of all of the actuation testing data is included in Table S3 in Supporting Information. The alginate/PNIPAAm ICE gels show promise for application as actuators because they can repetitively achieve a large free strain and blocked stress. We demonstrated how these materials can be fabricated into a smart valve for control of water flow29-33 with 4D printing. A multi­component computer-aided design (CAD) model of the smart valve was prepared with Solidworks software (Figure 3a) and printed with an Envisiontec 3D-Bioplotter (Figure 3b). Each of the four components was assigned a different material ink as follows: rigid Emax epoxy compound for the valve design contained tubing and a cap;34 alginate/PNIPAAm ICE gel ink (10% (w/v) NIPAAm) for actuators that open and close the valve; alginate/poly(acrylamide) ICE gel ink for the gasket and a channel for the flow of water; and an alginate-based ink for the sacrificial support structures. The valve was printed in the open configuration (Figure 3c), and when hot water made contact with the actuator components the valve experienced a 4D printing transformation. During heating the actuators contracted, to close the outlet and blocking the flow through the water channel (Figure 3d). The actuating gel was arranged as six parallel strips so as to decrease response times and generate sufficient force to seal the valve when closed. A video that illustrates the function of the valve is included in the Supporting Information. The valve design was evaluated by measuring the flow rate of water through the printed valve. The flow rate was calculated by measuring the time required for a 5 mL column of water to pass through the valve. Water passed through the open valve at a rate of 1.40 ± 0.04 mL s−1 (Figure 3e). The flow rate was then assessed after exposing the valve to 60 °C water for 1 min. The flow rate through the valve dropped by more than 50% to 0.63 ± 0.07 mL s−1. The reversible nature of the valve was demonstrated by passing 20 °C water over the hydrogel actuators which caused the valve to reopen. A video showing the valve actuation is provided in the Supporting Information. The average flow rate through the closed valve was 0.60 ± 0.10 mL s−1 after 1 min heating time and when repeated over five cycles of opening and closing. The valves show a reproducible degree of control. The valve could be almost completely closed by increasing the amount of time that 60 °C water was passed over the hydrogel actuators. After 2 min in the 60 °C water the flow rate through the valve reduced to 13 ± 2 μL s−1 (or over 99% reduction in flow rate) and after 3.5 min the flow rate stabilized at 8 ± 1 μL s−1 (or 99.4 ± 1.2% closed). This behavior can be understood from the first step of the alginate/PNIPAAm ICE gel actuation test. When 60 °C water passes over the open valve the hydrogel is free to contract until it caps the end of the flow cannel. Once capped the hydrogel continues to undergo constrained deswelling until a maximum stress is produced. The flow of water through the valve decreases until this maximum stress is achieved. Modifications to the valve design to improve the sealing surface are likely to allow complete 100% closure. In summary, we have designed a new ink for 3D printing of hydrogels that are both mechanically robust and thermally actuating. By modifying the amount of thermally responsive PNIPAAm network in the hydrogels, the gels showed reversible length changes of 41%–49% when heated and cooled between 20 and 60 °C. Blocked stresses generated in tension were in the range of 10–21 kPa. We were able to fabricate a smart valve that controls the flow of water by printing a dynamic alginate/PNIPAAm ICE gel ink alongside other static materials. The valve automatically closed upon exposure to hot water, reducing the flow rate by 99%, and opened in cold water. With CAD modeling this 4D printing technique can be easily extended to make other types of moving structures. The ability to 4D print robust, actuating hydrogel materials opens a new avenue for fabricating hydrogel-based sensors, soft robots, medical devices, and self-assembling structures. Materials: All materials were used as-received and all solutions were prepared using Milli-Q water (resistivity = 18.2 Ω cm). α-Keto glutaric acid photoinitiator was purchased from Fluka (Australia). Acrylamide solution (40%, for electrophoresis, sterile-filtered), alginic acid sodium salt (from brown algae with Brookfield viscosity 2% in H2O at 25 °C of 250 mPa s), calcium chloride (minimum 93.0% granular anhydrous), ethylene glycol (rheology modifier), N-isopropylacrylamide and crosslinker were purchased from (Australia). A epoxy based was purchased from (Australia). Hydrogel A range of PNIPAAm hydrogel solutions with NIPAAm concentrations between 10% and 20% (w/v) were prepared by between and of NIPAAm, 15 of and of α-Keto glutaric acid in 15 mL of m calcium Alginate/PNIPAAm ICE gel solutions were prepared by of alginic acid sodium salt to the PNIPAAm hydrogel solutions with The ICE gel solutions were then to for use as extrusion ICE gel ink was prepared using a previously The ink was prepared by mL of the of and of glutaric acid in mL of Milli-Q water, mL of the 0.1 m calcium chloride and mL of ethylene A was then used to of sodium salt into the printing alginate-based ink for printing sacrificial support structures was also prepared using a previously The alginate ink was prepared by mL of the 0.1 m calcium chloride solution and mL of ethylene glycol in mL of Milli-Q water. A was then used to of sodium salt into the printing Alginate/PNIPAAm ICE hydrogels and a valve were fabricated with a 3D-Bioplotter that was coupled with a A UV using a UV source with a 1 m was used to the printed of tensile and the valve were prepared with computer-aided design software software was used to the model into a of to determine the print and internal structures were then to each within the The extrusion inks were into the 3D-Bioplotter in which were into the temperature controlled print Each was with a 23 and maintained at temperature of 10 °C during printing. A between and 1 was used to the inks and a between 5 and 20 mm s−1 was used to the inks with a mm a that was cooled to 10 °C. The UV was passed over each of ink for 50 and for the final A was used to of the alginate-based ink for printing sacrificial support structures the print was Finally, the printed ICE hydrogels were immersed in 0.1 m calcium chloride to fully crosslink the Mechanical The mechanical properties of the printed alginate/PNIPAAm ICE hydrogels were determined using a Universal Mechanical Tensile tests were performed using a 50 load tensile were to at a rate of 10 mm The stress was calculated using the average cross-sectional area of the unstrained specimen at the The stress and strain were calculated from the of of was calculated as the area under the stress–strain curve to tests were performed on each material with from one The of an external load on the actuation performance was investigated using a Universal Mechanical Tester with a 50 load and water sleeve. tensile were clamped to the mechanical tester inside a sleeve that was filled with 20 °C water. The force was measured over time as each specimen was extended to a stroke of 1 mm at = to ensure that the hydrogels were under The water was drained from the sleeve the hydrogel and replaced with water heated to 60 °C in min while the stroke was maintained at 1 mm for min. The behavior of the hydrogels was then determined by the stroke between 1 mm and a stroke with an stress of zero at a stroke rate of 10 mm for a of three Finally, the stroke was maintained at 1 mm for a further 120 min and the water sleeve was drained and refilled with 20 °C water. The performance was characterized by measuring the flow rate of water through the valve. The flow rate was calculated by measuring as the time required for a 5 mL column of water to pass through the valve. The valve was closed by passing 60 °C water over the hydrogel actuators for up to 5 min and opened by passing 20 °C water over hydrogel actuators. The the for partial support of this work through of and and The also the for to 3D As a to and this by the materials are and be for but are not or support from than be to the The is not for the content or of by the than be to the corresponding for the