Bioinspired Hydrogels with Muscle-Like Structure for AIEgen-Guided Selective Self-Healing

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE1 Apr 2021Bioinspired Hydrogels with Muscle-Like Structure for AIEgen-Guided Selective Self-Healing Xiaofan Ji†, Zhao Li†, Yubing Hu, Huilin Xie, Wenjie Wu, Fengyan Song, Haixiang Liu, Jianguo Wang, Meijuan Jiang, Jacky W. Y. Lam and Ben Zhong Tang Xiaofan Ji† *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Department of Chemistry, Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction and Institute for Advanced Study, and Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong School of Chemistry and Chemical Engineering, Huazhong University of Science and Technology, Wuhan 430074 HKUST Shenzhen Research Institute, Shenzhen 518057 †X. Ji and Z. Li contributed equally to this work.Google Scholar More articles by this author , Zhao Li† *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Department of Chemistry, Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction and Institute for Advanced Study, and Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong Institute of Engineering Medicine, Beijing Institute of Technology, Beijing 100081 HKUST Shenzhen Research Institute, Shenzhen 518057 †X. Ji and Z. Li contributed equally to this work.Google Scholar More articles by this author , Yubing Hu Department of Chemistry, Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction and Institute for Advanced Study, and Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong HKUST Shenzhen Research Institute, Shenzhen 518057 Google Scholar More articles by this author , Huilin Xie Department of Chemistry, Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction and Institute for Advanced Study, and Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong HKUST Shenzhen Research Institute, Shenzhen 518057 Google Scholar More articles by this author , Wenjie Wu Department of Chemistry, Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction and Institute for Advanced Study, and Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong HKUST Shenzhen Research Institute, Shenzhen 518057 Google Scholar More articles by this author , Fengyan Song Department of Chemistry, Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction and Institute for Advanced Study, and Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong HKUST Shenzhen Research Institute, Shenzhen 518057 Google Scholar More articles by this author , Haixiang Liu Department of Chemistry, Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction and Institute for Advanced Study, and Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong HKUST Shenzhen Research Institute, Shenzhen 518057 Google Scholar More articles by this author , Jianguo Wang Department of Chemistry, Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction and Institute for Advanced Study, and Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong HKUST Shenzhen Research Institute, Shenzhen 518057 Google Scholar More articles by this author , Meijuan Jiang Department of Chemistry, Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction and Institute for Advanced Study, and Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong HKUST Shenzhen Research Institute, Shenzhen 518057 Google Scholar More articles by this author , Jacky W. Y. Lam Department of Chemistry, Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction and Institute for Advanced Study, and Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong HKUST Shenzhen Research Institute, Shenzhen 518057 Google Scholar More articles by this author and Ben Zhong Tang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Department of Chemistry, Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction and Institute for Advanced Study, and Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong HKUST Shenzhen Research Institute, Shenzhen 518057 Center for Aggregation-Induced Emission, SCUT-HKUST Joint Research Institutes, State Key Laboratory of Luminescent Materials and Devices, South China University of Technology, Guangzhou 510640 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.020.202000302 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Living systems, including human beings, animals, and plants, display the power to self-heal spontaneously after being damaged. The self-healing is usually selective, which means that the healing efficiency is related to the spatial distribution of dynamic interfacial interactions of the two rupturing surfaces. Current artificial systems use noncovalent interactions or dynamic covalent bonds to prepare self-healing materials. However, they can only show nonselective self-healing due to their homogeneous internal structures. Herein, we report the construction of a composite hydrogel Gel- C consisting of three different self-healing hydrogels (Gel- Y, Gel- G, and Gel- O) through the use of classic bilayer hydrogel technology. When the composite hydrogel was cut into two pieces, the relative orientation of the parts was rotated through different angles to study the differences in self-healing. Owing to the heterogeneous internal structure of the composite hydrogel and the recognition specificity of each included hydrogel, the interfacial dynamic interactions distribution of the two rupturing surfaces is diverse. The results of tensile tests demonstrated that these rotated samples exhibited different self-healing efficiencies. This system realized the transformation of artificial materials from nonselective self-healing to selective self-healing, providing inspiration for the development of novel biological materials and engineering materials. Download figure Download PowerPoint Introduction Biological systems are usually composite, which means that they are formed by the combination of different components.1,2 These components are assembled in a highly ordered and multidimensional manner, and thereby behave hierarchically.3–5 Each component has a distinct function, but they also work synergetically.6,7 For example, muscle contains many important components, such as muscle fiber, blood vessels, nerves, and so on (Figure 1a).8–10 When a severe accident happens, such as the breaking of a limb, the muscle of the limb does not heal effectively if the two injured parts are mismatched (Figure 1a), namely each component in contact with another but not itself. However, when the two injured parts are perfectly matched (Figure 1a), these components are continuous across the break, and then the whole system could self-heal effectively.11 This naturally selective self-healing phenomenon is a result of the intricate and sophisticated internal structure of tissues: each component has its own recognition specificity, thereby the mismatch will not induce recognition between different components.12 In addition to the human body, selective self-healing phenomena are also widely seen in other biological systems.13 Self-healing is a property that a material can heal the wound by itself when it is broken. Therefore, to better create smart materials that are much closer to the natural systems, one challenge is to construct a material showing selective self-healing. In current artificial self-healing systems, the efficient self-healing is achieved based on noncovalent interactions14–16 or dynamic covalent bonds.17–19 Generally, dynamic covalent bonds are more stable, while noncovalent interactions show rapid dynamic equilibria. The noncovalent interactions include hydrophobic interactions,20,21 hydrogen bonds,22,23 electrostatic interactions,24,25 metal coordination,26 π–π stacking,27,28 donor–acceptor interactions,29,30 host–guest interactions,31–34 and so on. For example, due to hydrophobic interactions, aggregative hydrophobes such as surfactant micelles act as cross-linking points to prepare self-healing materials. Hydrogen bonds, such as the interactions between terminal carboxyl groups and amide groups of aminocaproic acids, or the dimerization of 2-ureido-4-pyrimidone (UPy) units, are also commonly used in the preparation of self-healing materials. Electrostatic interactions, metal coordination, donor–acceptor complexes, or aromatic π–π stacking between polymer chains are other important kinds of noncovalent interactions that can provide driving force for self-healing. In addition, macrocycle-based host–guest interactions are unique because of their combination of multiple noncovalent interactions between the host and guest compounds.35 Among various macrocyclic hosts, cyclodextrin (CD),36–40 crown ethers,41,42 and cucurbit[8]uril (CB[8])43,44 have been widely used to prepare self-healing host–guest materials. On the other hand, imine bonds,45,46 disulfide bonds,47,48 boronate ester complexation,49,50 Diels–Alder reaction,51,52 and reversible radical reactions53,54 have been intensively used as important dynamic covalent bonds for robust self-healing materials. Although these artificial self-healing materials are inspired by natural systems, their self-healing properties are still in the preliminary stage, implying that they can only realize the repair of exterior structure. For example, when these materials are cut into two pieces, their self-healing efficiency will not change even though one broken piece is rotated; namely, displacement of the rupture plane will not affect the interfacial dynamic interactions distribution of the two rupturing surfaces. Interfacial dynamic interactions distribution is the distribution of dynamic covalent bonds or noncovalent bonds which are responsible for the self-healing between the two rupturing surfaces of the materials. Among current artificial self-healing materials, most of them are homogeneous systems. Thus once the materials are broken, dynamic covalent bonds or noncovalent bonds which are responsible for self-healing performance on the two rupturing surfaces distribute evenly. This nonselective self-healing phenomenon reflects the homogeneous internal structure of current artificial self-healing materials. Figure 1 | (a) Cartoon representation of selective self-healing of muscle. Cartoon representations of (b) formation of a composite hydrogel Gel-C and (c) its selective self-healing process. Download figure Download PowerPoint To develop such a material with selective self-healing, the key points are (1) the design of heterogeneous internal structure and (2) how to realize the perfect match of two injured parts. As to the heterogeneous internal structure, we can learn from natural systems to construct a composite hydrogel, which contains different self-healing hydrogel components inside. On the other hand, luminescent materials are good candidates to help visualize the perfect match of different components.55 However, luminescence is often weakened or quenched at high concentration, which has frequently been referred to as “aggregation-caused quenching (ACQ).” To solve the problem, we found that some materials were nonemissive in dilute solution but became highly luminescent when their molecules were aggregated in concentrated solutions or cast into solid films, which we termed “aggregation-induced emission (AIE).”56–59 Therefore, the introduction of AIE materials would help visualize the perfect match of different components.60–65 Thus, herein a heterogeneous composite hydrogel containing internal structures, each with its own different self-healing and luminescent properties, was constructed taking advantage of bilayer hydrogel technology and the AIE phenomenon. Three hydrogel parts with different self-healing mechanisms were integrated in the composite hydrogel. The different parts also showed different luminescent colors due to introductions of different AIE luminogens (AIEgens). The composite hydrogel had selective self-healing properties owing to the heterogeneous internal structure and the recognition specificity of each included hydrogel part. The selective self-healing property of the composite hydrogel was shown through a series of rotation-mode healing experiment guided by AIE luminescence. This system realized the transformation of artificial materials from nonselective self-healing to selective self-healing, providing inspiration for the development of novel biological and engineering materials. Experimental Methods AIE-Y,66 AIE-G,67 AIE-O,68 acylhydrazine-terminated poly(ethylene oxide) (PEO) (AH-PEO-AH),55 and tetraaldehyde-terminated PEO (TAPEO)55 were synthesized according to literature methods. Other starting materials and reagents were purchased from commercial suppliers and used without further purification unless otherwise stated. All solvents were purified before use. The solvent was carefully dried and distilled from a suitable desiccant before use. 1H NMR spectra were measured on a Bruker AVIII 400 MHz NMR spectrometer. Fluorescent emission spectra were measured on a HORIBA Fluorolog-3 research spectrofluorometer equipped with a solid sample holder. Suspension samples were measured using a quartz cuvette. Hydrogel samples were measured with the solid sample holder. Tensile tests were performed using a strain rate of 0.6 mm s−1 at room temperature and recorded on an Instron Tensile Tester with a 10 N transducer. At least five replicate samples were recorded for each sample in the test. Detailed procedures about preparation of the composite hydrogel and tests of its selective self-healing property are described in Supporting Information. Results and Discussion Formation of the composite hydrogel As shown in Figure 1b, three self-healing hydrogels69–72 (Gel- Y, Gel- G, and Gel- O) with different dynamic interactions and AIE dots66–68 are initially prepared and then fixed in a cylindrical mold (10 mL syringe). Subsequently, an aqueous solution containing monomer N,N-dimethylacrylamide (DMA), covalent cross-linker poly(ethylene glycol) diacrylate (PEGDA), and initiator potassium persulfate (KPS) is added into the mold. After the addition of accelerator N,N,N′,N′-tetramethylethylenediamine (TEMED), the aqueous solution gradually becomes a hydrogel. During this period, some of the solution will diffuse into the three hydrogels, inducing the formation of covalent hydrogels in each of the three fluorescent hydrogels. As a result, a composite hydrogel (Gel- C) with three different self-healing hydrogels inside is generated. When it is cut into two parts and one of them is rotated axially, the two parts then contact again for self-healing (Figure 1c). Because of the heterogeneous internal structure of the composite hydrogel Gel- C and the different components of each included hydrogel, the interfacial dynamic interactions distribution of the two rupturing surfaces is varying; therefore, these self-healing effectiveness are expected to be different (Figure 1c). The structures of three self-healing hydrogels are illustrated in Figure 2. As shown in Figure 2a, hydrogel Gel- Y is produced by mixing AH-PEO-AH and TAPEO in water. Due to the formation of dynamic acylhydrazone bonds between the two polymers, a dynamic covalent hydrogel is formed. Hydrogel Gel- G (Figure 2b) contains poly(N-isopropyl acrylamide) (PNIPA)/inorganic clay network structure. Its self-healing property is based on the ionic interactions between the polymer and clay surface. Hydrogel Gel- O (Figure 2c) is a poly(vinyl alcohol) (PVA) hydrogel whose self-healing ability comes from the hydrogen bonding between PVA chains. The incorporated AIEgens (AIE-Y, AIE-G, and AIE-O) endow the three hydrogels with yellow, green, and orange fluorescent color, respectively. The introduction of various fluorescent colors is intended to aid the visualization of the subsequent selective self-healing process. In addition, hydrogel Gel- M (Figure 2d) is designed as a covalently cross-linked hydrogel to form the stable outside matrix. The preparation of the fluorescent hydrogels is shown in Supporting Information Figures S1–S7. Figure 2 | Cartoon representations and chemical structures of hydrogels: (a) Gel-Y, (b) Gel-G, (c) Gel-O, and (d) Gel-M. (e) Cartoon representation and fluorescence images (λex = 365 nm) from different views of hydrogel Gel-C. Download figure Download PowerPoint To construct the composite hydrogel Gel- C, the connection strength between each component should be examined. To form a stable composite hydrogel, a bilayer hydrogel technique was applied in this system. Hydrogel Gel- M was used to form bilayer hydrogels with hydrogels Gel- Y, Gel- G, and Gel- O. For example, an aqueous solution containing the monomer, covalent cross-linker, and initiator for hydrogel Gel- M was poured into a mold already half filled by hydrogel Gel- Y. After waiting for 2 h, the solution diffused into the interior of hydrogel Gel- Y. Accelerator was then added to induce the formation of the covalent hydrogel Gel- M both inside and outside hydrogel Gel- Y. In this way, a bilayer hydrogel YM composed of hydrogels Gel- Y and Gel- M was formed. As shown in Supporting Information Figure S8, the fluorescence image of YM exhibits yellow and colorless, in agreement with the incorporated two hydrogels. To examine the connection strength between hydrogels Gel- Y and Gel- M, tensile test was used to demonstrate that bilayer hydrogel YM could be stretched obviously (see Supporting Information Figure S8a and Supporting Information Movie S2). This indicates that two layers (Gel- Y layer and Gel- M layer) in bilayer hydrogel YM connect with each other tightly so that YM can be regarded as an integrated hydrogel. Similarly, bilayer hydrogels GM and OM were also prepared (see Supporting Information Figures S8b and S8c and Supporting Information Movie S2). The tensile stress–strain curves of these three bilayer hydrogels were obtained from the tensile tests (see Supporting Information Figures S8d–S8f and Supporting Information Movie S2), confirming the stable connection between Gel- M and Gel- Y/Gel- G/Gel- O. After confirming the connection strength between each component hydrogel, we next construct the final composite structure (see Supporting Information Figure S9). Cylindrical hydrogel blocks of Gel- Y, Gel- G, and Gel- O were prepared with the same length. They were then inserted in a plastic syringe and fixed with a rubber plug. Pregel solution of Gel- M without TEMED was prepared and injected in the syringe. After standing at room temperature for 2 h, TEMED was injected in the syringe and the sample was held at room temperature for 16 h to form the composite hydrogel Gel- C. The fluorescsence images (Figure 2e) and movies (see Supporting Information Movie S3) from different views of Gel- C clearly demonstrated its composite structure. Since a cylindrical sample is not suitable for tensile testing, the corresponding flat hydrogel sample was used instead to demonstrate the tensile behavior. The flaky composite hydrogel sample was then cut into two pieces for the self-healing test. After putting the two broken parts back to their original position for a range of times, their tensile behaviors were examined (see Supporting Information Figures S11–S13 and Supporting Information Movie S4). The results illustrated that the fracture strain increased with the extension of healing time, indicating the time-dependent self-healing behavior. Self-healing of pair-wise hydrogels Self-healing experiments were also investigated among the component hydrogels Gel- Y, Gel- G, Gel- O, and Gel- M. Each hydrogel was contacted with itself and the other three ones, separately. After waiting for 50 h, these samples were lifted up. The healed samples Y– Y, G– G, and O– O are shown in Figures 3a–3d and Supporting Information Movie S1, which reflected that all of hydrogels Gel- Y, Gel- G, and Gel- O can display good self-healing ability. However, when they heal with other ones, the healing efficiency decreased (Figure 3d) or even disappeared (see Supporting Information Figure S14 and Supporting Information Movie S1). To further quantify the self-healing effects, tensile tests were conducted. Based on the tensile stress–strain curves (Figure 3e), it can be seen that the moduli of O– O and G– G are respectively the largest and smallest ones, indicating that hydrogel Gel- O is more rigid while hydrogel Gel- G is much more stiff. The corresponding breaking stress (Figure 3f-ii) and break strain values (Figure 3f-ii) can be summarized as follows. For hydrogel Gel- Y, the mechanical parameters were only observed when contacted with itself (Figure 3f). As to the highly compliant hydrogel Gel- G, the G– G sample exhibited the greatest breaking strain values among all involved samples (Figure 3f-ii). In terms of the more rigid hydrogel Gel- O, the O– O sample showed greater breaking stress values than those of G– O (Figure 3f-i). The results indicated that each of the three fluorescent hydrogels showed the best healing ability when contacting with itself. Figure 3 | Cartoon representations, fluorescence images (λex = 365 nm), and tensile tests of pair-wise hydrogels (a) Y–Y, (b) G–G, (c) O–O, and (d) G–O. (e) Tensile stress–strain curves of the pair-wise hydrogels. (f) Summary of breaking stress values (i) and breaking strain values (ii) of the pair-wise hydrogels after contacted for 50 h. Download figure Download PowerPoint Qualitative characterization of selective self-healing To study whether the composite hydrogel Gel- C exhibits the ability of selective self-healing, seven cylindrical hydrogels Gel- C were all cut into two parts. In this test, all the separated lower parts were fixed, while the upper parts were rotated through different angles: 0°, 60°, 120°, 180°, 240°, 300°, and 360° (Figure 4). The lower part and the corresponding rotated upper part were then put back into contact for healing. The detailed fluorescence images from different views of these samples are shown in Figure 4, Supporting Information Figures S15–S21, and Supporting Information Movie S5. After waiting for 50 h, these seven healed cylindrical hydrogels were lifted up (see Supporting Information Movie S6). As shown in Figure 4, no fracture was observed for the samples whose upper parts have been rotated through 0°, 60°, 300°, and 360° (Figures 4a, 4b, 4f, and 4g). However, only the upper parts could be lifted up as seen from the samples with rotation angles of 120°, 180°, and 240° (Figures 4c, 4d, and 4e). These results implied that the self-healing effectiveness is related to interfacial dynamic interactions distribution of the two rupturing surfaces. Moreover, with the help of AIEgens, the selective self-healing process could be visualized. Figure 4 | Cartoon representations, fluorescence images (λex = 365 nm) from different views, and lift up processes of the self-healed cylindrical hydrogel Gel-C when one broken part was rotated throughdifferent angles: (a) 0°, (b) 60°, (c) 120°, (d) 180°, (e) 240°, (f) 300°, and (g) 360°. Note: the healing time was 50 h. Download figure Download PowerPoint Quantification of selective self-healing To further quantify the difference in self-healing efficiency, seven self-healed flaky hydrogels corresponding to the aforementioned self-healed cylindrical hydrogels were prepared (see Supporting Information Figure S22). Tensile tests were carried out to compare their self-healing abilities (see Supporting Information Movie S7). As shown in Figures 5a–5g, all these healed flaky hydrogels could be stretched but with different breaking strain values. According to their tensile stress–strain curves (see Supporting Information Figure S23), the breaking strain values can be summarized as follows. As can be seen in Figure 5h, the samples corresponding to the rotation angles of 0° and 360° show the greatest breaking strain values. The data representing 60° and 300° decreased obviously, and those for 120°, 180°, and 240° displayed much lower values. The heterogeneous internal structure of the composite hydrogel and the recognition specificity of each included component account for the aforementioned selective self-healing phenomenon. As proved in Figure 3, each of the incorporated fluorescent hydrogels exhibits the best healing property with itself, while the healing efficiency decreases when contacting with others. Therefore, when the upper part of the broken composite hydrogel was rotated through different angles, the interfacial dynamic interaction distributions of the two rupturing surfaces changed. In the case of 0° and 360°, all included hydrogels (Gel- Y, Gel- G, and Gel- O) contacted with themselves completely, thus these two samples showed the greatest breaking strain values. However, for the systems at 60° and 300°, each fluorescent hydrogel contacted partially with itself, leading to the decrease of self-healing efficiency. As to the cases of 120°, 180°, and 240°, all the internal components contacted completely or partially with others but had no contact with themselves, producing much lower breaking strain values. Tensile property of original Gel- C was also tested ( Supporting Information Figure S10 and Supporting Information Movie S8). From a comparison between the original sample and a healed sample (0°, Figure 5h), it can be seen that the final breaking strain of the original sample was about 240%, while that of the healed sample (0°) was only about 89%. Figure 5 | Cartoon representations, fluorescence images (λex = 365 nm), and tensile tests of the self-healed flaky hydrogels whose upper part has been slided with different angles: (a) 0°, (b) 60°, (c) 120°, (d) 180°, (e) 240°, (f) 300°, and (g) 360°. Note: the healing time was 50 h. (h) Summary of the final breaking strain values of the aforementioned self-healed flaky hydrogels. Download figure Download PowerPoint Conclusion In summary, we report the selective self-healing behavior of a composite hydrogel Gel- C. This hydrogel was constructed by using the typical bilayer method among three fluorescent self-healing hydrogels and a covalently cross-linked hydrogel. Tensile tests were carried out to confirm the stable connection among each component hydrogel. The self-healing driven forces of three included fluorescent hydrogels were based on dynamic covalent bonds, ionic interactions, and hydrogen bonding. Due to the recognition specificity of each involved interaction, the three hydrogels show the best healing efficiency with themselves but lower strength when contacted with others. These properties were then applied in the self-healing of the composite hydrogel Gel- C. When Gel- C was cut into two pieces, its upper part was rotated through different angles (0°, 60°, 120°, 180°, 240°, 300°, and 360°) to study the difference in self-healing. The results proved that these rotated samples displayed different self-healing abilities. The systems with rotation angles of 0° and 360° showed the biggest breaking strain values. The values corresponding to 60° and 300° decreased obviously, and those for 120°, 180°, and 240° showed much lower values. Because of the heterogeneous internal structure of the Gel- C and the recognition specificity of each included hydrogel, the interfacial dynamic interactions distribution in space of the two rupturing surfaces is diverse, which ac