Microsphere‐Loaded and Borax‐Reinforced Polyacrylic Acid (PAA)/Polyvinyl Alcohol (PVA) Hydrogels Incorporating Tannic Acid: Ultra‐High Toughness and Antibacterial Properties for Artificial Intelligence Skin

ABSTRACT

Hydrogel‐based dressings are widely used in wound healing. Herein, we report a polyacrylic acid (PAA)/polyvinyl alcohol (PVA) composite hydrogel fabricated via self‐catalyzed free radical polymerization, with borax serving as the reinforcing phase. Meanwhile, metal‐ligand coordination bonding between tannic acid (TA) and Fe^{3 +} further elevates the crosslinking density of the hydrogel network. Magnetic chitosan microspheres (MCMs) were synthesized by emulsion cross‐linking and loaded with two antibacterial agents, namely tetracycline hydrochloride (TH) and berberine hydrochloride (Bbh). The incorporation of MCMs into the hydrogel matrix resulted in the development of a multifunctional composite hydrogel suitable for wound dressing applications. Results demonstrated that the composite hydrogel containing a specific concentration of 4‰ (w/v) borax and 20 mg/mL MCMs exhibited superior performance, including enhanced mechanical strength, improved responsiveness, sustained drug release, and potent antibacterial efficacy. The core novelty of this work lies in the synergistic integration of borax‐based mechanical reinforcement, MCM‐mediated dual drug loading and sustained release, and the self‐catalyzed polymerization system. This innovative structural and functional collaboration effectively optimizes the mechanical stability of the hydrogel dressing and achieves synergistic antibacterial and intelligent responsive therapeutic performances, providing a reliable and high‐efficiency candidate for advanced wound care and next‐generation wound dressing applications.

The core novelty of this work lies in the synergistic integration of borax‐based mechanical reinforcement, MCM‐mediated dual drug loading and sustained release, and the self‐catalyzed polymerization system, providing a reliable and high‐efficiency candidate for advanced wound care and next‐generation wound dressing applications.

1. Introduction

Wound dressings represent a pivotal advancement in wound care technology, integrating enhanced functionalities to optimize the healing process [1, 2, 3, 4, 5]. These dressings encompass films, foams, fibers, and hydrogels, which specifically engineered to maintain a moist wound micro‐environment, manage exudate efficiently, and leverage endogenous enzymes to facilitate the breakdown of necrotic tissue [6]. Hydrogel dressings, a widely adopted category in contemporary wound management, are particularly suitable for dry wounds, mild‐to‐moderate burns, abrasions, lacerations, and other injuries requiring a hydrated healing milieu. Typically composed of water and crosslinked polymers, e.g., polyvinyl alcohol (PVA), polyacrylate salts, or carboxymethyl cellulose, their defining characteristic is a high water content (exceeding 90%), which imparts cooling and soothing effects upon contact with wound surfaces [7].

PVA is a pivotal synthetic biomacromolecule with excellent biocompatibility, non‐toxicity, biodegradability, and processability, which render it widely applicable in biomedical fields [8, 9, 10]. However, PVA exhibits inert bioactivity, and its standalone form is unsuitable for wound dressings due to three key limitations, e.g., insufficient exudate absorption capacity, low elasticity, and a lack of drug delivery capabilities [11, 12]. In contrast, polyacrylic acid (PAA), a representative biomacromolecule polymerized from acrylic acid, features abundant carboxyl groups (–COOH) on its molecular chains. These groups confer exceptional hydrophilicity, pH responsiveness, and ion‐exchange capacity, establishing PAA as a critical bio‐intelligent material for applications such as smart drug delivery systems, pH‐responsive biosensors, and tissue engineering scaffolds [13, 14, 15, 16, 17, 18].

Generally, hydrogel dressings prepared by only these two biopolymers are unable to meet the requirements for artificial intelligence skin application. The term “intelligence” for artificial intelligence‐based skin refers to the skin's autonomous intelligent capabilities: it can actively perceive external mechanical, temperature stimuli via integrated flexible sensors, automatically convert collected physical signals into readable electrical data, and realize real‐time signal feedback without extra manual control. Therefore, the introduction of the reinforcing phases is needed to improve the properties of the PAA/PVA hydrogel dressings [19, 20]. Borax is a water‐soluble compound with low acute oral toxicity and no dermal or respiratory sensitization [21]. Wang et al. [22] demonstrated that incorporating borax into PVA hydrogels enhances their mechanical properties. Tannic acid (TA), a polyphenolic compound derived from natural plants, contains abundant pyrogallol/catechol groups. These functional moieties enable strong tissue adhesion, which prevents hydrogel displacement caused by wound exudate or hemorrhage at the injury site [23, 24, 25].

Furthermore, biopolymer‐based microspheres enable efficient drug encapsulation and controlled release [26, 27, 28, 29, 30, 31]. For example, magnetic chitosan microspheres (MCMs), characterized by their small particle size and structural stability, can be incorporated into hydrogels to enhance functionality while avoiding undesirable interactions between therapeutic agents and hydrogel precursors. By loading antibacterial drugs into magnetic microspheres, slow drug release can be targeted to specific regional sites and reduce the toxic effects on other sites. We speculate that development of a multifunctional PAA/PVA composite hydrogel dressing with ultra‐high toughness and enhanced antibacterial activity would broaden the applications for artificial intelligence skin.

Herein, we integrate PAA's high water absorption capacity, pro‐healing properties, and tunable mechanical characteristics with PVA's biocompatibility and superior mechanical performance to fabricate a multifunctional hydrogel dressing. Borax and MCMs were employed as reinforcing phases, while TA reacted with Fe^{3 +} ions to form the TA‐Fe^{3 +} redox pair, a system that enables rapid self‐catalyzed hydrogel formation at room temperature [19, 20, 32]. Two model antibacterial drugs were incorporated into the hydrogel, e.g., tetracycline hydrochloride (TH), a water‐soluble agent, and berberine hydrochloride (Bbh), a poorly soluble agent. In vitro experiments were conducted to systematically investigate the effects of borax and MCM content on the hydrogel's morphology, gelation time, swelling behavior, mechanical properties, and sensing performance. Additionally, the hydrogel's drug release kinetics and antibacterial efficacy were evaluated.

2. Experimental Section

2.1. Materials

Polyvinyl alcohol (PVA‐12) was purchased from Shanghai Macklin Biochemical Technology Co., Ltd. Acrylic acid (AA, >97%) and tannic acid (TA, 96%) were obtained from Shanghai Macklin Biochemical Technology Co., Ltd. and Shanghai Yuanye Bio‐Technology Co., Ltd., respectively. Ferric chloride hexahydrate (Fe₃O₄•6H₂O) was supplied by Nanjing Chemical Reagent Co., Ltd. Sodium tetraborate decahydrate (Borax) was procured from Chengdu Kelong Chemical Co., Ltd. Chitosan (CS, deacetylation degree: 93%, MW: 20 kDa) and N,N′‐methylenebisacrylamide (BIS, 99%) were acquired from Aladdin Industrial Corporation. Ammonium persulfate (APS, AR ≥98%), glycerol, and glutaraldehyde (GA) were sourced from Sinopharm Chemical Reagent Co., Ltd. Tetracycline hydrochloride (TH) and berberine hydrochloride (Bbh) were provided by Aladdin Industrial Corporation. Staphylococcus aureus (ATCC 25923, S. aureus) and Escherichia coli (ATCC 25922, E. coli) were obtained from Nanjing Lezhen Biotechnology Co., Ltd. All chemicals were used without further purification.

2.2. Preparation of Hydrogels

0.1 g of PVA was dissolved in 7 mL of deionized water at 95°C to form a transparent homogeneous solution. After cooling the solution to room temperature, specified amounts of borax were added to achieve final concentrations of 2‰, 4‰, 6‰, and 8‰ (m/v). Subsequently, 3 mL of AA was added to the mixture under vigorous stirring. Next, 0.02 g of FeCl₃•6H₂O, 0.02 g of TA, and 2 mg of BIS were sequentially incorporated into the borax/PVA/AA solution, resulting in a yellow transparent mixture. Finally, 0.02 g of APS was added to the mixture. Once APS was completely dissolved, the precursor solution was quickly transferred into molds, which were then sealed. The mixture was allowed to stand for 1 h to complete the gelation process [20].

2.3. Preparation of Composite Hydrogels

Magnetic chitosan microspheres (MCMs) were synthesized via emulsion polymerization. The aqueous phase was prepared by dissolving 0.1 g of chitosan powder and 0.01 g of Fe₃O₄ MNPs in 10 mL of a 3% (v/v) AA solution. The oil phase comprised 50 mL of liquid paraffin and 1.5 mL of Span‐80. Under continuous stirring at 60°C and 800 rpm, 5 mL of the aqueous phase was gradually added to the oil phase, followed by emulsification for 90 min to form a homogeneous water‐in‐oil (W/O) emulsion. Subsequently, 1 mL of a 25% (v/v) GA solution was added as a crosslinking agent. After reacting for 10 min, MCMs were collected by centrifugation at 6,000 rpm for 5 min. The collected microspheres were then alternately washed with petroleum ether and ethanol (2–3 cycles) and finally freeze‐dried [33, 34]. Composite hydrogels were fabricated by incorporating different ratios of MCMs into the hydrogel matrix, using the same preparation conditions as described above. The particle size distribution of the microspheres was analyzed using a laser particle size analyzer (BT‐9300S, China). The drug release profiles under different pH conditions were evaluated with TH as the model drug, via UV spectrophotometry (Biomate 3S, USA).

2.4. Morphological and Structure Characterization

Surface morphology of hydrogels and microspheres was observed using a field‐emission environmental scanning electron microscope (SEM) (Quant 250FEG, FEI, USA) at an accelerating voltage of 3 kV. Samples were gold‐coated with a sputter coater (Q150T, QT, USA) for 1.5 min (hydrogels) and 2 min (microspheres). Fourier‐transform infrared (FTIR) spectra of hydrogels and composite hydrogels were recorded on a Nicolet iS10 spectrometer (Thermo Fisher Scientific, USA) at room temperature using KBr pellets. Spectral acquisition covered the wavenumber range of 500–4,000 cm^{− 1} with a resolution of 0.48 cm^{− 1}. X‐ray diffraction (XRD) analysis was performed using a Bruker D8 diffractometer (Bruker, USA) with Cu‐Kα radiation (λ = 1.5406 Å). Scans were conducted over a 2θ range of 10–90° at a scanning rate of 2°·min^{− 1}.

2.5. Swelling Behavior

Swelling behavior of hydrogels was evaluated by immersing lyophilized hydrogels (initial mass denoted as W_d) in PBS solution maintained at 37°C. At predetermined time intervals, the hydrogels were removed, blotted to eliminate surface moisture, and weighed (recorded as Wt). The swelling ratio (SR) was calculated using the formula of (W_t‐W_d)/W_d.

2.6. Rheological Testing

Rheological properties of hydrogels were analyzed using a rheometer (MCR102, Anton Paar, Austria). Cylindrical hydrogel specimens (20 mm diameter × 2 mm thickness) were subjected to frequency sweep tests from 0.01 to 100 Hz at a constant strain amplitude (γ = 1%). All measurements were performed at 37°C with a temperature‐controlled stage. Storage modulus (G′) and loss modulus (G″) were recorded as functions of angular frequency.

2.7. Tensile and Sensing Testing

Tensile and sensing properties of hydrogel specimens (15 mm length × 10 mm width × 2 mm thickness) were evaluated using a universal testing machine (Instron 5943, USA). The cross‐head speed was set to 40 mm/min, and tensile tests were conducted until the yield point was reached. Sensing capability of the hydrogels was evaluated coupled with a digital source meter. Real‐time resistance changes were monitored during controlled elongation until the fracture strain threshold was reached, with a constant test current of 0.002 A. Cyclic loading‐unloading tests were conducted on a custom‐built apparatus to investigate the relative resistance variation (ΔR/R₀) under programmed strains, thus quantifying the strain sensitivity of the hydrogels. When used as a wearable strain sensor, the hydrogel was employed to detect biomechanical motions via resistance responses, including joint flexion (wrist, elbow, and knee), multi‐angle finger bending, swallowing, blinking, and phonation. The authors declare that this study was conducted in accordance with the principles of the Declaration of Helsinki. As the Nanjing University of Science and Technology Ethics Committee does not issue formal ethics approval or accreditation numbers for this type of research, an official approval number was not applicable. However, it was explicitly stated that informed consent was obtained from all participants involved in this study prior to their participation.

2.8. Drug Release

10 mg (W_m) of microspheres were dispersed in 10 mL of PBS and disintegrated via ultrasonication to ensure complete drug release. The characteristic absorption peaks of TH and Bbh at 363 and 344 nm, respectively, were measured using UV spectrophotometry. Drug concentrations were quantified using the standard curves of TH and Bbh, allowing calculation of the actual drug loading amount (W_drug). The drug loading rate (L) and encapsulation efficiency (E) were calculated using the formulas of (W_drug/W_m)*100% and (W_drug/Q_d)*100%, respectively, where Qd denotes the initial amount of drug added to the microspheres.

For the drug release study, 8 mg of MCMs and 8 mg of MCMs‐incorporated hydrogel were separately placed in centrifuge tubes containing 8 mL of PBS. The tubes were incubated in a thermostatic shaker at 37°C and 70 rpm. At predetermined time intervals, 2.5 mL aliquots were withdrawn to measure absorbance at specific wavelengths (363 nm for TH and 344 nm for Bbh). After each sampling, 2.5 mL of fresh PBS was replenished to maintain a constant total volume.

2.9. Antibacterial Testing

Antibacterial efficacy of the hydrogels was initially evaluated via the agar diffusion assay against Gram‐positive (S. aureus) and Gram‐negative (E. coli) bacteria. Bacterial suspensions (10⁷–10⁸ CFU/mL) were evenly spread on LB agar plates, followed by the placement of hydrogel discs (10 mm in diameter × 2 mm in thickness). After 24 h of incubation at 37°C, the diameters of the inhibition zones were measured to quantify the antimicrobial activity [35]. For the direct contact assessment, hydrogel discs (10 mm × 2 mm) were placed in 48‐well plates. A 100 µL aliquot of bacterial suspension (10⁶ CFU/mL of S. aureus or E. coli) was pipetted onto the surface of each hydrogel, followed by 6 h of incubation at 37°C in a humidified chamber [36]. Subsequently, 1900 µL of sterile PBS was added to each well, and the bacterial cells were resuspended via ultrasonication. A 100 µL aliquot of the resuspended solution was plated on solid agar and incubated at 37°C for 24 h [37]. The bactericidal efficiency (η) was calculated using the colony counting method with blank controls.

2.10. Statistical Analysis

All data were analyzed using analysis of variance (ANOVA). Statistical significance was defined as p < 0.05. Unless otherwise specified, results were expressed as mean ± standard deviation (n = 3).

3. Results and Discussion

3.1. Preparation of Microspheres and Hydrogels

The fabrication strategy for the MCMs/gel composite is illustrated in Scheme 1. Borate anions form borate ester bonds with the hydroxyl groups on the PVA backbone, establishing the first crosslinking network. Subsequently, TA coordinates with Fe^{3 +} to generate TA‐Fe^{3 +} redox pairs. These pairs activate APS to produce sulfate radicals (SO₄•⁻), which further generate hydroxyl radicals (•OH). The hydroxyl radicals then initiate the polymerization of acrylic acid (AA) monomers to construct the second crosslinking network, ultimately forming the hydrogel [20, 32]. By synergistically integrating the advantages of PVA and PAA, including high water absorption capacity, pro‐healing properties, and tunable mechanical characteristics, as well as biocompatibility and superior mechanical performance, a multifunctional composite hydrogel dressing was fabricated. Self‐catalysis eliminates extra exogenous catalysts, simplifying the preparation procedure and avoiding residual catalyst impurities inside hydrogel matrices. It enables mild polymerization conditions and improves biocompatibility of the obtained hydrogel.In‐situ generated catalytic sites continuously accelerate monomer crosslinking throughout the reaction. It regulates crosslinking rate homogenously, facilitating the construction of uniform and compact three‐dimensional crosslinked network.

SCHEME 1.

Schematic of the preparation principles of polyacrylic acid (PAA)/polyvinyl alcohol (PVA) composite hydrogel (a) and magnetic chitosan microspheres (MCMs) (b). The composite hydrogel was fabricated via self‐catalyzed free radical polymerization, with borax serving as the reinforcing phase. MCMs were synthesized by emulsion cross‐linking and loaded with two antibacterial agents, namely tetracycline hydrochloride (TH) and berberine hydrochloride (Bbh).

For the preparation of drug‐loaded MCMs, TH and Bbh were separately dissolved in a chitosan‐acetic acid solution (Scheme 1b). The TH/MCMs system exhibited a drug loading capacity (DLC) of 13.53 ± 1.02 µg/mg and an encapsulation efficiency (EE) of 14.89 ± 1.12%, whereas the Bbh/MCMs system showed corresponding values of 8.68 ± 0.05 µg/mg and 11.94 ± 0.07%, respectively. The distinct differences in loading performance between the two drugs originate from their inherent physicochemical properties and interfacial interaction behaviors with the chitosan microsphere matrix. Specifically, TH is a highly hydrophilic small‐molecule antibiotic that can be uniformly and stably dispersed in the aqueous chitosan solution during emulsion cross‐linking. Its favorable water compatibility enables sufficient contact and effective physical adsorption and electrostatic interaction with the porous chitosan network, thereby achieving higher encapsulation efficiency and loading content. In contrast, Bbh possesses poor aqueous solubility and strong hydrophobic molecular characteristics, which fundamentally limit its homogeneous dispersion in the aqueous reaction system. The incomplete dissolution and partial molecular aggregation of Bbh greatly reduce the effective concentration of free drug molecules available for encapsulation. Furthermore, the hydrophobic interaction between Bbh aromatic structures and the chitosan skeleton weakens the binding affinity between drug molecules and the carrier, further inhibiting drug embedding during microsphere formation. Combined with the larger molecular steric hindrance of Bbh compared with TH, these comprehensive factors collectively lead to the relatively lower drug loading and encapsulation efficiency of the Bbh/MCMs system. Nevertheless, the moderate and stable loading of Bbh avoids excessive short‐term drug accumulation, which complements the high and rapid loading characteristics of TH, laying a foundation for the subsequent dual‐drug synergistic release and long‐term antibacterial performance of the composite hydrogel [33, 34].

3.2. Morphology of Microspheres and Hydrogels

As shown in Figure 1a, SEM images revealed that the microspheres exhibited a spherical morphology with smooth surfaces. The incorporation of TH and Bbh caused negligible alterations to the microspheres’ morphology. Further analysis confirmed the absence of particle coalescence, with most microspheres maintaining discrete structures, a critical feature that ensures uniform dispersion in composite materials. Corresponding particle size distribution was statistically quantified based on more than 200 randomly selected microspheres from multiple SEM fields of view. Statistical sizing results demonstrated that the prepared microspheres possessed a narrow size distribution ranging from 100 to 300 µm, with an average diameter of 186.4 ± 22.7 µm. The relatively small and centralized particle size minimized structural disruption when the microspheres were embedded in hydrogel matrices, which is consistent with the discrete morphological characteristics observed from SEM observation.

FIGURE 1.

(a) Morphology of MCMs. (b) Size distribution profiles of chitosan microspheres, magnetic chitosan microspheres, and drug‐loaded variants.

SEM images of the pure hydrogel (Gel) and MCMs‐incorporated hydrogel (MCMs/gel) are presented in Figure 2. The incorporation of borax effectively regulates the pore microstructure of the hydrogel system. Specifically, borax acts as a pore‐regulating additive and crosslinking auxiliary, which optimizes the nucleation and growth of ice crystals during freezing and lyophilization, thereby significantly increasing the density of uniform micropores in the hydrogel matrix. The increased micropore density not only constructs a more refined and stable network skeleton that enables the hydrogel to withstand higher tensile stress and achieve enhanced tensile strength, but also provides abundant attachment sites for cells, effectively facilitating cell adhesion, growth, and migration and thus improving the overall biocompatibility of the hydrogel material. Moreover, the introduction of microspheres exerts a concentration‐dependent dual regulatory effect on the internal pore structure of the hydrogel. A moderate amount of uniformly dispersed microspheres can be well embedded in the 3D porous network without damaging the original pore structure. The well‐distributed microspheres achieve homogeneous stress dispersion under external force, and further optimize the pore size distribution and structural uniformity of the hydrogel, which significantly improves the flexibility and ductility of the composite hydrogel and retains the structural advantages of the porous matrix. However, excessive addition of microspheres will cause severe particle aggregation due to interfacial incompatibility. The aggregated microspheres squeeze and occupy the original pore space of the hydrogel matrix, leading to gradual shrinkage, diminishment and even partial closure of intrinsic pores. This irreversible pore collapse and structural disruption destroys the integrity of the 3D porous network, eliminates the structural advantages of the porous matrix, and results in dense stacking of microspheres in the matrix. Such abnormal structural evolution inevitably deteriorates the internal pore connectivity and uniformity, which ultimately impairs the mechanical properties and structural stability of the composite hydrogel.

FIGURE 2.

SEM images of hydrogels at low magnification (a–e) and high magnification (a’–e’).

3.3. Swelling Behavior

The swelling behavior of hydrogel dressings directly reflects their capacity to absorb wound exudate and maintain a moist wound micro‐environment, both critical for facilitating wound healing. The swelling profiles of the pure hydrogel (Gel) and MCMs‐incorporated hydrogel (MCMs/gel) are presented in Figure 3. For the pure Gel group, the swelling ratio rises rapidly at the initial fast‐water‐absorption stage driven by capillary penetration and hydrophilic group hydration, followed by a markedly slowed growth after 24 h; the testing was prolonged continuously until the swelling curve levelled off with negligible fluctuation of swelling ratio, marking the arrival of equilibrium swelling. With increasing borax content, the equilibrium swelling ratio of pristine hydrogel follows an initial decline and subsequent ascending tendency, and all formulations maintain a relatively high saturated swelling value at equilibrium, which guarantees efficient exudate uptake and long‐lasting moist surroundings at wound sites. In terms of MCMs/gel composites, their equilibrium swelling capacity varies with MCMs loading and displays an upward‐then‐downward tendency. At moderate microsphere dosage, abundant hydrophilic chitosan‐based MCMs introduce extra hydroxyl and amino groups to the hydrogel network; these polar functional groups form extensive hydrogen bonds with surrounding aqueous molecules to capture abundant water, improving saturated water absorption at equilibrium. Nevertheless, excessive MCMs addition triggers particle aggregation, which destroys partial interconnected porous channels inside the matrix, blocks water infiltration pathways and accordingly lowers the final equilibrium swelling ratio. Such concentration‐dependent variation in equilibrium swelling is originated from the intrinsic hydrophilic property of chitosan microspheres as well as the pore structural alternation induced by filler dosage.

FIGURE 3.

(a) Swelling ratios of hydrogels with varying Borax contents. (b) Swelling ratios of composite hydrogels with different MCMs loadings. Values reported are an average of n = 3, ± standard deviation.

3.4. Characterization of MCMs and Hydrogels

Figure 4 presents the XRD patterns of chitosan microspheres (CMs), magnetic chitosan microspheres (MCMs), and Fe₃O₄ magnetic nanoparticles (MNPs). Comparative analysis revealed distinct diffraction peaks at 35.6° and 63.0° for both Fe₃O₄ MNPs and MCMs, peaks that were absent in CMs. This confirms the successful incorporation of Fe₃O₄ MNPs into the MCM structure [33, 38]. The magnetic responsiveness of MCMs is demonstrated in Figure 4d. Under a magnetic field, MCMs were rapidly attracted upward in microcentrifuge tubes. When magnets arranged to form the characters “NJUT” were placed beneath a petri dish containing MCMs dispersed in deionized water, the microspheres migrated and assembled into identical patterns, verifying their strong magnetic actuation capability.

FIGURE 4.

(a) XRD patterns of Fe₃O₄, CMs, and MCMs. (b) FTIR spectra of GA, CS, CMs, and MCMs. (c) FTIR spectra of pure drugs and drug‐loaded microspheres. (d) Magnetic responsiveness of MCMs under external fields.

FTIR spectral analysis (Figure 4b) provides critical structural information for glutaraldehyde (GA), chitosan (CS), CMs, and MCMs. A prominent absorption band at 3285 cm^{− 1} originates from the overlapping stretching vibrations of O‐H and N‐H bonds—characteristic of the hydroxyl and amine functionalities in chitosan. Notably, the appearance of a carbonyl (C = O) stretching vibration at 1712 cm^{− 1} confirms the Schiff‐base reaction between GA and the amino groups of chitosan, while the emergent peak at 1650 cm^{− 1} provides direct evidence of imine bond (C = N) formation. Importantly, the FTIR spectral profile remained virtually unchanged after the incorporation of Fe₃O₄ MNPs. This strongly suggests that the magnetic nanoparticles were physically entrapped within the microsphere matrix rather than chemically bonded to it. Comparative analysis (Figure 4c) further confirms successful drug encapsulation. TH‐loaded MCMs (TH/MCMs) and Bbh‐loaded MCMs (Bbh/MCMs) exhibit distinct new bands at 1643 cm^{− 1} and 1504 cm^{− 1}, respectively. These bands precisely match the characteristic vibrations of crystalline TH and Bbh, as documented in previous studies [39, 40].

Figure 5a presents the FTIR spectrum of the hydrogel, with key absorption peaks at 2928 cm^{− 1}, corresponding to C‐H stretching vibrations of the methylene groups in PAA/PVA, and 1690 cm^{− 1} (attributed to –COOH vibrations). The disappearance of the C═C stretching band of acrylic acid at 1634 cm^{− 1} confirms the complete conversion of acrylic acid monomers to PAA via polymerization. A critical observation is evident in the spectrum of the borax‐modified hydrogel. The appearance of a B─O stretching vibration at 2358 cm^{− 1}, a peak absent in the unmodified hydrogel. This verifies the formation of borate ester bonds between borax and the hydroxyl groups of PVA.

FIGURE 5.

(a) Infrared spectra of Borax, PVA, TA, AA, gel, and Borax/gel. (b, c) Infrared spectra of MCMs/gel.

The structural integrity of the hydrogel composites is evident in Figure 5b,c. While the FTIR spectra of the MCMs‐incorporated hydrogel (MCMs/gel) bear striking similarity to those of the pure hydrogel, indicating the preservation of the hydrogel's crosslinking architecture, examination reveals a subtle 2 cm^{− 1} redshift in the carbonyl (C═O) peak (from 1693 cm^{− 1} to 1691 cm^{− 1}) following MCMs incorporation. This minor yet consistent shift, observed across various microsphere concentrations (Figure 5b), confirms the successful physical embedding of MCMs without disrupting the hydrogel's fundamental chemical structure, an observation that aligns with previous reports on composite material characterization [16, 41].

Figures 6a and 5b show the sol‐gel transition images of the pure hydrogel (Gel) and MCMs/gel, while their macroscopic morphologies are displayed in Figures 6c and 5d. The incorporation of borax did not alter the hydrogel's color or macroscopic structure, regardless of borax concentration. With the addition of MCMs, black particulate microspheres became visible within the hydrogel matrix. At MCMs loadings of 10 mg/mL and 20 mg/mL, the microspheres were uniformly distributed. In contrast, the composite with 30 mg/mL MCMs exhibited distinct black aggregates due to particle clustering.

FIGURE 6.

The images of gel (a) and MCMs/gel (b) in solution and gel states, and the images of gel (c) and MCMs/gel (d), along with the gelation time of gel and MCMs/gel. Values reported are an average of n = 3, ± standard deviation.

As shown in Figures 6e and 5f, the gelation time of the hydrogel exhibited a positive correlation with borax concentration. Specifically, when the borax content reached 8‰ (m/v), the gelation time increased to 126 s. This delay is attributed to the suppression of free radical activity by borax, which in turn decelerates the polymerization kinetics of acrylic acid monomers. Furthermore, the incorporation of MCMs extended the gelation time from 48 s to 201 s. This is because the microspheres act as physical barriers that hinder interactions between polymer chains and the crosslinking of PVA/PAA segments.

3.5. Mechanical Properties

The rheological properties of the hydrogels and composite hydrogels were evaluated by measuring their storage modulus (G′, reflecting elastic behavior) and loss modulus (G″, reflecting viscous behavior) at 37°C [42, 43], as shown in Figure 7a,d. Across the tested frequency range (from low to high), G′ consistently exceeded G″ by approximately one order of magnitude, clear evidence of a stable elastic 3D network formation. With increasing borax content, interactions between borax, PVA, and PAA were enhanced, generating additional crosslinking points and thereby improving shear resistance. However, excessive borax loading led to a reduction in G″, which is attributed to over‐densification of the crosslinked network (a structure that restricts viscous flow). For the MCMs‐incorporated hydrogel (MCMs/gel), both G′ and G″ increased with rising MCMs content. This phenomenon stems from two key effects: microsphere‐induced reinforcement of crosslinking density, and specific interactions between functional groups (hydroxyl and amino groups) on the microsphere surface and PVA/PAA polymer chains.

FIGURE 7.

The storage modulus and loss modulus of gel (a) and MCMs/gel (d). The stress‐strain curves of gel (b) and MCMs/gel (e). The maximum stress and strain of gel (c) and MCMs/gel (f). Values reported are an average of n = 3, ± standard deviation.

The tensile properties of the hydrogels were characterized via stress‐strain curves (Figure 7b–e), with the maximum tensile stress and elongation at break summarized in Figure 7c,f. As the borax concentration increased from 0‰ to 8‰, the maximum tensile stress first rose from 0.029 MPa to 0.083 MPa, then declined to 0.079 MPa. Concurrently, the elongation at break increased from 494.8% to 615.9%, followed by a drop to 470.4%. This trend reveals an inherent trade‐off between mechanical strength and ductility: moderate crosslinking strengthens the hydrogel by stabilizing the network, while excessive crosslinking restricts the mobility of polymer chains, ultimately inducing brittleness [43, 44, 45].

For the MCMs/gel composite, the tensile performance exhibited a non‐monotonic response to changes in MCMs content. Specifically, at the optimal MCMs loading, the elongation at break peaked at 952% (a significant increase from 579% for the pure hydrogel), and the tensile stress reached 0.08 MPa (up from 0.06 MPa for the pure hydrogel). At low MCMs loadings, the microspheres disrupted the continuous hydrogel matrix, weakening the material's mechanical integrity. In contrast, an optimal MCMs content facilitated effective stress distribution via microsphere bridging, where microspheres act as “load‐transfer nodes” between polymer chains. However, excessive MCMs loading (>30 mg/mL) reduced the hydrogel's free volume and triggered stress concentration, two factors that collectively compromised the hydrogel's elasticity and fracture resistance.

3.6. Sensing Performance

The sensing capability of the hydrogels was characterized by monitoring relative resistance changes (ΔR/R₀) during tensile deformation. As shown in Figure 8a,c, ΔR/R₀ increased parabolically with rising strain. The gauge factor (GF), defined as the ratio of ΔR/R₀ to strain, first increased and then decreased with increasing borax content, peaking at a borax concentration of 4‰ (Figure 8a). This trend is likely attributed to the formation of an optimized 3D network at moderate borax concentrations, which facilitates ionic conduction via FeCl₃. In contrast, excessive borax (>4‰) induces over‐crosslinking of the network, restricting ion mobility and thereby reducing the GF.

FIGURE 8.

The sensitivity of gel (a) and MCMs/gel (c). The sensitivity of gel with a Borax content of 4‰. The sensitivity of MCMs/gel with an MCMs content of 20 mg/mL. Values reported are an average of n = 3, ± standard deviation.

The MCMs‐incorporated hydrogel (MCMs/gel) with an MCMs loading of 20 mg/mL exhibited the highest GF (Figure 8c). Quantitative analysis revealed GF values of 1.47 (under low strain) and 4.26 (under high strain) for the hydrogel containing 4‰ borax (4‰ Borax/gel) (Figure 8b), whereas the MCMs/gel achieved GF values ranging from 3.08 to 5.63 (Figure 8d). The enhanced sensitivity of the MCMs/gel originates from the uniform distribution of microspheres, which ensures consistent modulation of conductive pathways during deformation.

Figure 9a demonstrates that the hydrogel exhibits rapid response (210 ms) and recovery (202 ms) times under instantaneous strain, confirming its suitability for real‐time monitoring. Stable sensing sensitivity was evidenced by the overlapping ΔR/R₀ curves obtained at different stretching rates (Figure 9b). Long‐term cyclic testing (at 80% strain for 1,000 cycles) showed minimal signal drift (Figure 9c), indicating the hydrogel's exceptional durability. Stepwise strain testing (with 50% increments up to 500% strain) revealed precise resistance gradation and recovery (Figure 9d). Notably, the strain steps maintained near‐90° angles, confirming the rapid response kinetics of hydrogel.

FIGURE 9.

Sensing performance of MCMs/gel. (a) Change in relative resistance during loading‐unloading process of hydrogel. (b) Change in relative resistance under different rates of strain. (c) Durability testing during cycling at 80% maximum strain of hydrogel. (d) Change in relative resistance as strain of hydrogel varies in steps. (e, f) Change in relative resistance of hydrogel during repeated stretching processes at low strain (1%–9%) and high strain (100%–500%).

Figure 9e–f illustrate consistent ΔR/R₀ response patterns across both small (1%–9%) and large (100%–600%) strain ranges. Identical strain magnitudes produced reproducible signal intensities, while varying strains induced proportional resistance changes. This validates the hydrogel's broad‐range strain discriminability (1%–600%) with high stability. The hydrogel's combined mechanical robustness and sensing precision enable real‐time monitoring of human biomechanical activities.

As shown in Figure 10, the hydrogel can detect both joint flexion (finger, wrist, elbow, and knee) and subtle physiological motions (blinking, swallowing, and phonation) via strain‐induced resistance changes [46, 47]. When attached to the finger (Figure 10a), it generated distinct resistance signals corresponding to 30°, 60°, and 90° flexion angles; the same principle was applied to wrist, elbow, and knee movements, enabling precise tracking of joint motion. When placed subdermally near the ocular and laryngeal regions, the hydrogel captured microstrains associated with blinking, swallowing, and vocal cord vibrations, confirming its capability for multimodal physiological signal detection [48, 49].

FIGURE 10.

Real‐time monitoring of human activities using MCMs/gel: (a) Finger flexion at 30°, 60°, and 90°. (b) Finger bending. (c) Wrist flexion. (d) Elbow flexion. (e) Knee flexion. (f) Blinking. (g) Swallowing. (h) Phonation.

3.7. In Vitro Drug Release

Figure 11a,b present the cumulative release profiles of TH‐loaded MCMs, Bbh‐loaded MCMs, and their corresponding hydrogel composites (MCMs/gel). The results demonstrate that both the cumulative release rate and release rate of the drugs were higher in standalone MCMs than in MCMs/gel, confirming the sustained‐release effect of the hydrogel matrix. Further analysis revealed that the interconnected 3D network structure of the PAA/PVA hydrogel effectively suppressed the initial burst release of drugs by constructing a multi‐stage diffusion barrier. Specifically, drug molecules first diffuse out from the interior of MCMs into the porous hydrogel framework, and then penetrate the hydrogel surface to migrate into the external PBS environment. This two‐step diffusion pathway significantly prolongs the drug release cycle and realizes long‐term drug delivery.

FIGURE 11.

Cumulative release of TH (a) and Bbh (b) from MCMs and MCMs/gel over 7 d. TH/MCMs/gel drug release at different pH levels. Values reported are an average of n = 3, ± standard deviation.

A comparative analysis of Figure 11a,b highlights distinct release behaviors and release efficiency differences between TH and Bbh. As a highly water‐soluble antibacterial drug, TH possesses excellent molecular diffusivity, enabling it to rapidly and uniformly diffuse through the hydrophilic hydrogel network into the surrounding medium. In contrast, Bbh exhibits relatively poor aqueous solubility and strong molecular hydrophobicity, which is the core reason for its lower loading capacity and cumulative release efficiency. On the one hand, the hydrophobic interaction between Bbh molecules and the chitosan microsphere framework enhances the intermolecular binding force, hindering the desorption and outward diffusion of Bbh. On the other hand, the aggregation tendency of partial Bbh molecules in aqueous environments further reduces the effective concentration of diffusible drug molecules. In addition, compared with the small‐molecule TH, Bbh has a larger molecular steric hindrance, which further limits its penetration efficiency through the microsphere pores and hydrogel network, ultimately leading to a significantly lower cumulative release amount than TH. Notably, the relatively low release rate of Bbh is conducive to avoiding excessive short‐term drug leakage, which complements the rapid release characteristics of TH and achieves a dual‐drug synergistic sustained‐release effect.

Figure 11c examines the pH‐dependent drug release behavior of MCMs/gel. Under acidic conditions (pH 5.5), the protonation of amino groups on chitosan loosens the microsphere cross‐linking structure and enlarges internal pore channels, thereby accelerating the outward diffusion and release of loaded TH, resulting in a cumulative release rate of 53.03%. Conversely, in neutral and alkaline environments (pH 7.4), the reduced protonation degree of chitosan amino groups maintains the compact and stable structure of microspheres, effectively slowing the drug diffusion rate and prolonging the sustained release duration. This intelligent pH‐responsive release behavior endows the composite hydrogel with superior microenvironment adaptability for wound therapy: rapid drug release in acidic, infection‐prone wound microenvironments can quickly inhibit acute bacterial proliferation and relieve inflammation, while stable and sustained drug release under neutral/alkaline conditions provides long‐term antibacterial protection for chronic wound repair, demonstrating great application potential in precise wound treatment [42].

3.8. Antibacterial Performance

The antibacterial efficacy of plain hydrogels, drug‐loaded hydrogels, and composite hydrogels (MCMs/gel) was evaluated against S. aureus and E. coli, with drug‐free hydrogels serving as controls. As shown in Figure 12, the control hydrogels (without drugs) produced modest inhibition zones, indicating inherent antibacterial properties of the hydrogel matrix itself. In contrast, the drug‐loaded hydrogels exhibited significantly larger inhibition zones, confirming enhanced antibacterial activity. Notably, both TH and Bbh demonstrated superior inhibitory effects against S. aureus compared to E. coli. This discrepancy stems from structural differences in bacterial cell walls. The thick peptidoglycan layer of Gram‐positive S. aureus may facilitate drug penetration, whereas the lipopolysaccharide‐rich outer membrane of Gram‐negative E. coli can hinder drug entry.

FIGURE 12.

Photograph of the circle of inhibition of hydrogel against S. aureus and E. coli.

Interestingly, the drug‐loaded hydrogels (with direct drug incorporation) showed smaller inhibition zones than the composite hydrogels (MCMs/gel) (Figure 13a). This suggests potential drug‐polymer interactions in the direct‐loaded system that compromise drug efficacy. In contrast, microsphere encapsulation in the composite hydrogels preserved drug activity, enabling optimal antibacterial performance. Figure 13b,c demonstrate that the hydrogels achieved >96% bactericidal efficiency against both bacterial strains within 6 h of contact, confirming their rapid and potent antibacterial action. Figure 13d compares the optical density (OD) values of bacterial suspensions exposed to different hydrogels. All hydrogel groups significantly reduced bacterial growth (evidenced by lower OD values) compared to the controls. However, the composite hydrogels (MCMs/gel) outperformed the direct drug‐loaded hydrogels, as indicated by their lower OD values, validating the superiority of the microsphere encapsulation strategy [50, 51, 52, 53].

FIGURE 13.

(a) Photograph of S. aureus and E. coli in contact with hydrogel. Bactericidal efficiency of hydrogel against S. aureus (b) and E. coli (c). (d) OD values of S. aureus and E. coli. Values reported are an average of n = 3, ± standard deviation. ^∗∗∗ p < 0.005.

4. Conclusions

In summary, this study successfully fabricated a dual‐crosslinked hydrogel via self‐catalyzed free radical polymerization, with borax employed as the reinforcing phase. The composite hydrogel containing 4‰ (w/v) borax and 20 mg/mL magnetic chitosan microspheres (MCMs) exhibited optimal comprehensive performance, including superior swelling capacity, robust mechanical properties and excellent strain‐sensitive responsiveness. Notably, the hydrogel demonstrated significant antibacterial efficacy against both E. coli and S. aureus, achieving a bactericidal rate of >96% within 6 h of contact. It also exhibited pH‐modulated drug release kinetics, with a cumulative release of 53.03% under acidic conditions (pH 5.5), a feature that enables targeted drug delivery in infection‐prone wound environments. These integrated attributes position the composite hydrogel as a promising multifunctional platform for wound dressing applications, capable of simultaneously realizing drug delivery, bacterial inhibition, and real‐time wound micro‐environment monitoring. Furthermore, this work provides a scalable fabrication strategy for developing next‐generation smart wound care systems, laying a foundation for their potential translation to clinical practice.

Author Contributions

Linhan Hu: conceptualization, methodology, formal analysis, writing – original draft. Zheng Yang: conceptualization, formal analysis. Xinwei Tao: validation, formal analysis. Bowen Ren: investigation, resources. Jianliang Li: investigation, data curation. Tianle Zhou: resources, investigation, validation, methodology. Huaping Tan: writing – review & editing, supervision, project administration, funding acquisition.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.