Mechanically robust PVA/SA semi-IPN hydrogels for highly effective temperature-triggered linagliptin delivery

Abstract

Hydrogels with excellent mechanical properties and high drug-delivery capacity are highly advantageous for wound healing applications. However, reconciling mechanical properties with drug-delivery performance remains challenging. Herein, a semi-interpenetrating polymer network (semi-IPN) hydrogel composed of poly(vinyl alcohol) (PVA) and sodium alginate (SA) was fabricated via freeze–thaw cycling. Through component optimization, this hydrogel achieved a unique combination of desirable properties, including softness (Young's modulus of 32.5 kPa), high stretchability (fracture strain of 283.9%), and high water content (96.46%). Linagliptin (LIN), a DPP-4 inhibitor crucial for diabetic wound repair, was loaded into the system with an effective loading efficiency of 89.25%. The hydrogel system demonstrated temperature-responsive release kinetics: minimal release occurred at low temperatures, while rapid, sustained release was achieved at physiological temperature, reaching a cumulative release efficiency of up to 85.17%, which was highly beneficial for the storage and application of drug-loaded hydrogels. This study presents a hydrogel platform with effectively integrated material robustness, temperature-triggered drug delivery and high-efficiency loading and release of LIN, which showcases significant potential as a novel therapeutic material for diabetic wound healing.

---

Introduction

Chronic wounds, particularly in diabetic patients, are exacerbated by inflammation, vascular dysfunction, and impaired healing, often leading to severe complications such as amputations.^3–5 Clinical treatment methods usually rely mainly on drug therapy. However, improper dosage of the drugs can often lead to serious complications.⁸ Therefore, it is necessary to develop a controllable drug administration method to eliminate the harm caused by excessive drug administration. Hydrogels are polymer materials with a three-dimensional network structure, which is very easy to be endowed with excellent mechanical, optical, and electrical properties, as well as the ability to be tailored across multiple scales.¹² Due to their versatility, hydrogels are widely used in various fields such as bioelectronics,¹⁵ water purification,^16–20 strain sensors,^21–23 electrochromic devices,²⁴ and soft robotics.^22,25 Moreover, hydrogels are promising drug delivery systems because of their high-water content, excellent biocompatibility, tunable physicochemical properties, and effective drug-loading capacity, which together enhance therapeutic outcomes.^26–28 They enable sustained and controlled drug release, improving drug utilization, while also offering high flexibility and environmental adaptability to support stable drug delivery. In wound healing applications, hydrogels help maintain a moist environment that promotes cell migration and proliferation, accelerating tissue repair.^29,30 Nevertheless, reconciling robust mechanical properties with efficient drug loading and controlled release persists as a critical challenge.

In recent years, microenvironment-responsive drug-release hydrogels, such as those responsive to reactive oxygen species (ROS), pH variations, and specific enzymes, have been widely developed for application in intelligent drug delivery systems.^31–34 However, the complexity of drug release mechanisms, the dynamic activity of the microenvironment, variability in substrate concentration, and the presence of false positive signals often results in incomplete or uncontrolled (burst) drug release, thereby prolonging the healing process of chronic wounds.^35–37 Furthermore, these responsive drug delivery systems are susceptible to oxidation, degradation, and inactivation during storage, leading to unnecessary drug waste and economic losses. Developing a storage-stable and highly efficient drug delivery system for wound treatment in humans remains a significant challenge.

Polyvinyl alcohol (PVA) is a polar polymer characterized by excellent water solubility, inherent non-toxicity, and favorable mechanical properties. Through chemical or physical crosslinking, PVA can form hydrogels that are extensively utilized as matrix materials in wound dressing applications.^19,38–40 Sodium alginate (SA), a naturally derived polyanionic, linear carbohydrate biopolymer extracted from seaweed, exhibits excellent biocompatibility, non-toxicity, and strong hydrophilic properties.^41,42 Both PVA and SA are considered promising materials for hydrogel design. PVA contributes structural integrity through physical crosslinking, whereas SA enhances biocompatibility and hydrophilicity. The integration of PVA and SA results in the formation of a semi-interpenetrating network (semi-IPN) structure, which facilitates efficient drug encapsulation and enables temperature-responsive drug release.⁴³

In addition, linagliptin (LIN), as a DPP-4 inhibitor, is widely used in the treatment of type 2 diabetes, where it works by enhancing insulin secretion and lowering blood glucose levels.^44–47 LIN exerts its pharmacological effects while exhibiting absolute safety.^6,9,48 However, LIN faces several challenges during treatment as an oral drug, such as incomplete absorption in the gastrointestinal tract, which can make it difficult to achieve adequate drug concentrations.^49,50 Furthermore, traditional oral administration methods struggle to provide sustained and efficient drug release at the local site of diabetic wounds. Therefore, applying LIN in wound dressings for local drug delivery can more directly promote wound healing while avoiding systemic side effects.^46,49–51

Therefore, in this study, we developed a three-dimensional semi-IPN SA-PVA hydrogel through a simple freeze-thawing cycle approach, following the regulation and optimization of the solid content of PVA. Among the prepared hydrogels, the one with a solid content of 90.9 wt% PVA (SA–PVA_90.9) exhibited the best overall performance, characterized by superior mechanical properties (maximum stress: 84.5 kPa; maximum strain: 283.9%), a high water content of 96.46%, and an ideal swelling rate of 2275%. When immersed in a LIN solution, the optimized PVA–SA_90.9% hydrogel demonstrated an exceptionally high drug-loading capacity. Notably, the drug-loaded hydrogel exhibited minimal release at 18 °C, while showing rapid, sustained release at human body temperature (37 °C), indicating excellent storage stability and temperature responsiveness. The aim of this study was to investigate the influence of different concentrations of PVA and SA on the structural and functional properties of the hydrogel, aiming to provide a novel material support for the treatment of diabetic wounds.

Materials and methods

Materials

Polyvinyl alcohol (PVA-124, 99.0% alcoholysis, Mr ≈ 195 000), sodium alginate (SA), and dimethyl sulphoxide (DMSO, 99%) were purchased from Aladdin (Shanghai, China). Phosphate buffered saline (PBS, pH = 7.4) was purchased from Solarbio (Beijing, China). Linagliptin (LIN, ≥99%) was purchased from Shanghai Sekerui Biotechnology Co., Ltd. Acidic electrolyzed oxidizing water (EOW) was purchased from Panike (Shenzhen, China) and used as received. All other chemicals were of analytical grade and used without additional purification.

Preparation of SA–PVA composite hydrogels

4 g of SA powder was added to 196 mL of deionized water and stirred until completely dissolved at room temperature to obtain a 2 wt% SA solution. PVA-124 powder was dissolved in deionized water at 90 °C to prepare PVA solutions of different mass concentrations (10, 12.5, 15, 17.5, and 20 wt%). Different concentrations of PVA solution (20 g) were mixed with 2 wt% SA solution (20 g) in a 1 : 1 mass ratio by vortex mixing. After degassing, sticky SA–PVA mixtures with different PVA concentrations were obtained, namely 83.3, 86.2, 88.2, 89.7, and 91.9 wt%. Finally, the mixtures were poured into molds and subjected to three cycles of freezing and thawing (i.e., frozen at −20 °C for 8 h and thawed at 25 °C for 4 h) to form mechanically stable SA–PVA.

Preparation of the SA–PVA@LIN system

The LIN was dissolved in dimethyl sulfoxide (DMSO) (10 mg mL⁻¹) to prepare a storage solution. The SA–PVA composite hydrogel solution was prepared as described above, and LIN was then added to the solution at concentrations of 50, 100, 200 and 400 µg mL⁻¹. After thorough mixing, the SA–PVA@GLIN was obtained through three freeze–thaw cycles. In addition, the SA–PVA@SLIN hydrogel system was developed by equilibrating PVA–SA hydrogels in a LIN solution. The drug loading hydrogel obtained through the soaking method was referred to as SA–PVA@SLIN. It should be noted that during the initial stage of mechanical property research, since the mechanism of drug delivery was not being investigated, we uniformly prepared the hydrogel dressings using the immersion method.

Characterization

Scanning electron microscopy (SEM). The optimized SA–PVA_90.9 and SA–PVA_90.9@LIN₁₀₀ samples were rapidly frozen using liquid nitrogen and then freeze-dried for 24 h. The morphology was observed and photographed by scanning electron microscopy (SEM, JSM-7500F, JEOL) to investigate the microscopic surface and cross-sectional morphological features, such as porous channels and differences in porosity.

Fourier transform infrared (FT-IR) spectroscopy. The freeze-dried samples were ground with potassium bromide (KBr) and pressed into pellets to prepare test samples. FT-IR spectra of PVA, SA, SA–PVA_90.9, and optimized SA–PVA_90.9@LIN₁₀₀ were tested in the wavelength range of 4000–500 cm⁻¹ by employing an FT-IR spectrometer (Spectrum Two, PerkinElmer). The spectra were fitted and analyzed to identify the corresponding functional groups, thereby confirming the presence of certain substances.

Water content

The initial weight of SA–PVA composite hydrogels with different solid contents was calculated, and then they were dried in a vacuum freeze dryer for 24 h before weighing again. Three measurements for each sample were carried out and the average result was taken. The water content (W_c) of the samples was calculated using formula (S1).

Swelling test

The swelling behavior was evaluated by immersing the hydrogel in excess PBS solution (pH = 2.8, 6.2, 7.0, 8.2, 10.8). The freeze-dried hydrogels were soaked in a large volume of PBS solution, and swollen samples were taken out at fixed time intervals until swelling equilibrium was reached. Each sample was tested three times, and the average value was taken. The swelling ratio (ESR) of the hydrogel was calculated using formula (S2).

Mechanical test

To address the propensity of hydrogels to dehydrate and subsequently undergo structural collapse and alterations in mechanical properties, an underwater tensile testing apparatus (ZQ-990L with 5 N load cell, Zhiqu Precision Instrument) was utilized to evaluate the mechanical properties of dumbbell-shaped SA–PVA specimens. The width of all dumbbell-shaped hydrogels was 2 mm, the thickness was 2.5 mm, and the gauge length was 20 mm. The stretching rate was set to be 100 mm min⁻¹, and the samples were stretched until fracture to obtain the stress–strain curves. Each sample was tested 10 times, and the average value was taken. The Young's modulus was calculated from the slope of the curve in the strain range of 5–15%. The toughness was calculated from the area under the stress–strain curve. The elongation at break was determined from the strain value. Additionally, a total of 1000 stretching cycles were performed on the 90.9 wt% SA–PVA (SA–PVA_90.9) in the 0–100% strain range.

Drug release

Gel method: SA–PVA@GLIN with different solid contents and drug concentrations was fixed between two mesh disks of an intelligent dissolution tester (ZRS-8GD, Tianda Tianfa). Then, 500 mL of dissolution medium (PBS, pH = 7.4) was added, and the temperature was maintained at 18 ± 0.5 °C, 25 ± 0.5 °C and 37 ± 0.5 °C with a constant rotation speed of 100 rpm. Samples were taken at fixed intervals (0.5, 1, 2, 3, 6, 12, 24, 36, 48, 60 h), and 5 mL of dissolution solution was collected from each dissolution cup, to which an equal volume of PBS solution is added. The absorbance of all samples was then measured at 293 nm. The drug release at each time point was calculated based on the standard curve, and the drug release curve was plotted. The drug release efficiency (DRE) was calculated from the concentration of the drug in SA–PVA@LIN and the amount of drug release by formula (S3).

Soaking method: SA–PVA with different solid contents of PVA (83.3, 86.2, 88.2, 89.7, 90.9 wt%) was soaked in 20 mL of a standard LIN solution (50, 100, 200 and 400 µg mL⁻¹) for 24 h to verify the drug loading efficiency of the hydrogel. SA–PVA@SLIN was suspended in PBS at 37 °C for 24 h to release LIN. The releasing solution was diluted to 20 mL with deionized water, and the absorbance of the sample was measured at 293 nm. Based on the absorbance–concentration calibration curve, the absorbance was converted into concentration, and the drug loading efficiency (DLE) was calculated using formula (S4).

Statistical analyses

Origin 2024 software was used for statistical analysis, and the results were expressed as mean ± standard deviation (SD), and the results of significant differences were as follows: P-values were calculated using Student's t-test; *P < 0.05, **P < 0.01, ***P < 0.001.

Results and discussion

Design and fabrication of SA–PVA

A physically crosslinked network is constructed via a freeze–thaw cycle method, as illustrated in Fig. 1. Between SA and PVA, a loose hydrogel structure is formed through hydrogen bonding. When this hydrogel is frozen, the water within it will form ice crystals, which promotes the aggregation of SA and PVA chains and the formation of more hydrogen bonds to stabilize the gel system.^52,53 Upon thawing, the melted ice crystals release water, leaving a stable three-dimensional framework maintained by hydrogen bonds and physical entanglements. With repeated freeze–thaw cycles, the PVA and SA chains further aggregate and re-arrange to form the gel system with a more stable structure. Finally, the resulting hydrogel, rich in hydroxyl groups, exhibits excellent water absorption and swelling properties.^30,54

Fig. 1 Preparation and drug release process of SA–PVA@(G/S)LIN drug-loaded hydrogels.

Furthermore, the linear SA, rich in a large amount of carboxyl groups, forms hydrogen bonds with the hydroxyl groups in PVA, and simultaneously penetrates through the network voids of PVA to form a physical entanglement structure, thereby forming a semi-IPN structure.^41,55,56 By adjusting the density of the crosslinking points and the number of hydrogen bonds in the semi-IPN, macroscopic control of the mechanical properties can be achieved.^43,57–59 In addition, LIN is a water-soluble molecule, which can easily be encapsulated within the three-dimensional hydrogel system. LIN is continuously released from this three-dimensional network via diffusion. By providing an appropriate dose of linagliptin over time, it may directly or indirectly induce keratinocytes to undergo epithelial–mesenchymal transition, thereby promoting keratinocyte migration and improving diabetic wound healing.^51,60

Characterization of SA–PVA

Fig. 2a shows the SEM images of the surfaces of the SA–PVA and SA–PVA@SLIN₁₀₀ hydrogels with different PVA formulations. SA–PVA_90.9 features a distinct porous structure with abundant continuous pores, forming a semi-interpenetrating polymer network (semi-IPN) architecture. In contrast, the SA–PVA_90.9@SLIN₁₀₀ drug-loaded hydrogels possess a highly porous three-dimensional network with more densely packed pores. Pore size and morphology vary with increasing PVA content, leading to a structural transition from a compact to a more porous configuration. This phenomenon can be attributed to the formation of the semi-IPN structure, which allows both components to retain their intrinsic chemical structures. Specifically, SA is not chemically crosslinked but physically entangled within the PVA network – SA chains diffuse into the PVA matrix as linear chains to form the semi-IPN structure.⁵⁵ Furthermore, SA may interfere with the crosslinking sites between the PVA molecular chains, thereby restricting the effective physical crosslinking scope of PVA.^61,62

Fig. 2 Characterization of the SA–PVA@SLIN properties and water content. (a) The SEM of SA–PVA and SA–PVA@SLIN₁₀₀ with different PVA contents on a scale of 10 µm. (b) FT-IR spectra of PVA, SA, SA–PVA_90.9 and SA–PVA_90.9@SLIN₁₀₀. (c) FT-IR spectra and (d) water contents of SA–PVA with varying PVA content. (e) Swelling ratios of SA–PVA within 0–24 h. N = 3.

In Fig. 2b, all the characteristic peaks of PVA and SA are observed. For instance, the peak at 3417 cm⁻¹ is attributed to the O–H stretching vibration of the hydroxyl group in PVA, the peak at 1628 cm⁻¹ corresponds to the C=O stretching vibration of SA, the peak at 1417 cm⁻¹ is associated with the bending vibration of O–H, and the peak at 1087 cm⁻¹ is attributed to the C–O stretching vibration. FT-IR analyses are performed on PVA, SA, SA–PVA_90.9, and SA–PVA_90.9@SLIN₁₀₀ to investigate the interactions among the constituent components. The intensities of the O–H and C–O peaks primarily reflect the extent of hydrogen bonding; when PVA is incorporated into SA, no significant shift in peak positions is observed, suggesting the absence of hydrogen bond formation. However, the slight shift in the O–H peak from 3431 cm⁻¹ to 3437 cm⁻¹, along with the shift in the C–O peak from 1087 cm⁻¹ to 1089 cm⁻¹, indicates the formation of a synergistic hydrogen bonding system between SA and PVA. As the PVA content increases in the hydrogel, the intensities of the peaks at 1628 cm⁻¹ and 1087 cm⁻¹ decrease, which is attributed to the reduction in the number of free C=O groups (Fig. 2c).

Water content and swelling ratio of SA–PVA

Regardless of the PVA content, the water content of all SA–PVA hydrogels remains around 90% (Fig. 2d). This high water content is crucial for biomedical applications, particularly in wound care and tissue engineering, as it helps create a moist environment, promotes tissue healing, and aids in drug delivery.^63–65 Additionally, the swelling rate of different types of SA–PVA composite hydrogels is demonstrated in Fig. 2e. During the initial swelling stage, the hydrogel's swelling rate increases rapidly as water molecules penetrate the network and interact with internal macromolecules. This is followed by a slower rise in swelling degree, which stabilizes by approximately 12 h. For SA–PVA with PVA contents of 83.3 wt%, 86.2 wt%, 88.2 wt%, 89.7 wt%, and 90.9 wt%, the maximum swelling degrees are approximately 1792%, 2055%, 2275%, 1728%, and 1476%, respectively, showcasing that the swelling rate first increases and then decreases upon increasing PVA content, which is mainly due to the hydroxyl groups in the PVA chains. The hydrogen bonds forming between hydroxyl groups and H₂O endow the hydrogel with strong water-absorption capacity. However, higher PVA content reduces the hydrogel's inherent water content, limiting the enhancement of osmotic pressure. This, in turn, restricts the hydrogel's ability to absorb water molecules, leading to a subsequent decrease in swelling rate.^66,67

In addition, the swelling behavior of SA–PVA is also dictated by the ionization state of the alginate fraction. In neutral to weakly alkaline environments (pH = 7–8), carboxylate moieties (–COO⁻) along the alginate backbone undergo deprotonation, inducing robust electrostatic repulsion that drives polymer network expansion and yields maximal swelling capacity.^68,69 Under acidic conditions (pH = 2.8–6), these moieties are protonated to form carboxyl groups (–COOH), abrogating electrostatic repulsion while facilitating intermolecular hydrogen bonding, thereby inducing polymer network shrinkage. In strongly alkaline solutions (pH = 9–10.8), the high ionic strength may screen electrostatic repulsions, and excess ions can cause an “anti-polyelectrolyte” effect, also resulting in reduced swelling (Fig. S1).

Mechanical properties

Hydrogel dressings for biomedical applications are required to possess excellent stretchability, flexibility, and structural stability. The designed hydrogels should facilitate drug release regulation, promote wound healing, enhance patient comfort, and be biocompatible with human skin. To evaluate the mechanical properties of SA–PVA, tensile and cyclic tensile tests were performed. Comparative analysis of SA–PVA with different component ratios reveals that as the PVA solid content increases from 83.3 wt% to 90.9 wt%, the maximum stress (27.6–86.1 kPa), elongation at break (102.0–253.9%), and toughness (18.9–122.3 kJ m⁻³) all show a continuous increasing trend (Fig. 3a–c). In contrast, the Young's modulus (32.4–38.8 kPa) first decreases and then increases with rising PVA content (Fig. 3d). Notably, this combination of superior tensile performance and tunable flexibility is highly favorable for maintaining conformability and compatibility between the hydrogel dressing and human skin (typical skin strain: 6–60%; ideal Young's modulus range for dressings: 5–2600 kPa). These mechanical characteristics thus establish a robust foundation for the biomedical application of the SA–PVA dressing (Fig. S1).^70,71 This phenomenon may be attributed to the significant increase in the network density of PVA as the monomer content increases, thereby ensuring the stability of the physical network during stretching. Moreover, at a strain of 100%, continuous stress–strain cyclic stretching demonstrates that the 90.9 wt% SA–PVA has good mechanical stability (Fig. 3e), maintaining 83% stability in maximum stress (46.6–38.6 kPa) after 1000 cyclic stretches (Fig. 3f). Dynamic mechanical analysis (DMA) confirmed the viscoelastic properties of the hydrogel, demonstrating its solid-like behavior (G′ > G″) and mechanical stability (Fig. S2). In conclusion, SA–PVA hydrogels not only exhibit excellent mechanical properties in terms of stretchability in coordination with body movements, but also possess effective drug release capabilities, making them a promising solution with significant potential for enhancing the comfort and effectiveness of diabetic wound healing treatments.

Fig. 3 The mechanical properties of SA–PVA. (a) Tensile stress–strain curves of SA–PVA with various PVA contents. (b) Elongation at break of SA–PVA, N = 3. (c) Toughness and (d) Young's modulus of SA–PVA with various PVA contents, N = 3. (e) Cyclic tensile stress–strain curves of SA–PVA_90.9 under the strain of 100%. (f) Stress variation of SA–PVA_90.9 at the 1st, 2nd, 10th, 100th, and 1000th cycles.

Drug release analysis

The concentration of LIN solution is plotted on the X-axis, and the UV absorption intensity is plotted on the Y-axis. The fitted absorbance–concentration standard curve is manifested in Fig. S4a, with the equation: y = 0.0094x + 0.021, R² = 0.9993.

PVA hydrogels exhibit a sparsely physically cross-linked network, enabling drugs to diffuse readily through the pores. Additionally, their strong hydrophilic nature induces significant network swelling upon water absorption, expanding the pore structure and further enhancing drug diffusion. Therefore, PVA-based hydrogels generally exhibit high drug release efficiency; however, since LIN is a non-hydrophilic drug, its release efficiency is relatively low (Table 1). The SA–PVA@GLIN and SA–PVA@SLIN systems are quantitatively assessed for drug release over time, along with the drug loading capacity of SA–PVA@SLIN. Under a drug concentration of 100 µg mL⁻¹, as the solid content of PVA in the hydrogel increases, the drug loading efficiency initially increases and then decreases (78.31–83.34%) (Fig. 4a and Table S1). The trend is similar to the principle governing the swelling rate. However, the drug release performance first decreases and then increases with increasing PVA content, meaning that the drug release amount (63.32–82.13 µg mL⁻¹) and drug release efficiency (72.46–85.17%) both exhibit a trend of first decreasing and then increasing (Fig. 4b–f and Tables S1 and S2). This may be because when PVA and SA form a semi-IPN structure, the increase in PVA leads to more crosslinking points in the network, which restricts the free movement of polymer chains, thus limiting the drug release to some extent.^43,59,72 However, when the water content of the hydrogel is relatively low, it generates a stronger osmotic pressure, further promoting drug release, which results in an increased drug release amount as the water content decreases in SA–PVA with 89.7 wt% and 90.9 wt% PVA. By comparing the drug release efficiency of SA–PVA@GLIN₁₀₀ and SA–PVA@SLIN₁₀₀ at different PVA contents, it is found that SA–PVA@SLIN₁₀₀ has a significantly higher release efficiency than SA–PVA@GLIN₁₀₀ (Fig. 4e and Tables S1 and S2). LIN is a poorly water-soluble drug and tends to form a gel easily when mixed with the SA–PVA precursor solution. Therefore, SA–PVA@SLIN₁₀₀ provides more flexibility in drug loading and release.

Table 1 Comparison of drug-loading efficiency, drug release efficiency, and release time of the PVA-based drug-loaded hydrogels

Loaded drugs	Drug loading efficiency (%)	Drug release efficiency (%)	Release time (h)	Ref.
Diclofenac sodium	—	95.96	8	2
Diflunisal	79	95	30	7
Diphenhydramine HCl	4.23	93.96	12	10
Theophylline	—	80.73	336	11 and 76
Curcumin	—	89.56	48	14
SA–PVA@SLIN	89.25	85.17	24	This work

---

Fig. 4 The drug release performance of SA–PVA@(S/G)LIN₁₀₀ at different PVA solid contents. (a) Drug loading efficiency of SA–PVA@SLIN₁₀₀ at different solid contents of PVA. (b) Cumulative LIN release of different contents of SA–PVA@SLIN₁₀₀ within 0–24 h. (c) Drug release efficiency of SA–PVA@SLIN₁₀₀ at different temperatures. (d) The cumulative release of LIN from various SA–PVA@GLIN₁₀₀ over a period of 0–24 h. (e) Drug release efficiency of SA–PVA@SLIN₁₀₀ and SA–PVA@GLIN₁₀₀ at different solid contents. (f) Drug release efficiency of SA–PVA@GLIN₁₀₀ at various temperatures. N = 3, P-values were calculated using Student's t-test; *P < 0.05, **P < 0.01, ***P < 0.001.

When the hydrogel solid content was fixed, the water content remained constant. As the drug concentration increases, both the drug loading efficiency (52.42–89.25%) and the drug release amount (22.91–298.65 µg mL⁻¹) exhibited an increasing trend (Fig. 5a, b and d and Tables S3 and S4). A concentration of 400 µg mL⁻¹ demonstrated the highest drug loading efficiency (Fig. 5a and Table S3), with rapid release of LIN in the first 12 hours, followed by a slower, sustained release phase that maintains a stable concentration within 24 hours (Fig. 5b and d and Table S4). The release mechanism was governed by concentration gradient-driven diffusion, enabling prolonged LIN release at the target site.⁷³ Similarly, when comparing the drug release efficiency of SA–PVA_90.9@SLIN and SA–PVA_90.9@GLIN at different LIN concentrations, SA–PVA_90.9@SLIN consistently showed significantly higher release efficiency than SA–PVA_90.9@GLIN; however, both exhibited a decreasing trend with increasing concentration (Fig. 5e and Table S4). Generally, higher initial drug concentrations result in greater drug loading within the hydrogel. Consequently, higher drug concentrations might enhance the loading capacity, leading to improved early-stage release efficiency. However, at elevated concentrations, intermolecular interactions (e.g., hydrogen bonding, van der Waals forces) might intensify, and drug molecules might occupy hydrogel pores, reducing free volume and subsequently decreasing the diffusion rate of the drug. Collectively, these findings indicated that SA–PVA demonstrated excellent drug loading and release performance for LIN (Table 2). Ultimately, based on the optimization of the mechanical properties, drug loading and drug release capabilities, SA–PVA_90.9@SLIN₁₀₀ is deemed to be the most promising system, which not only has good skin compatibility but also demonstrates efficient drug delivery capabilities, and shows great potential for application in hydrogel wound healing.

Fig. 5 The drug release performance of SA–PVA_90.9@(S/G)LIN₁₀₀ loading different LIN concentrations. (a) Drug loading efficiency of SA–PVA_90.9@SLIN. (b) The cumulative release amount of SA–PVA_90.9@SLIN within 0–24 h. (c) Drug release efficiency of SA–PVA_90.9@SLIN₁₀₀ at various temperatures. (d) The cumulative quantity of SA–PVA_90.9@GLIN₁₀₀ released between 0 and 24 h. (e) Efficiency of drug release from SA–PVA_90.9@SLIN and SA–PVA_90.9@GLIN at various LIN concentrations. (f) Drug release efficiency of SA–PVA_90.9@GLIN at different temperatures. N = 3, P-values were calculated using Student's t-test; *P < 0.05, **P < 0.01, ***P < 0.001.

Table 2 Comparison of the bioavailability, drug release efficiency and release time of the different hydrogel dressings loaded with LIN

Administration methods	Drug release efficiency (%)	Release time (d)	Ref.
LNG-HA	43.56	2	1
LIN@PG10@PDA	55.19	0.58	6
SA@LIN/BMS-202	55.7	2	9
AG-CD@LIN	78.6	0.58	13
SA–PVA@SLIN 100	85.17	1	This work

---

It is particularly noteworthy that both SA–PVA_90.9@SLIN₁₀₀ and SA–PVA_90.9@GLIN₁₀₀ consistently exhibit strong thermosensitivity, that is the drug-release ability is controllable at different temperatures. When the temperature rises to 37 °C, LIN in PVA_90.9@SLIN₁₀₀ releases rapidly with a high efficiency of up to 85.17%. However, when the temperature drops to 18 °C, the network of the hydrogel successfully returned to the contracted and low-release state, with the drug release efficiency being less than 20% (Fig. 4c, f and 5c, f and Fig. S5–S8). This is because at lower temperatures, the SA–PVA_90.9 network contracts, narrowing the diffusion channels for LIN molecules and physically entrapping the drug within the network. However, at elevated temperatures, phase separation occurs throughout the gel, generating larger micropores that offer lower diffusion resistance.^72,74,75 Meanwhile, the cyclic heating–cooling experiments confirmed its robust reversibility without damaging the structure or function of the hydrogel (Fig. S4b and c).

Conclusions

A semi-IPN hydrogel composed of PVA and SA was successfully fabricated through freeze–thaw cycling. The optimized hydrogel of SA–PVA_90.8 exhibited remarkable mechanical properties (tensile strength: 84.5 kPa, fracture strain: 283.9%, Young's modulus: 32.5 kPa) and high-water content (96.46%), fulfilling the requirements for flexible wound dressings. Efficient LIN loading (89.25%) and thermally triggered release (85.17% cumulative efficiency at 37 °C) demonstrated its capability for intelligent drug-delivery ability. This system overcomes the limitations of oral administration, enabling sustained local release at the wound site under normal human body temperature and allowing low-temperature preservation. The combination of mechanical robustness and stimuli-responsive functionality positions the PVA/SA hydrogel as a promising platform for diabetic wound therapy.

Author contributions

Y. W.: writing – conceptualization, data curation, software, formal analysis, and original draft. G. L., L. P., Y. L., M. Y. and Q. W.: data curation and formal analysis. L. P. and H. G.: resources, conceptualization, supervision, data curation, writing – review and editing, and funding acquisition. All authors have read and agreed to the published version of the manuscript.

Conflicts of interest

The authors declare no conflicts of interest.

Data availability

All data generated or analyzed during this study are included in this published article.

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information: Fig. S1. The swelling rate of hydrogels at different pH values. Fig. S2. Dynamic frequency-sweep measurements of SA–PVA. Fig. S3. Adhesion performance and compression performance of SA–PVA. (a) and (b) Photo showing adhesive effects of SA–PVA adhering to various substrates (heart, liver, lung, kidney, spleen, rubber, glass, PI, PET and Ti); (c) compression performance of SA–PVA; (d) applications of hydrogels in different skin regions. Fig. S4. Standard curve of LIN. Fig. S5. The cumulative LIN release of SA–PVA@GLIN₁₀₀ at various temperatures: (a) 83.3 wt%, (b) 86.2 wt%, (c) 88.2 wt%, (d) 89.7 wt%, and (e) 90.9 wt%. Fig. S6. The cumulative LIN release of SA–PVA_90.9@GLIN at various temperatures: (a) 50 µg mL⁻¹, (b) 100 µg mL⁻¹, (c) 200 µg mL⁻¹, and (d) 400 µg mL⁻¹. Fig. S7. The cumulative LIN release of SA–PVA@SLIN₁₀₀ at various temperatures: (a) 83.3 wt%, (b) 86.2 wt%, (c) 88.2 wt%, (d) 89.7 wt%, and (e) 90.9 wt%. Fig. S8. The cumulative LIN release of SA–PVA_90.9@SLIN at various temperatures: (a) 50 µg mL⁻¹, (b) 100 µg mL⁻¹, (c) 200 µg mL⁻¹, and (d) 400 µg mL⁻¹. Table S1. SA–PVA@SLIN₁₀₀ at different PVA solid contents. Table S2. SA–PVA@GLIN₁₀₀ at different PVA solid contents. Table S3. SA–PVA_90.9@SLIN at different LIN concentrations. Table S4. SA–PVA_90.9@SLIN at different LIN concentrations. See DOI: https://doi.org/10.1039/d5cp03379a.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (22308049 and 82360763), the Jiangxi Provincial Key Laboratory of Flexible Electronics (20242BCC32010), the China Postdoctoral Science Foundation (2023M740486), Ningbo Natural Science Foundation (2024J096), and the Graduate School-level Innovation Fund of Jiangxi Science and Technology Normal University (YC2024-X55). The authors express their heartfelt gratitude to Prof. Baoyang Lu and Caicai Jiao (Jiangxi Science and Technology Normal University) for valuable discussions and extend their thanks to Manting Song, Yu Lin, Gen Li, Juan Teng, Changchen Gong and Ziyang Qiu for the invaluable assistance rendered throughout the course of this research.

Notes and references

1. J. Liu, S. Guo, S. Hong, J. Piao and M. Piao, Curr. Drug Delivery, 2024, 21, 1537–1547.
2. S. Hua, H. Ma, X. Li, H. Yang and A. Wang, Int. J. Biol. Macromol., 2010, 46, 517–523.
3. H. Bhardwaj, S. Khute, R. Sahu and R. K. Jangde, Curr. Drug Targets, 2023, 24, 1239–1259.
4. R. Lobmann, G. Schultz and H. Lehnert, Diabetes Care, 2005, 28, 337–442.
5. V. Ramachandran, T. Mohanasundaram, D. Karunakaran, M. Gunasekaran and R. Tiwari, Curr. Diabetes Rev., 2023, 19, 127–139.
6. W. Zhao, L. Qiang, C. Zhang, S. Li, Y. Liu, C. Wang, X. Ma, J. Wang and Y. Bao, ACS Appl. Mater. Interfaces, 2024, 16, 50175–50187.
7. M. Figueroa-Pizano, I. Vélaz, F. Peñas, P. Zavala-Rivera, A. Rosas-Durazo, A. Maldonado-Arce and M. Martínez-Barbosa, Carbohydr. Polym., 2018, 195, 476–485.
8. Q. Wang, W. Zheng, J. Wang, C. Jiao, D. Gao, J. Liu and B. Lu, ACS Appl. Mater. Interfaces, 2025, 17, 30747–30758.
9. G. Xu, K. Liu, X. Chen, Y. Lin, C. Yu, X. Nie, W. He, N. Karin and Y. Luan, Nanoscale Horiz., 2024, 9, 295–304.
10. S. Khan and N. Ranjha, Polym. Bull., 2014, 71, 2133–2158.
11. J. Varshousaz and N. Koopaie, Iran. Polym. J., 2002, 11, 123–131.
12. Y. S. Zhang and A. Khademhosseini, Science, 2017, 356, eaaf3627,  DOI:10.1126/science.aaf3627.
13. P. Yang, X. Chen, Y. Qin, L. Yu, G. Ge, W. Yin, W. Zhang, W. Li, W. Li and W. Xia, Biomaterials, 2025, 320, 123273.
14. L. Liu, Q. Zhu, Q. Zhang, Z. Dong, C. Chen, G. Zhao and P. Ma, J. Ind. Text., 2024, 54, 15280837241293156,  DOI:10.1177/15280837241293156.
15. J. Yu, R. Wan, F. Tian, J. Cao, W. Wang, Q. Liu, H. Yang, J. Liu, X. Liu, T. Lin, J. Xu and B. Lu, Small, 2024, 20, 2308778.
16. Y. Gong, Z. Xu, J. Wu, J. Zhong and D. Ma, Appl. Surf. Sci., 2024, 657, 159790.
17. J. Qiu, X. Xu, Z. Li, Y. Hu, G. Liu, X. Lv, J. Xu and B. Lu, Nano Energy, 2024, 130, 110057.
18. Z. Li, Z. Qiu, J. Xu and B. Lu, Chem. Eng. J., 2025, 169656.
19. X. Xu, Q. Zhao, Q. Liu, J. Qiu, J. Li, W. Zheng, J. Cao, L. Wang, W. Wang, S. Yuan, A. Fu, H. Yang, C. Wang, J. Xu and B. Lu, Desalination, 2024, 577, 117400.
20. Q. Zhao, J. Liu, Z. Wu, X. Xu, H. Ma, J. Hou, Q. Xu, R. Yang, K. Zhang, M. Zhang, H. Yang, W. Peng, X. Liu, C. Zhang, J. Xu and B. Lu, Chem. Eng. J., 2023, 442, 136284.
21. G. Li, C. Li, G. Li, D. Yu, Z. Song, H. Wang, X. Liu, H. Liu and W. Liu, Small, 2022, 18, 2101518.
22. Z. Zhang, G. Chen, Y. Xue, Q. Duan, X. Liang, T. Lin, Z. Wu, Y. Tan, Q. Zhao, W. Zheng, L. Wang, F. Wang, X. Luo, J. Xu, J. Liu and B. Lu, Adv. Funct. Mater., 2023, 33, 2305705.
23. L. Wang, W. Wang, R. Wan, M. Yao, W. Chen, L. Zhang, J. Xu, X. Liu and B. Lu, Soft Sci., 2025, 5, 2.
24. X. Luo, R. Wan, Z. Zhang, M. Song, L. Yan, J. Xu, H. Yang and B. Lu, Adv. Sci., 2024, 11, 2404679.
25. H. Li, J. Cao, R. Wan, V. R. Feig, C. M. Tringides, J. Xu, H. Yuk and B. Lu, Adv. Mater., 2025, 37, 2415151.
26. M. Dattilo, F. Patitucci, S. Prete, O. I. Parisi and F. Puoci, J. Funct. Biomater., 2023, 14, 55.
27. S. Jacob, A. B. Nair, J. Shah, N. Sreeharsha, S. Gupta and P. Shinu, Pharmaceutics, 2021, 13, 357.
28. Z. Sun, C. Song, C. Wang, Y. Hu and J. Wu, Mol. Pharmaceutics, 2019, 17, 373–391.
29. F. Huang, X. Lu, Y. Yang, Y. Yang, Y. Li, L. Kuai, B. Li, H. Dong and J. Shi, Adv. Sci., 2023, 10, 2203308.
30. H. Liu, C. Wang, C. Li, Y. Qin, Z. Wang, F. Yang, Z. Li and J. Wang, RSC Adv., 2018, 8, 7533–7549.
31. P. Gupta, K. Vermani and S. Garg, Drug Discovery Today, 2002, 7, 569–579.
32. J. Liu, B. Jia, Z. Li and W. Li, Front. Bioeng. Biotechnol., 2023, 11, 1115603.
33. M. Sobczak, Int. J. Mol. Sci., 2022, 23, 4421.
34. S. Zhang, G. Ge, Y. Qin, W. Li, J. Dong, J. Mei, R. Ma, X. Zhang, J. Bai and C. Zhu, Mater. Today Bio, 2023, 18, 100508.
35. S. Fuchs, A. U. Ernst, L.-H. Wang, K. Shariati, X. Wang, Q. Liu and M. Ma, Chem. Rev., 2020, 121, 11458–11526.
36. Z. Hao, X. Li, R. Zhang and L. Zhang, Adv. Healthcare Mater., 2024, 13, 2400513.
37. Q. Sun, Z. Wang, B. Liu, F. He, S. Gai, P. Yang, D. Yang, C. Li and J. Lin, Coord. Chem. Rev., 2022, 451, 214267.
38. R. Murmu, D. Roy, H. Sutar, P. Senapati and S. Mohapatra, J. Polym. Mater., 2022, 39, 89–109.
39. X. Xu, Q. Sun, A. Xu and X. Guo, J. Polym. Mater., 2022, 39, 167–181.
40. W. Zheng, L. Wang, H. Jiao, Z. Wu, Q. Zhao, T. Lin, H. Ma, Z. Zhang, X. Xu, J. Cao, J. Zhong, J. Xu and B. Lu, Chem. Eng. J., 2023, 455, 140553.
41. I. Sterin, A. Tverdokhlebova, E. Katz and O. Smutok, ACS Appl. Mater. Interfaces, 2024, 16, 28222–28229.
42. T. Wang, W. Yi, Y. Zhang, H. Wu, H. Fan, J. Zhao and S. Wang, Colloids Surf., B, 2023, 222, 113096.
43. X. Kuang, M. O. Arıcan, T. Zhou, X. Zhao and Y. S. Zhang, Acc. Mater. Res., 2022, 4, 101–114.
44. J. Doupis, Drug Des., Dev. Ther., 2014, 431–446,  DOI:10.2147/dddt.s59523.
45. M. Long, L. Cai, W. Li, L. Zhang, S. Guo, R. Zhang, Y. Zheng, X. Liu, M. Wang and X. Zhou, Diabetes, 2016, 65, 780–793.
46. M. Long, L. Cai, W. Li, L. Zhang, S. Guo, R. Zhang, Y. Zheng, X. Liu, M. Wang and X. Zhou, Diabetes, 2018, 67, 518–531.
47. L. B. Pucar, E. P. Pugel, D. Detel and J. Varljen, Wound Repair Regen., 2017, 25, 25–40.
48. P. Yang, X. Chen, Y. Qin, L. Yu, G. R. Ge, W. L. Yin, W. Zhang, W. M. Li, W. H. Li, W. Y. Xia, Z. B. Wu, F. Ding, J. X. Bai, F. W. Meng and D. C. Geng, Biomaterials, 2025, 320, 123273.
49. M. Cao and Z. Bloomgarden, J. Diabetes, 2020, 12, 844–847.
50. T. Karve and A. Banga, Int. J. Pharm., 2024, 654, 123992.
51. R. Marfella, F. Sasso, M. Rizzo, P. Paolisso, M. Barbieri, V. Padovano, O. Carbonara, P. Gualdiero, P. Petronella and F. Ferraraccio, J. Diabetes Res., 2012, 2012, 892706.
52. W. Yu, J. Chen, Q. Gao, Y. Guo, S. Zhang, Y. Pan, B. Nie, X. Zhang, L. Jiang and J. Qiu, Chem. Eng. J., 2025, 519, 164695.
53. Y. Zheng, H. Huang, J. Yu, Z. Hu and Y. Wang, Chem. Eng. J., 2023, 454, 140054.
54. P. Kaur, V. Gondil and S. Chhibber, Int. J. Pharm., 2019, 572, 118779.
55. Y. Jiang, Y. Hou, J. Fang, W. Liu, Y. Zhao, T. Huang, J. Cui, Y. Yang and Z. Zhou, Int. J. Polym. Anal. Charact., 2019, 24, 132–141.
56. M. Saraiva, M. d S. Campelo, J. Neto, M. d C. Gonzaga, M. Bastos, S. d A. Soares, N. Ricardo, G. Cerqueira, R. d C. Leitão and M. Ribeiro, Int. J. Biol. Macromol., 2023, 244, 125278.
57. M. Afshar, G. Dini, S. Vaezifar, M. Mehdikhani and B. Movahedi, J. Drug Delivery Sci. Technol., 2020, 56, 101530.
58. K. Zhang, Q. Feng, Z. Fang, L. Gu and L. Bian, Chem. Rev., 2021, 121, 11149–11193.
59. X. Zhao, X. Chen, H. Yuk, S. Lin, X. Liu and G. Parada, Chem. Rev., 2021, 121, 4309–4372.
60. N. Ta, Y. Li, C. Schuyler, M. Virella and Y. Huang, Atherosclerosis, 2010, 213, 429–435.
61. P. Candry, B. Godfrey, Z. Wang and F. Sabba, Sci. Rep., 2022, 12, 20822.
62. Z. Lin, Y. Gao, L. Chen, W. Wang and T. Chen, J. Environ. Chem. Eng., 2025, 117584,  DOI:10.1016/j.jece.2025.117584.
63. Y. Huang, T. Dong, X. Zhang, S. Xu, X. Song, Z. Wang, M. Qin, L. Deng and Y. Li, J. Polym. Mater., 2025, 42 DOI:10.32604/jpm.2025.060699.
64. M. Liao, Y. Zhao, Y. Pan, J. Pan, Q. Yao, S. Zhang, H. Zhao, Y. Hu, W. Zheng and W. Zhou, J. Mater. Sci., 2023, 58, 5756–5772.
65. W. Zhang, Y. Wu, H. Chen, Y. Gao, L. Zhou, J. Wan, Y. Li, M. Tang, Y. Peng and B. Wang, Sep. Purif. Technol., 2024, 336, 126261.
66. H. Liu, J. Gu, Y. Liu, L. Yang, A. Li, J. Yu, L. Wang and X. Qin, Langmuir, 2023, 39, 13641–13648.
67. J. Ren, J. Li, Z. Xu, Z. Du and F. Cheng, J. Environ. Chem. Eng., 2021, 9, 105370.
68. L. Gu, X. Sun, J. Pan, D. Liu, L. Huang, Y. Yu, B. Yu and H. Cong, Biomaterials, 2025, 123432.
69. J. Xu, W. Song, L. Ren, N. Wu, R. Zeng, S. Wang, Z. Wang and Q. Zhang, Int. J. Biol. Macromol., 2024, 263, 130282.
70. H. Joodaki and M. Panzer, Proc. Inst. Mech. Eng., Part H, 2018, 232, 323–343.
71. S. Liu, Y. Rao, H. Jang, P. Tan and N. Lu, Matter, 2022, 5, 1104–1136.
72. J. Li and D. Mooney, Nat. Rev. Mater., 2016, 1, 1–17.
73. A. Dravid, B. Raos, Z. Aqrawe, S. Parittotokkaporn, S. O’Carroll and D. Svirskis, Front. Chem., 2019, 7, 638.
74. A. C. Marques and M. Vale, Materials, 2021, 14, 4247.
75. K. Nakanishi and N. Tanaka, Acc. Chem. Res., 2007, 40, 863–873.
76. G. Cheng, S. Wang, W. Li, X. Liu and Q. Huang, Ind. Crops Prod., 2024, 220, 119234.

---