Hierarchical hybrid crosslinking multifunctional gelatin-based hydrogel: ideal platforms for flexible wearable devices, brain–computer interfaces and biomedical applications

Abstract

Hydrogels are promising candidates for flexible wearable technology, biomedicine, and even brain–computer interfaces (BCI). However, the mismatch in the mechanical properties and high biotoxicity of the materials cast a shadow on their application prospects. Herein, we developed a multifunctional hydrogel matrix primarily based on the natural polymer gelatin. Multilevel hybrid dynamic crosslinking (MHC) enables the adjustment of the hydrogel molecular network and endows the material with satisfactory mechanical properties and self-healing behavior. The excellent biocompatibility of the hydrogel is enough to support the growth and proliferation of cells while avoiding tissue rejection or inflammation. Moreover, the excellent self-adhesive performance allows the hydrogel to directly adhere to the surface of human skin and tissues, enabling real-time monitoring of body movements and non-invasive sensing of electroencephalogram (EEG) signals. Therefore, the multifunctional hydrogel matrix with self-adhesive behavior, self-healing properties, appropriate mechanical performance, and excellent biocompatibility can be regarded as a promising platform for applications in flexible wearable devices, biomedical materials, and BCI devices.

---

1 Introduction

Hydrogels, with their unique 3-dimensional (3D) network structure and easily adjustable mechanical and rheological behaviors, show bright prospects in the fields of flexible wearable devices, biomedical membranes and BCIs.^1–4 Resourceful and renewable natural polymers with excellent biocompatibility and a large number of reactive groups (hydroxyl and carboxyl groups) are considered one of the most promising classes of candidate materials for fabricating hydrogels.^5–7 However, the shortcomings of single-component natural polymer-based hydrogels, such as poor mechanical properties, easy breakage,⁸ and lack of self-healing ability,⁹ make it difficult to cope with repetitive and frequent external stimuli or complex deformation in practical applications.^10–12 Furthermore, these hydrogels commonly lack bio-adhesive properties, which hinders close and stable contact between the hydrogel and biological tissues and thus affects the stability and detection sensitivity of sensing and BCI testing.^13–15 Therefore, the construction of natural polymer-based hydrogels with good mechanical, self-healing, and self-adhesive properties is key to promote the future development and application of hydrogels in various fields, including biomedicine, flexible sensing and BCIs.

The single covalent crosslinking network and lack of functional groups are the main reasons for these defects. The single covalent crosslinking structure inhibits the dissipation of energy, causing pressure to rapidly exceed the tolerance threshold of natural polymer hydrogels, resulting in material breakage.^16,17 The existence of multiple crosslinking forms (hydrogen bonding, dynamic chemical bonding, and covalent bonding) allows for a gradual dissipation of energy, increasing the hydrogel's resistance to external stress stimuli and preventing irreversible permanent damage to the material.¹⁸ For example, Guan et al.¹⁹ introduced ionic crosslinking and hydrogen bonding into a covalent crosslinking hydrogel network, realizing a significant improvement in the mechanical properties of the gel, with an increase in elongation at break and toughness by 240% and 160%, respectively. Similarly, Sofie Houben's team and Bo Yang's team combined hydrogen bonding and dynamic cross-linking to achieve efficient tuning of gel mechanical properties.^20,21

The absence of functional groups weakens the interaction between hydrogel molecular chains or between molecular chains and other materials and prevents the smooth occurrence of self-healing or self-adhesion behavior.²² For example, dopamine (DA),^23,24 tannic acid (TA)²⁵ or modified lignin-containing catechol groups can interact with various polar substrates to form strong non-covalent interactions, endowing hydrogels with unique self-adhesive properties. Typically, Bai et al.²⁶ reported a self-adhesive hydrogel consisting of polyvinyl alcohol (PVA), phytic acid, and polydopamine (PDA), in which the catechol functional group of PDA can form noncovalent interactions with materials such as skin, glass, and plastics, and provide self-adhesive ability. Zhang et al., on the other hand, introduced o-phenol hydroxyl groups into the material system by relying on the direct cross-linking of gelatin by tannic acid and achieved the smooth preparation of hydrogels with both good mechanical properties and desirable self-adhesive behavior.²⁷ The self-healing process of most hydrogels relies on reversible dynamic chemical bond reactions between specific chemical groups, such as the aldol condensation between aldehyde groups and amino groups,^28,29 and the reversible amidation between maleic anhydride and secondary amines,³⁰ without the need for an additional healing component (capsule).³¹ When these chemical bonds are broken by external stimuli, the forces acting on the fresh broken site lead to intermolecular self-assembly, which in turn drives the rearrangement of the fractured gel network and the self-healing of the broken sample.^32,33 Therefore, it can be speculated that constructing a multilevel interpenetrating dynamic crosslinking network and introducing appropriate functional groups may be an effective way to endow natural polymer hydrogels with self-healing properties, self-adhesive behavior, and good mechanical performances at the same time.

Natural polymer gelatin derived from animal skin, bone and other tissues can easily undergo intermolecular helical entanglement to produce gelation,³⁴ is an ideal raw material for hydrogel construction. Here, multilevel hybrid dynamic crosslinking (MHC) hydrogel networks were constructed using gelatin as the main gel matrix. The introduction of PVA, borax and tannic acid into the gelatin macromolecular network brings new intermolecular interactions such as dynamic covalent bonding (borate bonding) and multiple hydrogen bonding to hydrogels (Fig. 1). They hybridize with the existing molecular helix entanglement interactions, which together achieve the construction of MHC networks and endow the hydrogels with excellent comprehensive performance. Different from the previously reported research papers, our study focuses on exploring the effect of hierarchical cross-linking or multilevel cross-linking on the gel mechanical properties, realizing the synergistic unification of dual-network structure with multilevel interactions. This unique construction strategy effectively improves the flexibility of hydrogels and brings unexpected multifunctionality to them. The MHC network improves the hydrogel's resistance to external stimuli, and the hierarchical release of multilevel energy provides a dramatic increase in the elongation at break and toughness of more than 3500% and 4000%, respectively. Besides, the large number of dynamic chemical bonds offers the hydrogel a satisfactory self-healing rate and efficiency. The o-phenolic hydroxyl groups in tannic acid enhance the interaction between the gel and human tissues, endowing the hydrogel with certain self-adhesive behavior. The hydrogel adhered to the human skin has the capability to capture various human movements (joint motion, coughing, etc.) and can detect changes in EEG signals and serve as an electrode, demonstrating its potential use in the field of flexible wearable devices and BCI. Moreover, due to the excellent biocompatibility of gelatin and PVA, the MHC hydrogels do not cause massive cell death and obvious rejection of biological tissues, creating the premise for playing the role of tissue engineering carrier. In summary, MHC hydrogel is worthy of being used as a hoping medium for flexible wearable devices, biomedicine and BCI, and the construction of an MHC network is also an effective strategy to develop high-performance and functional hydrogels (Fig. 1).

Fig. 1 Schematic of the fabrication processing of MHC hydrogels.

2 Experimental

2.1 Materials

Gelatin and borax (99%) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd (Shanghai, China). PVA (1750 ± 50) was obtained from Sinopharm Chemical Reagent Co., Ltd (Shanghai, China). Tannic acid was purchased from Macklin reagent Co., Ltd (Shanghai, China). All chemicals were used directly without further purification.

2.2 Preparation of pure gelatin (G) and tannic acid-gelatin (TG) hydrogel

4.0 g of gelatin was added to 27 mL of deionized water and stirred at 70 °C for 1 h to obtain a clear gelatin solution. Subsequently, the gelatin solution was cooled at 4 °C for 1 h, and the G hydrogel was obtained. For TG hydrogel, 0.15 g tannic acid and 4 g gelatin were dissolved in 5 mL and 22 mL deionized water, respectively, to obtain aqueous solutions of tannic acid and gelatin. Next, the two solutions were mixed and fully stirred, and then cooled at 4 °C for 1 h to obtain TG hydrogel.

2.3 Preparation of the gelatin-tannic-acid-borax (GTB), gelatin-tannic-acid-PVA (GTP), and PVA-tannic-acid-borax (PTB) hydrogels

4 g gelatin was added to 17 mL of deionized water and stirred at 70 °C for 1 h to obtain solution A. 0.15 g tannic acid and 0.25 g borax were dissolved in 10 mL of deionized water to obtain solution B. Solution A was then slowly poured into solution B under continuous stirring. After stirring for 2 min, the mixture was cooled at 4 °C for 1 h to obtain GTB hydrogel.

4.0 g of gelatin and 1.5 g of PVA were dissolved in 22 mL of deionized water at 95 °C to obtain solution C. Then, 0.15 g of tannic acid was dissolved in 5 mL of deionized water to obtain solution D. The two solutions were then heated up to 95 °C, and then quickly mixed. After stirring for 2 min, the mixture was poured into the mould and cooled to obtain the GTP hydrogel.

1.5 g of PVA was dissolved in 17 mL of 95 °C deionized water to obtain PVA solution E. Then, 0.15 g tannin and 0.25 g borax were added to 10 mL of deionized water and dissolved under thorough stirring to obtain solution F. Solutions E and F were heated to 95 °C and mixed under vigorous stirring. Finally, the mixture was poured into a mould and cooled to yield the PTB hydrogel.

2.4 Preparation of the MHC gelatin (MHC-PG) hydrogel

4.0 g gelatin and 1.5 g PVA were added to 17 mL deionized water, and the mixture was continuously stirred and dissolved at 95 °C for 2 h to obtain PVA/gelatin solution. Then, 0.15 g tannic acid and 0.25 g borax were dissolved in 5 mL deionized water separately and the two solutions were heated to 95 °C. Then, the tannic acid solution was added to the PVA/gelatin solution at 95 °C with rapid stirring. After 5 min, the aqueous borax solution was added to the above mixture and poured into the mold after sufficient stirring for 2 min to obtain the MHC hydrogel.

2.5 Characterization

The 3D network structure of hydrogels was observed and recorded using a scanning electron microscope (SEM, SU5000, Hitachi). The chemical cross-linking structures of different samples were analyzed using Fourier transform infrared spectroscopy (FTIR, iS50, Thermo Fisher Scientific, USA). The mechanical properties of the materials were recorded using an electronic universal testing machine (INSTRON2344, USA). Rectangular hydrogel samples were gradually stretched at a constant rate of 50 mm min⁻¹, while samples for compression testing were cylindrical with a diameter of 20 mm and tested at a speed of 10 mm min⁻¹. The rheological behavior of hydrogels at room or variable temperatures was investigated using a rheometer (AR2000ex, TA, USA). The crystalline properties were investigated using an X-ray diffractometer (XRD, D8 advance, Bruker, Germany) with operating voltages and currents of 40 kV and 40 mA, respectively. The hydrogel resistance was recorded using a four-probe tester (RST-8, Guangzhou Four-Probe Technology, China). The strain-sensing behavior of hydrogels was demonstrated using a high-precision source meter (Agilent B2902A, Agilent, USA) to record their real-time resistance changes. EEG signals were obtained from an internationally recognized 32-channel EEG acquisition system with MHC hydrogel and commercial wet electrodes as experimental and control groups, respectively. The signals were first amplified by an eegoTM mylab amplifier (ANT Neuro, Enschede, Netherlands) with a sampling frequency of 1000 Hz and then collected for analysis. The swelling behavior of the hydrogels was analyzed by recording the mass changes. The hydrogels were soaked in sufficient deionized water for a period of time. Subsequently, the hydrogels were removed from the deionized water and their masses were recorded. The swelling ratio was calculated by the following equation:³⁵where M_t and M₀ are the masses of the initial and swollen samples, respectively.

2.6 Cytocompatibility assessment

The cell proliferation was observed by utilizing CCK-8 assay. Briefly, 1 × 10⁴ mL⁻¹ of human umbilical vein endothelial cells (HUVECs) was inoculated in 96-well plates. After 24 h, 48 h and 72 h of culture, the proliferation of cells was evaluated by Cell Counting Kit-8 (CCK-8, Solarbio). At a fixed time, 10 μL CCK-8 solution and 100 μL fresh culture medium extracting solution were added to the cells washed by PBS in turn, and incubated at 37 °C in the dark for 1 h. The proliferation ability was evaluated by measuring the absorbance at 450 nm with the enzyme-labeled instrument. Live/dead cells were stained with a Calcein AM/PI double staining kit (KeyGen, China), and the cells were observed and photographed with a confocal laser scanning microscope (CLSM, Olympus, Japan).

2.7 Blood compatibility assessment

Firstly, diluted goat blood was obtained by mixing 10 mL PBS buffer solution and 8 mL fresh goat whole blood. Subsequently, an ultraviolet sterilized hydrogel sample (0.5 g) was immersed in a centrifuge tube containing 10 mL PBS and 0.2 mL diluted goat blood and then incubated at 37 °C for 1 h. Diluted goat blood was incubated for 60 min at 37 °C using PBS solution and deionized water as negative and positive controls, respectively. Finally, all tubes were centrifuged at 3000 rpm for 5 min, the color of the supernatant was observed and the optical density (OD) at 540 nm was analyzed and recorded. The hemolysis ratio was calculated using the following equation:³⁶where OD_Exp, OD_Ne and OD_Po are the OD values of the experimental group, negative control group and positive control group, respectively.

2.8 Experimental animals and treatments

All animal studies were approved by the Animal Care and Use Committee of the Central Hospital of Dalian University of Technology (YN2024-022-56). Black mice were housed in ventilation cages with a constant temperature of 25 °C, 40–70% humidity, 12 h/12 h light/dark cycle, and regular water and food were provided. Animals were randomly subdivided into the control group and the MHC-PG hydrogel-treated group. To establish an experimental model, mice were anesthetized and MHC-PG hydrogel was implanted into the liver tissue of mice. After the operation, the weekly food intake and weight changes of mice were recorded. After the mice were killed, the liver tissue around the hydrogel implant was fixed with 4% paraformaldehyde solution, and the histocompatibility of MHC-PG hydrogel was analyzed by hematoxylin and eosin (H&E) staining.

3 Results and discussion

The main backbone of MHC-PG hydrogel is composed of gelatin and PVA, and the MHC together supports the excellent comprehensive performance of the hydrogel. As shown in Fig. 2a, gelatin and PVA are easily dissolved in deionized water and maintain the desired fluidity. After mixing the borax and tannic acid solutions, the cross-linking reaction occurs rapidly and drives the gelation process within 5 min. The mechanical properties of the hydrogels, such as breaking strength and elongation at break, are closely related to the type of crosslinking and the density of the crosslinking network. MHC can significantly improve the interpenetrating network structure and energy dissipation mechanism inside the hydrogel, and then improve the mechanical properties of hydrogel. The stress–strain curves of TG and GTB hydrogels showed no obvious difference, indicating that the introduction of borax does not significantly hinder the formation of gelatin hydrogels (Fig. 2b and c). The comparison results of PTB and MHC-PG curves further confirm ideal enhancement of hydrogel mechanical properties by the MHC network. The compression test results match the tensile curves well with a similar trend, and the MHC-PG hydrogel still had the best compression data (Fig. 2d). Notably, compared to the single crosslinked pure gelatin hydrogel (G hydrogel), the MHC-PG hydrogel exhibits much higher toughness and larger elongation at break, which are enhanced by 3500% and 4000%, respectively (Fig. 2b, c and S1†). Compared to similar studies, our hydrogels have higher flexibility, with elongation at break exceeding about 50% compared to that of these papers, which is attributed to the continuous and homogeneous dissipation of energy by the gradient crosslinked network.³⁷ From the above analyses, it can be seen that tannins and borax mainly have strong intermolecular interactions with gelatin and PVA, respectively. The performance analysis of G, TG and MHC-PG hydrogels can well reflect the difference between “single cross-linking” and “hierarchical cross-linking”. Therefore, we chose two hydrogels, G and TG, as control samples to analyze the dramatic changes in the hydrogel structure and properties brought about by hierarchical cross-linking in the course of the subsequent study.

Fig. 2 (a) Digital photos and schematic of the cross-linked network of MHC-PG hydrogels. (b) Stress–strain curves, (c) elongation at break and breaking strength of the hydrogels. (d) Compression curves of different hydrogels. (e) Optical microscopy images of the hydrogel self-healing process and healing mechanism. (f) SEM images of freeze-dried G, TG, and MHC-PG hydrogels. (g) Swelling behavior of the hydrogel. (h and i) Rheological behavior of the hydrogel at room temperature and under variable temperature conditions.

Relying on the dynamic properties of borate bonding and the phenolic hydroxyl of tannic acid, a fast and efficient gel transformation process was realized and it contributed to the smooth formation of hydrogels. The dynamic bonding network and the interpenetrating macromolecular structure ensure the mechanical stability and self-healing properties of the hydrogels. FTIR curves showed that all samples exhibited characteristic absorption bands attributed to –OH at 3100–3700 cm⁻¹ (Fig. S2†). Furthermore, in the curves of G and TG hydrogels, the absorption peaks in the regions of 1632, 1525, and 1233 cm⁻¹ are attributed to stretching vibrations of the C=O bond caused by amide I, the overlap between N–H bending vibrations and C–N stretching vibrations related to amide II, the vibrations of the C–N and N–H groups of the amide range related to amide III, respectively.³⁸ Notably, the addition of tannic acid causes a slight enhancement of the –OH absorption peak in the hydrogel, indicating the formation of hydrogen bonds between tannic acid and gelatin (Fig. S2†).

Whereas, the vibrational absorption peaks at 1405 cm⁻¹ and 840 cm⁻¹ belonging to the B–O–C and the B–O (in residual B(OH)₄⁻) bond, respectively, confirm the existence of borate bonds between PVA and borax (Fig. S2†).³⁹ These analyses demonstrate the successful construction of MHC network structures in hydrogels. Besides, MHC-PG can withstand more than 90% compression deformation and still does not crack (Fig. 2d and S3†). In contrast, all G and TG hydrogels begin to crack under about 70% deformation, and it is difficult to maintain macro-integrity.

In addition, due to the dynamic properties of borate bonds and multiple hydrogen bonds, damaged MHC-PG hydrogels can achieve a self-healing process at room temperature, facilitating the repair and regeneration of a fractured gel network caused by external stimuli. Both optical micrographs and digital photos show that the scar caused by scissors almost disappeared after 10 min of re-contact with fresh wounds, visually confirming the existence of dynamic crosslinking networks (Fig. 2e and S4†). In addition, the rapid recovery of conductivity further signals the completion of the hydrogel healing process (Fig. S5†). After 10 min of healing, the stress–strain curve of the MHC-PG hydrogel almost overlaps with that of the original state and demonstrates an elongation at a break of more than 800% (Fig. S6†). The structure changes of the internal 3D network can be directly confirmed by SEM. From the SEM images (Fig. 2f), it can be observed that a well-defined hierarchical pore network fills the interior of the freeze-dried MHC-PG hydrogel. However, uniform 3D pore structures are distributed in the G and TG hydrogels, suggesting a consistent yet single cross-linked network configuration (Fig. 2f). In addition, the hierarchical network formed after MHC hinders the movement of water molecules to some extent and reduces the swelling rate of MHC-PG hydrogel (Fig. S7†). The swelling rate of the dried G hydrogel in deionized water is close to 750% after 5 h, and it can reach the swelling equilibrium state soon. However, the swelling rate of the gelatin hydrogel containing TA exhibits a slight decrease, but it can still absorb water quickly. In contrast, it needs nearly 15 h for the swelling degree of MHC-PG hydrogel to reach 700%, proving increases in the crosslinking density and changes in the crosslinking network structure (Fig. 2g). The storage modulus (G′) of all the samples is higher than their loss modulus (G′′) and the tan δ value is less than 1, which is a typical characteristic of the hydrogel rheological behavior and means that all hydrogels can maintain a good elastic state within the applied frequency range (Fig. 2h). However, the introduction of the MHC network induced a decrease in G′ and tan δ, indicating the decrease of hydrogels' rigidity. It is speculated that the MHC network destroyed the rigid spiral entanglement of some gelatin molecules and enhanced the dynamic toughness crosslinking, which is consistent with the previous mechanical test results (Fig. 2h). In addition, with the increasing temperature, the G′ of all hydrogels decrease obviously because the physical cross-linking network was destroyed (molecular entanglement, hydrogen bonding, etc.) in hydrogels. However, the G′ of MHC-PG hydrogel is always higher than G′′, confirming the good temperature stability of the MHC network. In contrast, the G′ of both TG and G hydrogel is finally lower than G′′, indicating that the physical cross-linking network in the hydrogel has gradually dissociated and the material has changed from the gel state to flowable liquid (Fig. 2i). As shown in Fig. S8,† all different hydrogels show a broad diffraction peak at about 2θ = 20°, which is a typical XRD curve characteristics of most polymers, indicating the highly amorphous structure of the whole gel. However, the peak intensity of the MHC-PG hydrogel is lower than those of G and TG hydrogels, suggesting a more chaotic structure and lower crystallization behavior of the MHC-PG hydrogel.

As a drug-loading platform in some clinical applications, excellent histocompatibility and low cytotoxicity should be representative characteristics. In order to evaluate the biocompatibility of MHC-PG hydrogel, we analyzed in detail the effects of the hydrogel on the survival and proliferation of human umbilical vein endothelial cells (HUVECs) and the rejection caused by its implantation in mice (Fig. 3a). The CCK-8 assay show that the cell viability cultured with TG culture medium is the lowest (about 75%) (Fig. 3b), which can be attributed to the inhibitory effect of tannic acid on cell proliferation and growth. In contrast, the cell activity value of the MHC-PG hydrogel culture medium is close to 90%, because the MHC network can slow down the release rate of tannic acid and create a more favorable environment for cell growth. In addition, after 48 h of culture, almost no dead cells were detected in the hydrogel leaching solution, which is similar to that of the PBS control group (Fig. 3c). In addition, HUVECs cultured with hydrogel leaching solution demonstrated complete cell morphology and obvious actin filaments, which were almost the same as from the PBS control group (Fig. 3d). The above analysis confirms that the MHC-PG hydrogel has no obvious adverse effect on cell proliferation and growth, that is, extremely low cytotoxicity. In order to further evaluate the toxicity of the MHC-PG hydrogel to the living body, the changes in organs and tissues and the living state of mice after hydrogel implantation were analyzed. The H&E staining results show that compared with the blank control group, the implanted MHC-PG hydrogel does not cause obvious damage to the surrounding tissues and cells, and the liver tissue cells have full morphology and clear visible nuclear region (Fig. 3e). In addition, for a period of time, after MHC-PG implantation, the food intake and body weight of mice did not change significantly (Fig. S9†), indicating that the MHC-PG hydrogel did not affect the living status of mice. Given the above results, MHC-PG hydrogel exhibits ideal biocompatibility, suggesting its potential as a biological tissue engineering material or drug-loading platform. Some biomedical materials usually cause the rupture of red blood cells, which in turn leads to internal bleeding. Therefore, we co-cultured the materials with blood to evaluate the hemolytic behavior of the hydrogel. From Fig. S10,† it can be found that almost all the blood cells in the deionized water control group are damaged, and almost no precipitate was seen in the centrifuge tube. In contrast, the supernatant of the experimental group (MHC-PG hydrogel) was almost colorless and transparent, and obvious precipitates gathered at the bottom of the centrifuge tube, meaning that almost no broken erythrocyte, therefore, meeting the use standard of biomedical materials. Moreover, the hemolysis rates of G, TG and MHC-PG hydrogels were all below 10%, supporting their good blood compatibility (Fig. 3f). More attractively, the MHC-PG hydrogel can firmly adhere to the dermal tissue surface of mice without any other adhesives (Fig. S11†). In addition, the MHC-PG hydrogel can also be directly attached to the surface of organs including the liver, muscle and heart (Fig. 3g), proving the good self-adhesive performance of the hydrogel and can be attributed to the intermolecular interactions (hydrogen bond, etc.) between the ortho-phenolic hydroxyl group of tannic acid in the hydrogel and the surface of biological tissue. Quantitative tests prove that the adhesion strength of MHC-PG hydrogel to a wide range of materials, including paper, pigskin, wood and metal, exceeds 3 kPa, ensuring stable anchoring of the hydrogel to complex surfaces (Fig. S12†). Besides, after 72 h of immersion in simulated body fluids, MHC-PG undergoes no dissolution and still has good mechanical properties, which implies its great potential as a drug-loading platform (Fig. S13†).

Fig. 3 (a) Evaluation of cell compatibility and biocompatibility of the hydrogel. (b) Viability of HUVECs seeded on the MHC-PG hydrogels on days 1, 2, and 3. (c) Live/dead assay of HUVECs cultured on the MHC-PG hydrogel. (d) The morphology of HUVECs on the MHC-PG hydrogel. (e) H&E staining images of liver tissue at week 1 post-surgery. (f) Hemolytic index of G, TG and MHC-PG hydrogels. (g) Adhesion behavior of the MHC-PG hydrogel to various tissues and organs.

Hydrogel is a promising flexible electronics platform, and its conductivity and sensing characteristics are important indicators to evaluate its application potential. The mixture of tannic acid and borax endows MHC-PG hydrogel with a certain conductivity close to 0.055 S m⁻¹, which is higher than that of G and TG hydrogels (Fig. S14†). Fig. 4 shows the comprehensive sensing behavior of the MHC-PG hydrogel. It can be seen from Fig. 4a that the I–V electrical curve of MHC-PG hydrogel has ideal linearity in different strain ranges of 10–100%, indicating the stable ohmic characteristics of the sample.⁴⁰ Besides, there is a clear negative correlation between the slope of the I–V line and strain, meaning the distinct increase of the hydrogel resistance during stretching (Fig. 4a). Notably, the resistance change rate of the hydrogel always shows a good linear positive correlation with the deformation variable, both during stretching and recovery processes, and, moreover, the two curves demonstrate a satisfactory high degree of overlap, implying a negligible hysteresis of the hydrogel (Fig. 4b). When the hydrogel is subjected to step-like stress stretching, the resistance of the hydrogel also shows a step-like programmed increase and keeps better stability (Fig. 4c). Moreover, MHC-PG has an ultra-fast response rate, the response time is only about 0.45 s (Fig. 4d), meaning that it can immediately send electrical signal feedback at the moment of stretching and proving the excellent real-time sensing performance of the hydrogel. In practical applications, fatigue resistance is also regarded as one of the important performance evaluation criteria and has always been the focus of researchers' attention. It can be seen from Fig. 4e that the change of the MHC-PG hydrogel's resistance after 500 tensile–recovery cycles is still almost the same as that of the initial state, confirming a long service life, excellent anti-fatigue behavior of the MHC-PG and a brighter practical application prospect. Because of its excellent flexibility and self-adhesive performance, MHC-PG hydrogel is expected to be attached to the human body for effective monitoring of sports behavior. During the testing process, these stretchable conductive hydrogels can play the role of transducers, convert body movements into electrical signals, and record different body movements. As shown in Fig. 4f and g, the hydrogel resistance adhered to the volunteer's body would increase correspondingly due to the bending of the knee or wrist joints, and when the limb returns to the straight state, the resistance will also return to the initial value. Moreover, the bending angle of the finger can also be accurately recorded by MHC-PG hydrogel, and different bending angles of the finger would also lead to different resistance changes. For example, when the bend angle of the finger is 30°, the resistance increases by about 28%, while, when the bend angle is 90°, the resistance increases by more than 40% (Fig. 4h). More surprisingly, even subtle throat vibrations and coughs can be recorded accurately (Fig. 4i). However, it is important to note that hydrogels themselves contain a large amount of water, and dehydration caused by long-term use can lead to stiffening of the material, which in turn affects the flexibility and sensing behavior. Therefore, hydrogels, including MHC-PG, should be encapsulated to prevent water evaporation and ensure excellent performance.

Fig. 4 (a) I–V curves for different tensile deformations. (b) Relationship between MHC-PG hydrogel resistance change and strain deformation. (c) Change in hydrogel resistance with step strain. (d) Response and hysteresis time of the MHC-PG hydrogel. (e) Cyclic sensing performance of the MHC-PG hydrogel. The capture and sensing of wrist (f), knee (g), and finger (h) bending movements by the MHC-PG hydrogel. (i) Hydrogel resistance signal changes caused by throat vibration (coughing).

Event-related potentials (ERP) can effectively evaluate brain function by measuring EEG signals. ERP provides a reliable strategy to evaluate the performance of noninvasive MHC-PG BCI electrodes. Here, we conducted the N170 BCI test on volunteers in isolated rooms, and selected commercial electrodes and MHC-PG electrodes as control group and experimental group, respectively (Fig. 5a and b). During the N170 BCI test, two electrode positions, O1 and P3, were selected for signal analysis. From Fig. 5c, it can be confirmed that the contact impedance of the MHC-PG hydrogel is lower than that of the commercial electrode, which proves the close contact between the hydrogel and the head skin and is the premise for collecting high-precision EEG signals. In further tests, the waveforms of the EEG signals acquired by the MHC-PG hydrogel electrode and the commercial wet electrode are basically the same, and the amplitude of the voltage is within 5–100 μV, indicating the accuracy and reliability of the acquired EEG signals (Fig. 5d and S15†). The good self-adhesion of MHC-PG hydrogel enables it to be firmly attached to the substrate, which prevents the hydrogel from falling off and the interference of human motion on signals. Therefore, MHC-PG hydrogel electrodes can produce more pronounced signal fluctuations than commercial electrodes (Fig. 5e). Combined with the above analysis, the MHC-PG hydrogel electrode developed in this study has a stable ability to capture EEG signals, suggesting its bright application prospect in the field of non-invasive BCI.

Fig. 5 (a) Schematic of the EEG test. (b) Diagram of working and reference electrode positions. (c) Contact impedance of commercial wet and hydrogel electrodes. (d) Changes in EEG signals acquired by gel electrodes and commercially available wet electrodes during a 20 s test time. (e) Comparison of N170 potential at the O1 position between commercial wet and hydrogel electrodes.

4 Conclusions

In summary, a multilevel hybrid dynamic crosslinking gelatin-based hydrogel application matrix was synthesized using a simple one-pot method. In terms of molecular network, relying on the synergistic effect of the dynamic covalent bonds (reversible borate ester bonds) and non-covalent bonds (hydrogen bonds and molecular helix entanglement) as well as the introduction of functional groups (o-diphenol hydroxyl group), the MHC-PG hydrogel achieves an integration of excellent mechanical properties, strong self-healing behavior and satisfactory self-adhesive properties. In vitro and in vivo experiments confirmed that the hydrogel has extremely low cytotoxicity and ideal histocompatibility and does not cause breakage of erythrocytes or damage to mouse liver tissue. In addition, the hydrogel can be directly adhered to the surface of skin tissues for real-time monitoring of large-scale and subtle human movements with stable and repeatable signals. Moreover, benefiting from its suitable modulus, good flexibility and high conductivity, the hydrogel can serve as a BCI electrode and capture human brainwave signals. We believe that MHC-PG hydrogels will be promising candidates for future flexible wearable devices, biomedical carriers, and BCI materials.

Data availability

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

Author contributions

C. X. designed the concept and performed the main experiments. S. G., H. Z., and W. F. participated in the experimental testing and data analysis. X. Z. and X. D. designed the concept and planned and supervised this project. All authors wrote the manuscript.

Conflicts of interest

The authors declare no interest conflict. They have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was financially supported by the National Key Research and Development Program of China (Grant No. 2023YFB3810100), Dalian Science and Technology Innovation Program (Grant No. 2023JJ12SN031), Fundamental Research Funds for the Central Universities of China (Grant No. 2023YGZD03) and Dalian Medical Science Research Program (Grant No. 2311003).

Notes and references

1. Y. D. Shi, Y. J. Guan, M. J. Liu, X. C. Kang, Y. Tian, W. C. Deng, P. Yu, C. Y. Ning, L. Zhou, R. M. Fu and G. X. Tan, ACS Nano, 2024, 18, 3720–3732.
2. Y. Zhao, S. L. Song, X. Z. Ren, J. M. Zhang, Q. Lin and Y. L. Zhao, Chem. Rev., 2022, 122, 5604–5640.
3. G. Ge, Q. Wang, Y. Z. Zhang, H. N. Alshareef and X. C. Dong, Adv. Funct. Mater., 2021, 31, 2107437.
4. G. Ge, Y. Z. Zhang, X. Xiao, Y. T. Gong, C. Liu, C. Lyu, W. L. Ong, G. W. Ho, Z. Yang and W. Huang, Adv. Funct. Mater., 2024, 34, 2310963.
5. P. Yang, Z. Li, B. R. Fang and L. L. Liu, Int. J. Biol. Macromol., 2023, 253, 127612.
6. I. Y. Miranda-Valdez, M. Sourroubille, T. Mäkinen, J. G. Puente-Córdova, A. Puisto, J. Koivisto and M. J. Alava, Cellulose, 2024, 31, 1545–1558.
7. T. L. Wu, C. Y. Cui, C. C. Fan, Z. Y. Xu, Y. Liu and W. G. Liu, Bioact. Mater., 2021, 6, 2820–2828.
8. J. Y. Wu, Z. X. Zhang, Z. Y. Wu, D. S. Liu, X. X. Yang, Y. X. Wang, X. Jia, X. Xu, P. Jiang and X. L. Wang, Adv. Funct. Mater., 2023, 33, 2210395.
9. Y. N. Liang, Z. X. Wu, Y. N. Wei, Q. L. Ding, M. Zilberman, K. Tao, X. Xie and J. Wu, Nano-Micro Lett., 2022, 14, 52.
10. D. D. Ye, C. Y. Chang and L. N. Zhang, Biomacromolecules, 2019, 20, 1989–1995.
11. J. P. Zhang, Y. Hu, L. A. Zhang, J. P. Zhou and A. Lu, Nano-Micro Lett., 2023, 15, 8.
12. G. Ge, Y. Z. Zhang, W. L. Zhang, W. Yuan, J. K. El-Demellawi, P. Zhang, E. Di Fabrizio, X. C. Dong and H. N. Alshareef, ACS Nano, 2021, 15, 2698–2706.
13. J. Y. Yang, H. Wang, W. T. Huang, K. L. Peng, R. Shi, W. Tian, L. Z. Lin, J. J. Yuan, W. S. Yao, X. L. Ma and Y. Chen, Mater. Horiz., 2023, 10, 3797–3806.
14. Z. W. Wang, H. Wei, Y. J. Huang, Y. Wei and J. Chen, Chem. Soc. Rev., 2023, 52, 2992–3034.
15. X. Wang, X. Sun, D. Gan, M. Soubrier, H. Y. Chiang, L. Yan, Y. Li, J. Li, S. Yu, Y. Xia, K. Wang, Q. Qin, X. Jiang, L. Han, T. Pan, C. Xie and X. Lu, Matter, 2022, 5, 1204–1223.
16. T. D. Lei, J. J. Pan, N. Wang, Z. P. Xia, Q. S. Zhang, J. Fan, L. Tao, W. Shou and Y. Gao, Mater. Horiz., 2024, 11, 1234–1250.
17. Z. Q. Yu, X. Y. Li, X. H. Li, B. R. Zheng, D. Z. Li, D. K. Xu and F. H. Wang, Adv. Funct. Mater., 2023, 33, 2305995.
18. J. Y. Yu, S. E. Moon, J. H. Kim and S. M. Kang, Nano-Micro Lett., 2023, 15, 51.
19. S. Q. Guan, C. Xu, X. F. Dong and M. Qi, J. Mater. Chem. A, 2023, 11, 15404–15415.
20. S. Houben, A. A. Aldana, A. S. Huysecom, W. Mpinganzima, R. Cardinaels, M. B. Baker and L. M. Pitet, ACS Appl. Polym. Mater., 2023, 5, 1819–1827.
21. B. Yang, J. L. Song, Y. H. Jiang, M. Li, J. J. Wei, J. J. Qin, W. J. Peng, F. L. Lasaosa, Y. Y. He, H. L. Mao, J. Yang and Z. W. Gu, ACS Appl. Mater. Interfaces, 2020, 12, 57782–57797.
22. Y. Q. Liang, H. R. Xu, Z. L. Li, A. D. Zhangji and B. L. Guo, Nano-Micro Lett., 2022, 14, 185.
23. L. Han, K. Z. Liu, M. H. Wang, K. F. Wang, L. M. Fang, H. T. Chen, J. Zhou and X. Lu, Adv. Funct. Mater., 2018, 28, 1704195.
24. L. Wu, L. W. Li, M. J. Qu, H. Wang and Y. Z. Bin, ACS Appl. Polym. Mater., 2020, 2, 3094–3106.
25. J. Wang, T. Y. Dai, H. Wu, M. Y. Ye, G. L. Yuan and H. B. Jia, Compos. Sci. Technol., 2022, 221, 109345.
26. Y. Bai, S. Q. Yan, Y. B. Wang, Q. Wang and X. Duan, ACS Appl. Mater. Interfaces, 2023, 15, 48744–48753.
27. H. J. Zhang, X. Y. Wang, L. N. Wang, T. L. Sun, X. G. Dang, D. R. King and X. Y. You, Soft Matter, 2021, 17, 9399–9409.
28. H. H. Cui, B. B. Cui, H. Y. Chen, X. W. Geng, X. C. Geng, Z. R. Li, S. K. Cao, J. L. Shen and J. G. Li, Biomater. Sci., 2023, 11, 4226–4237.
29. Z. Z. Hui, X. Y. Pan, Y. Li, C. S. Zhang, X. Z. Zuo, J. Tang, Y. P. Wang, N. N. Qiu, S. B. Zheng, X. J. Ye, R. Z. Hu, D. P. Song, W. Fang, J. Yang and G. Q. Yan, Int. J. Biol. Macromol., 2024, 260, 129529.
30. B. Qin, S. Y. Liu, Z. H. Huang, L. D. Zeng, J. F. Xu and X. Zhang, Nat. Commun., 2022, 13, 7595.
31. S. Cho, S. Y. Hwang, D. X. Oh and J. Park, J. Mater. Chem. A, 2021, 9, 14630.
32. X. Wan, T. S. Mu and G. P. Yin, Nano-Micro Lett., 2023, 15, 99.
33. J. H. He, Z. X. Zhang, Y. T. Yang, F. G. Ren, J. P. Li, S. J. Zhu, F. Ma, R. Q. Wu, Y. Lv, G. He, B. L. Guo and D. K. Chu, Nano-Micro Lett., 2021, 13, 80.
34. Y. W. Chen, Y. Y. Zhou, Z. H. Hu, W. Y. Lu, Z. Li, N. Gao, N. Liu, Y. R. Li, J. He, Q. Gao, Z. J. Xie, J. C. Li and Y. He, Nano-Micro Lett., 2024, 16, 34.
35. H. Y. Jiang, N. B. Bao, J. G. Tang and H. J. Li, Chem. Eng. J., 2023, 455, 140573.
36. C. Xu, P. Xu, Y. Gao, F. Gao, X. J. Zhuang, H. Zhang and X. F. Dong, Compos. Commun., 2023, 43, 101743.
37. L. Zhao, Z. J. Ren, X. Liu, Q. J. Ling, Z. J. Li and H. B. Gu, ACS Appl. Mater. Interfaces, 2021, 13, 11344–11355.
38. S. Rahmani, A. Olad and Z. Rahmani, Polym. Bull., 2023, 80, 4117–4138.
39. S. S. Xue, Y. P. Wu, G. F. Liu, M. L. Guo, Y. H. Liu, T. Zhang and Z. H. Wang, J. Mater. Chem. A, 2021, 9, 5730–5739.
40. L. Wu, L. W. Li, M. S. Fan, P. Tang, S. T. Yang, L. J. Pan, H. Wang and Y. Z. Bin, Composites, Part A, 2020, 138, 106050.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ta04767b

---