Tailoring Physical Properties of Dual-Network Acrylamide Hydrogel Composites by Engineering Molecular Structures of the Cross-linked Network.

We demonstrate the impact of engineering molecular structures of poly(acrylamide) (PAAm) and poly(N-isopropylacrylamide) (PNIPAm) hydrogel composites on several physical properties. The network structure was systematically varied by (i) the type and the concentration of difunctional cross-linkers and (ii) the type of native or chemically modified natural polymers, including sodium alginate, methacrylate/dopamine-incorporated porcine skin gelatin and fish skin gelatin, and thiol-incorporated lignosulfonate, which are attractive biopolymers generated in pulp and food industries because of their abundance, rich chemical functionalities, and environmental friendliness. First, we added cross-linking agents of varying lengths at different concentrations to assess how the cross-linking agent modulates the mechanical properties of acrylamide-based composites with alginate. After chemically modifying gelatins from fish or porcine skin with methacrylate and/or dopamine, the acrylamide-based composites were fabricated with the chemically modified gelatins and thiolated lignosulfonate to assess the stress-strain behavior. Furthermore, swelling ratios were measured with respect to temperature change. The mechanical properties were systematically modulated by the changes in the molecular structure, that is, the length of the chemical unit between two end alkene groups in the difunctional cross-linker and the types of the additive natural polymers. Overall, PAAm hydrogel composites exhibit a significant, negative correlation between toughness and the volume fraction of the swollen state and between strain at fracture and the volume fraction of the swollen state. In contrast, PNIPAm hydrogel composites showed positive, but only moderate correlations, which is attributed to the difference in the network polymer structure.

Introduction

Hydrogel is a soft material with a covalently or physically cross-linked three-dimensional network structure comprising of polymeric chains capable of holding water over 30% of its weight.[1] As the hydrogel mimics the nature of biological tissues with highly hydrated polymer chains, the application of hydrogels with appropriate chemical modifications to attain desired functionalities has attracted much attention in tissue engineering, drug delivery, wound dressing, medical implant, and injury biomechanics.[2−8] Despite its advantages, the application of hydrogels needs chemical modification to overcome the limitation of inherently weak mechanical nature, attributed to the limitation of molecular motion of the polymer chains in the cross-linked network structure.[9]

Dual polymeric components in the hydrogel can synergistically compensate this shortcoming by which the first polymeric component confers a cross-linking network as a backbone and the second component reinforces to provide several desirable properties such as elasticity and toughness.[10−14] For example, the use of sodium alginate and cellulose was reported to fabricate tough hydrogels with poly(acrylamide) (PAAm),[15,16] exhibiting approximately a 20-fold increase of the length by stretching out with a high fracture energy close to 9000 J/m2, which is completely different from the PAAm homopolymer hydrogel.[15] In addition, the post-fabrication treatment with multivalent ion solutions, which induces strong interchain interaction of sodium alginate chains, further provides the means to systematically change the physical properties in hydrogels.[17,18] Another example is a poly(N-isopropylacrylamide) (PNIPAm) hydrogel, widely utilized for its thermo-sensitivity, biocompatibility, and flexibility, though it typically shows a low strength and modulus.[19−21] These shortcomings of physical properties can be effectively mitigated by achieving dual-network hydrogels with interpenetrating networks that can allow additional cross-links by covalent, ionic, or hydrogen bonds or by hydrophobic interactions.[22]

The physical properties of dual-network hydrogels are typically governed by several parameters, that is, cross-linking density, chemical or physical cross-linking, structure of a cross-linker, and chemical nature of filler materials. In fabrication of acrylamide-based hydrogels, typically a difunctional cross-linker is added for cross-linking. It has been known that the properties of cross-linked materials have strong dependencies on cross-linking density, swollen state, molecular weight, distance between two reactive functional groups in the cross-linker, molecular structure of cross-linkers, and nature of cross-linking.[23−27] Filler materials with a cross-linking ability can confer different material properties of the end product with respect to the extent of the interaction between the filler materials and the cross-linked end product. Accordingly, chain conformations and cross-linking density could be important design features to modulate mechanical properties, for example, elasticity and toughness of the cross-linked end product. Therefore, it is essential to understand the molecular structure–property relationships by systematic studies on the aforementioned parameters to attain dual-network hydrogels toward target applications where the specific mechanical properties such as modulus, elasticity, and adhesion strength are required: for example, hydrogels for the soft tissues including scaffolds that should bear loaded force and withstand shear in service,[28] strong and enzymatically degradable adhesives for biomedical applications,[29−31] and a structural component that can block the flow of water in the pipeline.[32]

We report the variation of molecular structures in PAAm and PNIPAm hydrogel composites and the modulation in the mechanical properties by varying the concentration and type of cross-linkers (Figure a). We hypothesized that the swollen state, the length between the cross-linking points, and the nature of bonding (weak physical interaction or covalent chemical bonding) modulate the mechanical properties of dual-network acrylamide hydrogel composites. To form a dual-network acrylamide hydrogel composite with fillers, we incorporated several natural polymers including sodium alginate (polysaccharide), gelatins from porcine skin and fish skin (protein), and lignosulfonate with thiolation (phenolic polymer). Since the application of these additive natural polymers is advantageous due to their abundance, biocompatibility, and biodegradability, their applicability has been largely expanded in various fields including biomedical applications such as scaffold materials, wound dressing, antioxidant materials, and drug delivery devices.[33−36] Chemical modifications were performed on the polymeric backbone to incorporate reactive groups such as methacrylate to gelatin, 3,4-dihydroxyphenethylamine (dopamine) to methacrylated gelatin, or sulfhydryl to lignosulfonate. The dual-network acrylamide hydrogel composites with the chemically modified gelatin and/or lignosulfonate were formed to enhance Young’s modulus or toughness (the area under the stress–strain curve). Our study showed that the mechanical properties of the acrylamide hydrogel composites can be tuned by chemical cross-linking or physical entanglement of acrylamide polymers with chemically modified gelatins and/or thiolated lignosulfonate (TLS). While PAAm hydrogel composites exhibited a high to very high correlation between toughness and the volume fraction of the swollen state, swelling ratios of PAAm hydrogel composites were insensitive to temperature changes. In contrast, PNIPAm hydrogel composites showed the inherent thermo-responsiveness of PNIPAm as evidenced by the changes in swelling ratios with low to moderate correlations between toughness and the volume fraction of the swollen state.

Figure 1

(a) Chemical structures of PAAm/PNIPAm, cross-linkers, and additive natural polymers for hydrogel fabrication, and (b) scheme showing the cross-linking reaction. (c) Synthesis of GelMA and GelMA–DOPA, and (d) incorporation of thiol into lignosulfonate.

Materials and Methods

Materials

Alginic acid sodium salt from brown algae (Alg), porcine skin gelatin (pGel), cold water fish skin gelatin (fGel), N,N,N′,N′-tetramethyl ethylenediamine (TEMED, 99%), N,N′-methylenebis(acrylamide) (MBAA, 99%), ethylene glycol diacrylate (EGDA, 99%), ammonium persulfate (APS, ≥98%), glycidyl methacrylate (GMA, 97%), 4-dimethlaminopyridine (DMAP, 99%), dimethyl sulfoxide (DMSO), and dopamine hydrochloride (DOPA) were purchased from Sigma-Aldrich. Sodium lignosulfonate (SLS), diethylene glycol diacrylate (DEGDA, >75%), N,N′-ethylenebisacrylamide (EBAA, >97.0%), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), N-hydroxysuccinimide (NHS, 98%), 3-mercaptopropionic acid (MPA, >98%), and N-isopropylacrylamide (NIPAm, >98%) were obtained from TCI Chemicals. Sodium bicarbonate (NaHCO3, 99.0%), acrylamide (AAm, 98.5%), acrylic acid (AAc, 99.5%), isopropyl alcohol (IPA), and phosphate-buffered saline (PBS) solution (pH = 7.4) were purchased from Samchun Chemical Co., Ltd. Hydrochloric acid aqueous solution (HCl solution, 37%) was purchased from Fisher Scientific. Otherwise noted, all reagents were used without additional purifications.

Methacrylation of Porcine Gelatin and Fish Skin Gelatin (pGelMA and fGelMA)

Methacrylate incorporation was conducted using previously published protocols.[34] Briefly, either porcine or fish skin gelatin (10 g) was mixed in 100 mL of DMSO and stirred at 50 °C until the solution was fully dissolved. GMA (4 mL) and DMAP (0.6 g) were added to the gelatin/DMSO solution, and the reaction mixture was stirred at 50 °C for 2 days. Upon cooling down to room temperature, the mixture was subjected to dialysis (tubing molecular weight cutoff = 3500 Da) against DI water at 40 °C for 5 days to remove impurities. The solution was lyophilized for 4 days, resulting in a white powder. The reaction scheme is provided in Figure c.

Modification of Gelatin Methacrylate with Dopamine (pGelMA–DOPA and fGelMA–DOPA)

Incorporation of dopamine into gelatin methacrylate was performed with the protocol adapted from the literature.[37] Briefly, synthesized pGelMA or fGelMA (2.00 g) was dissolved in 100 mL of PBS solution. A mixture of EDC (2.33 g) and NHS (2.30 g) was added to GelMA, and then, DOPA (2.00 g) was added to the reaction mixture. The reaction was carried out at 40 °C for 7 h with vigorous stirring. The solution was cooled down to room temperature, and then, the crude product was subjected to dialysis (tubing molecular weight cutoff = 3500 Da) against HCl solution (pH = 4) for 2 days. The dialysis medium was changed to deionized water (pH = 7), and the solution was further dialyzed for 1 day. The resulting solution was then lyophilized for 2 days to obtain a gray fluffy sample. The reaction scheme is provided in Figure c.

Synthesis of TLS

TLS was synthesized by following the protocol published in the literature.[34] SLS (1.00 g) was dissolved in 10 mL of deionized water in a round bottom flask equipped with a magnetic stirring bar. A mixture of HCl 0.1 mL and MPA 1.0 mL was added to the SLS solution with N2 purging. The system was heated to the temperature of 80 °C with stirring, and the reaction proceeded for 24 h. The flask was cooled down to room temperature, and the solution was added to the excess IPA for precipitation. The precipitate was collected with centrifugation and redispersed in IPA for further washing. The washing process was repeated twice, and the solid TLS was isolated and collected, followed by drying under vacuum overnight. The reaction scheme is provided in Figure d.

Fabrication of Acrylamide Hydrogel Composites: Variation of Natural Polymer Component and Difunctional Cross-linkers

To fabricate an AAm composite, AAm (3.1 g, 50 wt % aqueous solution), TEMED (0.1 g, 10 wt % aqueous solution), and MBAA (0.1 g, 1 wt % aqueous solution) were mixed. For a NIPAm composite, NIPAm (2.4 g, 50 wt % aqueous solution), TEMED (0.1 g, 10 wt % aqueous solution), and MBAA (0.1 g, 1 wt % aqueous solution) were mixed. To form acrylamide composites with additional natural polymers, a natural polymer component (Alg, fGel, fGelMA, pGel, pGelMA, pGelMA–DOPA, or fGelMA–DOPA) of 9.9 g (2 wt % aqueous solution) was added to the solution of AAm or NIPAm monomer. A composite with TLS was formed by adding TLS (0.1 g, 10 wt % aqueous solution) to the solution of AAm or NIPAm monomer. The prepared solution of AAm or NIPAm, TEMED, MBAA, natural polymer, with or without TLS, was poured into a rectangular mold (40 mm × 5 mm × 5 mm or 60 mm × 40 mm × 35 mm). Finally, APS (0.15 g, 10 wt % aqueous solution) was added to the prepared solution in the mold and quickly mixed, followed by annealing in the oven at 40 °C for 20 h for cross-linking. For the variation of the difunctional cross-linker, MBAA or EBAA or a mixture of MBAA with either EGDA (MBAA/EGDA) or DEGDA (MBAA/DEGDA) (varying mole ratios of MBAA to either EGDA or DEGDA ranging from 1:1.2 to 1:10.5 or 1:1 to 1:9, respectively) was added to the solution prepared with the same amount of AAm, TEMED, and Alg described above, followed by adding into the same mold. The details regarding the amount of cross-linkers are summarized in Tables S1 and S2. The hydrogel composites were formed by quickly mixing APS (0.15 g, 10 wt % aqueous solution) with the prepared solution in the mold, and subsequent cross-linking in the oven at 40 °C for 20 h.

Characterizations

1H and 31P nuclear magnetic resonance (NMR) spectroscopy was performed using a JEOL JNMECZ-400S with D2O and chloroform-d to confirm the structure of products and to determine the degree of methacrylation in fGelMA or pGelMA, the degree of dopamine incorporation in fGelMA–DOPA or pGelMA–DOPA, and the thiolation efficiency in TLS synthesis. Fourier transform infrared (FT-IR) spectra of various resulting hydrogels were recorded by a PerkinElmer Frontier with a UATR, over the wavenumber range from 4000 to 640 cm–1 at room temperature. Tensile tests were performed using a universal testing machine (UTM, YEONJIN S-Tech, TXA) at room temperature under a crosshead speed of 12 mm/min and a load cell of 305.9 N. Rectangular prism (40 mm × 5 mm × 5 mm) samples were prepared for tensile tests.

Swelling Ratio Measurements of Acrylamide Hydrogel Composites

Swelling and deswelling behaviors of acrylamide hydrogel composites were studied with gravimetric analysis. Fabricated acrylamide hydrogel composites with different natural polymers (Alg, fGel, fGelMA, pGel, pGelMA, fGelMA–DOPA, and pGelMA–DOPA in the absence and presence of TLS) were cut into a rectangular shape (10 mm × 10 mm × 30 mm). Then, samples were dried in a vacuum oven overnight and the weight of dried hydrogel composites (Wd) was measured. Then, the dried hydrogel composites were immersed in water for 30 min at various temperatures from 22 to 40 °C with 2 °C intervals. At each temperature, the weight of the swollen hydrogel composite (Ws) was measured to calculate the swelling ratio (S) with the equation of S = (Ws – Wd)/Wd.

Statistical Analysis

Pearson’s correlation coefficients assuming Gaussian distribution and p values (α = 0.05) were assessed using GraphPad Prism 8.

Results and Discussion

Methacrylation of Gelatin (GelMA) and Conjugation of Dopamine to Methacrylated Gelatin (GelMA–DOPA)

The incorporation of methacrylate into gelatins was attained by the ring opening reaction of epoxy groups in GMA with a primary amine group of lysine in gelatins. DMSO was used as a solvent, which is known to be effective to prevent the undesired reaction of the epoxy group with water.[38] The degree of methacrylate incorporation for both fGel and pGel was determined via1H NMR in D2O (Figure a,b). As glycidyl groups of GMA react only with free amine groups of lysine in gelatin, the reaction efficiency can be obtained by direct comparison of the lysine peak at 2.9–3.1 ppm to the phenylalanine peak at 7.2–7.5 ppm, which is the internal reference.[39,40] As the methacrylate incorporation of both gelatins proceeded, the intensity of the lysine peak in the 1H NMR spectrum gradually decreased. With the quantitative analysis of the 1H NMR spectra, the efficiency of the methacrylate incorporation into fGel or pGel (=1 – [Ilys(GelMA)/Iphe(GelMA)]/[Ilys(Gel)/Iphe(Gel)], where Ilys and Iphe are the intensity of the lysine and the phenylalanine peak, respectively) was estimated to be 0.93 or 0.92, respectively.

Figure 2

(a) 1H NMR spectra of fGel, fGelMA, and fGelMA–DOPA, and (b) 1H NMR spectra of pGel, pGelMA, and pGelMA–DOPA.

fGelMA–DOPA and pGelMA–DOPA were synthesized by EDC/NHS coupling of the carboxyl group in gelatin with a primary amine group in dopamine. The extent of the reaction was monitored with 1H NMR spectra (Figure a,b), in which the aromatic peak at 6.7–6.9 ppm assigned to protons of phenyl group of dopamine was observed.[41] The integrated intensity of the peak was compared with the phenylalanine peak at 7.2–7.5 ppm to yield the amount of incorporated dopamine in the fGelMA and pGelMA samples. In comparison to the phenylalanine peak, the estimated molar ratios of dopamine to phenylalanine for fGelMA–DOPA and pGelMA–DOPA were 0.65 and 0.85, converted to the concentration of dopamine of 0.046 and 0.035 mmol/g, respectively, with the literature value of the phenylalanine concentration on fGel and pGel.[42]

Formation of Acrylamide Hydrogel Composites with Difunctional Cross-linkers

Acrylamide hydrogel composites were fabricated by adding AAm, Alg, TEMED, APS, and various cross-linking agents into a mold and cured at 40 °C for 20 h.[17] First, cross-linking agents with different lengths were used to confirm the effect of length between cross-linking points. Figure a shows the FT-IR spectra of PAAm hydrogel composites with different cross-linkers of MBAA, EBAA, MBAA/EGDA, and MBAA/DEGDA, respectively. The characteristic peaks of the PAAm hydrogel composites were shown, attributed to N–H (≈3200 and ≈1620 cm–1), C–H (≈3200 cm–1), and C=O (≈1670 cm–1).[43,44] In PNIPAm hydrogel composites, broad bands of N–H (3050–3300 cm–1) were observed (Figure c). Furthermore, the characteristic bands of C–H (2870–3000 cm–1), C=O (1630 cm–1), and N–H (≈1530 cm–1) were observed.[45,46] The addition of Alg, fGel, or pGel as a filler was evidenced by the peak shape difference due to O–H (≈3420 cm–1), C–O (1456 cm–1), and C=O (1559 cm–1) (Figure b). The expected characteristic peaks in FT-IR spectra were all observed in acrylamide hydrogel composites with native and chemically modified natural polymers (Figure S1), suggesting that the structural variation of cross-linker and the addition of natural polymers do not prevent the formation of hydrogel composites via radical polymerization.

Figure 3

FT-IR spectra of (a) PAAm hydrogel composites with various cross-linking agents (MBAA, EBAA, MBAA/EGDA, and MBAA/DEGDA), (b) PAAm hydrogel composites with Alg, fGel, and pGel, and (c) PNIPAm hydrogel composites with Alg, fGel, and pGel.

Modulation of Mechanical Properties of Alg/PAAm with Difunctional Cross-linkers

A difunctional cross-linker significantly modulates the mechanical properties of acrylamide hydrogel composites.[24] We first investigated the mechanical properties of Alg/PAAm hydrogel composites fabricated with four different difunctional cross-linkers, MBAA, EBAA, MBAA/EGDA, and MBAA/DEGDA. In Alg/PAAm, the mechanical properties of the Alg/PAAm hydrogel composite was enhanced by physical cross-links between amine groups on PAAm chains and carboxyl groups on Alg chains.[47] With this system, we hypothesized that the molecular changes in the difunctional cross-linker modulate the elastic modulus of acrylamide hydrogel composites. As depicted in Figure a, the molecular weight between two terminal alkenes increases. The strain–stress curve (sample dimension = 5 mm × 5 mm × 40 mm; strain rate = 0.2 mm/s) was changed dramatically when the amount of MBAA added into Alg/PAAm hydrogel composite was increased from 1 to 4 mg (Figure S2a and Table S1). As shown in Figure S2a and Table , the strain at fracture and the stress at fracture were decreased to 4.42 and 20.62 kPa, respectively. Further increases of the amount of MBAA added into an Alg/PAAm hydrogel composite up to 4 mg reduced the strain at fracture to 2.68 and the stress at fracture to 18.47 kPa. The modulus of elasticity was increased from 2.63 to 10.09 kPa with respect to the increased amount of MBAA from 1 to 2 mg, while the toughness (the area under the stress–strain curve) was decreased from 409.65 to 34.09 kJ/m3 with respect to the increase of the amount of MBAA from 1 to 4 mg. All of these changes are attributed to the increase of the number of bridging points formed by difunctional cross-linkers, and hence, the increase of the density of chemical cross-linking.[24]

Table 1

Stress or Strain at Fracture, Toughness, and Modulus of Hydrogel Composites Fabricated with Various Concentrations of MBAA and EBAA

sample	strain at fracture (−)	stress at fracture (kPa)	toughness (kJ/m3)	modulus (kPa)
MBAA 1.0			409.65	2.63
MBAA 2.0	4.42	20.62	60.91	6.49
MBAA 3.0	3.60	18.98	42.63	6.84
MBAA 4.0	2.68	18.47	34.09	10.09
EBAA 1.1			299.88	3.13
EBAA 2.2	9.29	41.11	218.86	5.15
EBAA 3.3	3.74	19.62	53.26	7.07
EBAA 4.4	3.47	16.90	39.44	7.13

EBAA exhibited different stress–strain curves in comparison to MBAA due to a slightly higher molecular weight between the two terminal alkenes than MBAA. As shown in Figure S2b, Alg/PAAm hydrogel composites with 1.1 mg of EBAA exhibited a similar shape of stress–strain curves in comparison to those with MBAA; however, when the amount of EBAA was doubled to 2.2 mg, the strain and the stress at fracture decreased to 9.29 and 41.11 kPa, respectively (Table ). With the increase of the amount of EBAA up to 4.4 mg, the shape of stress–strain curves became similar to that of Alg/PAAm hydrogel composites with higher mass fractions of MBAA in Figure S2.

Alg/PAAm hydrogel composites fabricated with DEGDA were instantaneously deformed and did not form self-supporting hydrogel composites, making the characterization with the UTM challenging. Thus, MBAA with varying ratios of either EGDA (MBAA/EGDA) or DEGDA (MBAA/DEGDA) was introduced to assess the effect of difunctional cross-linker composition and molecular structures (Table S2). The molecular weights of EGDA and DEGDA are 170.16 and 214.22 g/mol, respectively. The ratio of MBAA to EGDA or DEGDA was varied from 1:1.2 to 1:10.5 or from 1:1 to 1:9, respectively. As shown in Figure S3, the stress–strain curves obtained as a function of the amount of EGDA did not result in significant changes of the stress–strain curves although the stress–strain curves with EGDA are comparable to the Alg/PAAm hydrogel composite with 1 mg of MBAA in Figure S2a. In other words, strain and stress at fracture of acrylamide hydrogel composites were enhanced when MBAA/EGDA were used together in comparison to the acrylamide hydrogel composites cross-linked with MBAA alone. For example, modulus of Alg/PAAm (MBAA 1.0) was 1.69 kPa (Table S3), while that of Alg/PAAm (MBAA and EGDA) was typically higher than 3.00 kPa (Table S2). These results indicate that the addition of EGDA to Alg/PAAm hydrogel composites cross-linked with MBAA alone can withstand higher strains. Furthermore, it was shown that increasing the ratio of EGDA with respect to MBAA affects Young’s modulus of acrylamide hydrogel composites in a relatively smaller range from 2.70 to 3.34 kPa (Table S2 and Figure S3a). Similar trends in the use of MBAA/DEGDA were also observed (Figure S3b). Increasing the ratio of DEGDA with respect to MBAA affects Young’s modulus of hydrogel composites in the range from 1.94 to 3.56 kPa (Table S2), while the trend was not completely linear. For example, the highest modulus of elasticity is at the ratio of MBAA/DEGDA (1:7) and the lowest at MBAA/DEGDA (1:3). Apparently, the acrylamide hydrogel composites with MBAA/DEGDA were generally able to withstand deformation to a similar extent as Alg/PAAm hydrogel composites could with MBAA/EGDA. In summary, these results suggest that a cross-linker with lower molecular weights (shorter lengths between two end alkenes) more directly (or linearly) impacts the mechanical properties of Alg/PAAm hydrogel composites than a cross-linker with higher molecular weights.

Incorporation of Alg, Gel, GelMA, and GelMA–DOPA into Acrylamide Hydrogel Composites

Mechanically tough and elastic acrylamide hydrogel composites are attainable with the addition of sodium alginate.[47] Thus, we hypothesized that the mechanical properties of acrylamide hydrogels can also be varied with Gel, GelMA, and GelMA–DOPA. Carboxyl groups in Alg and Gel interact with amino groups in PAAm via hydrogen bonds. As shown in Figure a, stress–strain curves of Alg/PAAm and pGel/PAAm are overlapped at lower strain, while the strain at fracture of pGel/PAAm is decreased to 12.87. This result confirmed that Alg interacts with amino groups in PAAm to a greater extent in comparison to Gel due to the presence of carboxyl groups, showing enhanced mechanical properties. Young’s modulus of fGel/PAAm is lesser than that of Alg/PAAm or pGel/PAAm hydrogel composites, which is likely attributed to the presence of lower number of proline and hydroxyproline and the lower molecular weight of fGel than those of pGel.[27,48] When fGelMA or pGelMA was incorporated, the stress–strain curves of PAAm hydrogel composites were dramatically changed (Figure b). First, Young’s modulus of fGelMA/PAAm or pGelMA/PAAm increased to 8.10 and 3.53 kPa, respectively, while the strain at fracture was largely decreased to 1.76 and 6.11, respectively. Since methacrylate groups on GelMA chains participate in the cross-linking reaction, the cross-linking density was apparently increased. Therefore, the hydrogel composites with GelMA became brittle as Young’s modulus was increased.[39] However, incorporation of fGelMA–DOPA or pGelMA–DOPA to PAAm adversely affected the final mechanical properties of PAAm hydrogel composites, as shown in Figure c. Both fGelMA–DOPA/PAAm and pGelMA–DOPA/PAAm exhibited stress or strain at fracture far less than those of fGelMA/PAAm or pGelMA/PAAm except the strain at fracture of fGelMA and fGelMA–DOPA (1.76 for both cases). Since two phenols in dopamine effectively interact with polar groups of PAAm and GelMA via hydrogen bonding, incorporation of dopamine possibly results in the increase of physical cross-linking density. However, the radical-based cross-linking can be partially inhibited by phenols in the dopamine group, which results in a hydrogel composite with low Young’s modulus (stiffness). It is expected that the presence of dopamine in the hydrogel composites offers an improvement of adhesion properties;[49] however, the PAAm hydrogel composites with both fGelMA–DOPA and pGelMA–DOPA did not exhibit an improvement of adhesion, suggesting that the dopamine groups interact with different functionalities in the matrix of PAAm.[50] As summarized in Table S3, the incorporation of methacrylate reduces the toughness of PAAm hydrogel composites (fGel or pGel to fGelMA or pGelMA, respectively, in PAAm), and the additional conjugation of dopamine further reduces the toughness of PAAm hydrogel composites (fGelMA or pGelMA to fGelMA–DOPA or pGelMA–DOPA, respectively, in PAAm).

Figure 4

Stress–strain curves of PAAm and PNIPAm hydrogel composites fabricated using native and chemically modified natural polymers. (a) PAAm hydrogels with Alg, fGel, and pGel, (b) PAAm hydrogels with fGelMA and pGelMA, (c) PAAm hydrogels with fGelMA–DOPA and pGelMA–DOPA, (d) PNIPAm hydrogels with Alg, fGel, and pGel, (e) PNIPAm hydrogels with fGelMA and pGelMA, and (f) PNIPAm hydrogel with pGelMA–DOPA.

We further examined the effect of incorporating GelMA and GelMA–DOPA into PNIPAm with the same method and conditions used for PAAm hydrogel composites. PNIPAm hydrogel composites showed decreased strain at fracture and toughness in comparison to PAAm hydrogel composites. In Figure d, while Alg/PNIPAm and fGel/PNIPAm hydrogel composites showed comparable mechanical properties to each other, pGel/PNIPAm hydrogel composites showed significantly higher toughness and about three times higher stress at fracture than Alg/PNIPAm or fGel/PNIPAm hydrogel composites. The less charged chemical nature of pGel compared to Alg could be physically entangled with more hydrophobic PNIPAm matrix,[51] improving Young’s modulus and toughness of the pGel/PNIPAm hydrogel composite. In contrast, fGel/PNIPAm hydrogel composites did not improve the mechanical properties when compared to the Alg/PNIPAm hydrogel composites, indicating that the reinforcement of the mechanical properties with fGel in the PNIPAm matrix is less synergistic due to the lower molecular weight of fGel.[27] Interestingly, methacrylation of gelatin affected the mechanical properties of PNIPAm hydrogel composites in a different way when compared to PAAm hydrogel composites. Addition of fGelMA or pGelMA to PNIPAm hydrogels exhibited increased strains at fracture at 2.14 and 1.79, respectively (Figure e and Table S3). Finally, the incorporation of pGelMA–DOPA into the PNIPAm hydrogel showed a decrease of strain at fracture and an increase of Young’s modulus in comparison to pGelMA in PNIPAm (Figure f), which was similar to the case of pGelMA–DOPA PAAm hydrogel (Figure c). The cross-linking reaction of the fGelMA–DOPA/PNIPAm hydrogel composite was not successful in forming a freestanding composite partially due to phenols in dopamine, which inhibited the cross-linking reaction. Furthermore, the chain length of fGelMA–DOPA is not likely long enough to make an effective cross-linking network for enough mechanical strength,[27,52] while the cross-linking was feasible with pGelMA–DOPA having a long enough chain length.

Incorporation of TLS into Acrylamide Hydrogel Composites

Sulfonated lignin is a water-soluble, non-linear, and cluster-like natural polymer with several chemical functionalities for additional applications.[53] TLS bears multiple sulfhydryl groups, enabling us to utilize the thiol–ene cross-linking reaction.[34,36] Therefore, TLS plays the role of nanoparticle cross-linker and filler in a hydrogel composite. Incorporation of TLS into PAAm and PNIPAm hydrogels with Alg, Gel, GelMA, and GelMA–DOPA (Figure b,d, summarized in Table S4) exhibited largely similar trends of stress–strain curves in comparison to the acrylamide hydrogel composites without TLS (Figure a,c, summarized in Table S3).

Figure 5

Tensile stress–strain curves with various native and chemically modified natural polymers in the matrix of (a) PAAm, (b) TLS–PAAm, (c) PNIPAm, and (d) TLS–PNIPAm hydrogel composites. Note that all curves in Figure are included in (a,c) and replotted for vis-à-vis comparison.

Incorporating TLS into Alg and Gel made significant decreases in toughness and strain at fracture (Figure b). The structure of TLS (Figure d) is highly branched and complex, which suggests the role of the cluster-like filler in the hydrogel matrix. Moreover, hydrophilic sulfonate groups may enable the interaction with AAm monomers of the PAAm matrix. Therefore, TLS is distributed homogeneously in the hydrogel composites with Alg or Gel, resulting in comparable toughness to each other with TLS. The addition of TLS (TLS–PAAm hydrogel with fGelMA and pGelMA) showed two to three-fold increase of toughness in comparison to fGelMA–PAAm and pGelMA–PAAm hydrogel composites without TLS, attributable to the covalent cross-linking between the methacrylate group in GelMA and the sulfhydryl groups of TLS. In contrast, incorporating TLS into the GelMA–DOPA/PAAm hydrogel composites decreased toughness and other mechanical properties, except Young’s modulus of pGelMA–DOPA (4.25 kPa pGelMA–DOPA/PAAm vs 7.48 kPa pGelMA–DOPA/TLS–PAAm).

However, NIPAm monomers are relatively more hydrophobic than AAm monomers, which may lead to clustering of TLS in the PNIPAm matrix. Consequently, relatively obscure mechanical properties appear. As shown in Figure d, Alg/PNIPAm and Gel/PNIPAm hydrogel composites showed slightly increased strain–stress at fracture and toughness with the addition of TLS. Incorporating TLS into fGelMA/PNIPAm decreased strain–stress at fracture and toughness, while Young’s modulus increased; however, incorporating TLS into pGelMA/PNIPAm exhibited higher toughness and Young’s modulus when compared to the hydrogel composites without TLS. It confirms that the methacrylate group in GelMA improved the cross-linking density with TLS in the matrix of both PAAm and PNIPAm. pGelMA/TLS exhibited improved toughness in both PAAm and PNIPAm hydrogels when compared to that of fGelMA/TLS. It is likely attributed to the higher density of cross-linking between thiol groups in TLS and methacrylate in pGelMA than fGelMA, showing in the literature that the number of proline and hydroxyproline in the chains of Gel and the molecular weight of Gel chains are higher in pGel than in those of fGel.[27,48] The mechanical properties summarized in Tables S3 and S4 were visualized using a heat map in Figure .

Figure 6

Heatmap (matrix) representation of mechanical properties. Native and chemically modified natural polymers are listed in rows, and PAAm, TLS–PAAm, PNIPAm, and TLS–PNIPAm are listed in columns. Blue colored rectangles represent lower values, and gray colored rectangles represent no data.

Swelling Ratio and Its Correlation to the Mechanical Properties of the Composites

We assessed swelling ratios of PAAm and PNIPAm hydrogel composites with Alg, Gel, GelMA, and GelMA–DOPA with and without TLS (Figure ). PNIPAm is known to exhibit lower critical solution temperature (LCST), approximately at which the hydrogel becomes hydrophobic, and hence, phase separation occurs at 32 °C. As expected, PAAm hydrogel composites with and without TLS did not show any changes in swelling ratio with respect to temperature changes (Figure a,b). In contrast, PNIPAm hydrogel composites clearly exhibited a decrease of swelling ratio. In the absence of TLS, the swelling ratio varied at or slightly below 32 °C as the type of incorporated natural polymers (Alg or Gel) was changed (Figure c,d). For both PAAm and PNIPAm hydrogel composites, fGelMA or pGelMA further reduced swelling ratios when compared to fGel or pGel, indicating that a higher cross-linking density limits the extent of swelling. The incorporation of fGelMA or pGelMA into the matrix of PNIPAm maintained the phase transition of PNIPAm at or around 32 °C (Figure c). Interestingly, the incorporation of TLS led to a slightly lesser extent of variation in swelling ratio and maintained the phase transition at or around the temperature of 34–35 °C except for the Alg/TLS–PNIPAm hydrogel composite (Figure d). This is possibly due to the noncovalent interactions between the carboxylate group of Alg and the multiple functional groups of TLS, resulting in reduced or limited interactions of Alg/TLS with water molecules in hydrogel composites when compared to those including Gel.[54] PNIPAm hydrogel composites without TLS exhibited swelling ratios comparable to the rest of the TLS–PNIPAm composites (except Alg/TLS–PNIPAm), and hence leading to a decrease in LCST below 34 °C.

Figure 7

Plots of swelling ratio as a function of temperature with various native and chemically modified natural polymers in the matrix of (a) PAAm, (b) TLS–PAAm, (c) PNIPAm, and (d) TLS–PNIPAm hydrogel composites.

Furthermore, we attempted to correlate the mechanical properties with a specific parameter that modulates the network of acrylamide hydrogel composites. Cross-linking density (or average molecular weight between cross-linking points) is directly related to Vs by the Flory–Rehner theory.[55] Of all the parameters tested and summarized in Tables S3 and S4, toughness as a function of the volume fraction of the acrylamide hydrogel composites in the swollen state (Vs = 1/S, where S is the swelling ratio) are shown in Figure . Since Vs is the reciprocal of the swelling ratio, the general trend of Vs increases from Alg to GelMA and GelMA–DOPA. Correlations with respect to strain–stress at fracture, toughness, and Young’s modulus are summarized in Table S5. The correlation coefficient of the PAAm hydrogel composites shows that the correlation between toughness and Vs is very high (PAAm vs toughness) or high (TLS–PAAm vs toughness) and is associated negatively, suggesting that the incorporation of physically and chemically cross-linkable units, such as carboxylate, amine, alcohol, methacrylate, and phenol in native and chemically modified natural polymers, into PAAm was effective to modulate the toughness of the PAAm hydrogel composite. Interestingly, the correlation between toughness and Vs of the PNIPAm hydrogel composites is low (PNIPAm vs toughness) or moderate (TLS–PNIPAm vs toughness) and positive; that is, the amide group of PNIPAm interacts weakly with the physically cross-linkable units present in native and chemically modified natural polymers in comparison to the amide group of PAAm. In the matrix of PAAm, the negative correlation between toughness and Vs implicates the weaker cross-linking in PAAm hydrogel composites as the complexity or the degree of functionalization of natural polymers increases (Figure a). In contrast, the moderate and positive correlation between toughness and Vs was observed in the matrix of PNIPAm. This is probably due to the thermo-responsiveness of PNIPAm, which confers the chain flexibility of NIPAm even in the presence of native and chemically modified natural polymers (Figure c). The toughness of TLS–PAAm and TLS–PNIPAm hydrogel composites shows high and moderate correlations, respectively, with respect to Vs (Figure b,d). It is also notable that the incorporation of TLS into both PAAm and PNIPAm matrices changed the slope of correlations in comparison to those without TLS. These observations imply that covalent bonding and noncovalent interactions such as hydrogen bonds or dipolar interactions conferred by the incorporation of TLS into acrylamide hydrogel composites affect the toughness of PAAm and PNIPAm hydrogel composites. Interestingly, PNIPAm hydrogel composites exhibited low correlation between toughness and Vs compared to TLS–PNIPAm composites. Similar trends for all acrylamide hydrogel composites with and without TLS were observed in the correlations between strain at fracture and Vs (Figure S4), where the correlation between the strain at fracture of TLS–PAAm vs Vs was only found to be significant (Figure S4b). The physical interaction increases via more hydrogen bonds between phenols of TLS and amine groups of acrylamide, and the chemical cross-linking increases between the sulfhydryl groups of TLS and alkene of acrylamide or methacrylate of GelMA, thereby showing the trend of forming tougher PNIPAm hydrogel composites than those without TLS. Interestingly, this similar increase was also observed in the matrix of PNIPAm, as evidenced by the increase of correlation coefficient from 0.353 (without TLS, Figure S4c) to 0.403 (with TLS, Figure S4d). The correlation between stress at fracture and Vs shows similar but obscure trends to the correlations of toughness and strain at fracture without any statistical significance (Table S5). The correlation between Young’s modulus and Vs also shows very low to moderate correlations without statistical significance. These results suggest that the correlation between toughness and Vs and that between strain and Vs implicate that Vs could be used to predict mechanical behaviors of acrylamide-based hydrogel composites regardless of the type of the filler materials. For ultimate and ideal prediction of the properties of the complex composite hydrogel systems, the fundamental understanding in depth is required, for instance, relation between the porous morphologies examined carefully with the well-established experimental technique[56−58] and physical properties and the cross-linking density and other related parameters to Vs and further connection to the molecular structure and physical or chemical cross-linking mechanism.[55,59]

Figure 8

Toughness values as a function of the volume fraction of the polymer in the swollen state (Vs) with various native and chemically modified natural polymers in the matrix of (a) PAAm (Pearson’s r = −0.8270, *p < 0.05), (b) TLS–PAAm (Pearson’s r = −0.7148, p > 0.05), (c) PNIPAm (Pearson’s r = 0.3177 for toughness, p > 0.05), and (d) TLS–PNIPAm (Pearson’s r = 0.5048 for toughness, p > 0.05).

Conclusions

We assessed the impact of engineering molecular structures of PAAm and PNIPAm hydrogel networks on their physical properties. For the variation of the molecular structure, we systematically changed the structure of a cross-linker from difunctional small molecules to sustainable natural polymers without or with chemical modifications. We first examined the effect of difunctional cross-linker length/molecular weight at different concentrations and ratios in Alg/PAAm hydrogels. We found that a cross-linker with lower molecular weights (shorter lengths between two end alkenes) more directly impacts the mechanical properties of Alg/PAAm hydrogel composites than a cross-linker with higher molecular weight. Then, we investigated the effect of incorporating Alg, Gel, GelMA, and GelMA–DOPA into the matrix of PAAm and PNIPAm hydrogels in the absence and presence of TLS on mechanical properties, swelling ratios, and temperature responsiveness. We decided to incorporate the sustainable materials that are largely wasted in pulp and food industries, which can be used as alternatives to petrochemical feedstock. With thorough examination of the stress–strain curves of these acrylamide hydrogel composites and further analyses acquiring stress–strain at fracture, toughness, and Young’s modulus, the effects of the type of the polymer and the chemical modification were elucidated. Furthermore, we correlated the observed physical properties of all composite hydrogels to Vs, which is directly related to the cross-linking density. The negative correlation between toughness and Vs in the matrix of PAAm implicates the weaker cross-linking in PAAm hydrogel composites as the complexity or the degree of functionalization of natural polymers increases. In contrast, the moderate and positive correlation between toughness and Vs was observed in the matrix of PNIPAm. These results provide the insight for a structure–property relationship in acrylamide hydrogels fabricated with different cross-linkers ranging from small molecules to macromolecules, and with the addition of native or chemically functionalized natural polymers with desired functionalities.