Towards injured joint rehabilitation: structural color hydrogels for accelerated wound healing and rehabilitation exercise monitoring

Abstract

Joint injuries caused by severe acute trauma seriously affect patients’ mobility and quality of life. Traumatic or postoperative wound healing and rehabilitation training are both essential for restoring joint functions, calling for effective wound healing materials that are also capable of monitoring rehabilitation training for joint condition evaluation and physical therapy guiding. Herein, a structural color hydrogel for wound care and naked-eye rehabilitation exercise monitoring of injured joints is designed by constructing a hybrid double-network, which contains a covalently crosslinked network and a Zn²⁺ coordination based dynamic network. The crosslinking formed by Zn²⁺ coordination endows the structural color hydrogel with enhanced mechanical properties for joint wounds with motion requirements, as well as antibacterial, anti-inflammatory, and pro-angiogenic properties that promote wound healing. Meanwhile, the Poisson's ratio of the structural color hydrogel can be easily tuned by varying the covalently-crosslink density to achieve sensibility ranging from 3.6 nm to 6.2 nm photonic-bandgap shift per 1% strain, achieving a remarkable color change responding to joint range-of-motion from minimal (0–2°) to wide-range (0–90°) bending during rehabilitation exercises. This structural color hydrogel provides an approach to the multi-stage management of joint injuries and real-time clinical insights into rehabilitation progress.

---

Meng Qin

Meng Qin received her PhD from the Institute of Chemistry, Chinese Academy of Sciences (ICCAS) in 2016 under the supervision of Prof. Yanlin Song. Afterwards, she worked as a research associate at the Department of Materials Science and Engineering, UCLA. She is currently an associate professor at the College of Polymer Science and Engineering, Sichuan University. Her current research is focused on biomedical polymers.

1. Introduction

Joint injuries caused by severe acute trauma (e.g., cuts, stabs, and burns) involve skin wounds and damage to ligaments, tendons, nerves, vessels, or bones.¹ These injuries significantly impair the mobility and quality of life. Whether surgical treatment is required or not, effective healing for traumatic wounds or postoperative wounds and rehabilitation training are essential to restore joint functions and ensure long-term health. Initially, protecting the injured site from infection and promoting tissue regeneration are priorities, while minimizing joint movement to prevent complications.² At this stage, early rehabilitation exercises, such as passive movement or gentle stretching, can enhance blood circulation and prevent stiffness.³ As healing progresses, weight-bearing and range-of-motion (ROM) exercises gradually intensify to strengthen muscles and stabilize joints. Evaluating rehabilitation progress is vital to verify the effectiveness of interventions and guide subsequent treatment.⁴ However, traditional assessments rely on subjective judgment or measurement by physical therapists, which may lead to inconsistent outcomes and inefficiencies. For patients with nerve injuries, the evaluation is even harder due to the insensitive tactile sense. In order to promote full recovery of the injured joints, the development of wound healing materials that are simultaneously capable of accurate and sensitive monitoring of rehabilitation exercises is in high demand.

Structural color hydrogels which are achieved by introducing a photonic crystal structure into the hydrogel matrix are promising materials to meet the requirements of simultaneous wound therapy and rehabilitation exercise monitoring. On the one hand, the hydrogel matrix can be functionalized with antibacterial, anti-inflammatory, and stimulating angiogenesis properties, etc., to promote wound healing.^5–12 Structural color hydrogel dressings have emerged in recent years.^13–16 For example, Zhang et al. proposed a self-healing hydrogel dressing based on structural color microspheres, which can be used for tissue regeneration, collagen accumulation, and vascular regeneration in diabetic wound models.¹⁷ Specifically, in regard to joint wounds, the frequent motion and bending of joints require improved mechanical properties of the hydrogels.¹⁸ On the other hand, the photonic crystal structure can provide colorimetric sensing signals in response to mechanical deformation, making them ideal for point-of-care monitoring of motion and bending.^19–21 Although structural color hydrogels have been developed for joint angle monitoring, current research studies are focused on large angular ranges (e.g. 30°, 60°, and 90°), lacking the high-sensitivity detection of a minor angular change.^22,23 Indeed, joint ROM is minimal in the beginning and increases gradually with the rehabilitation process. Therefore, monitoring both minimal and wide-range ROM is needed, depending on the rehabilitation stage. Taking the combination of therapy and monitoring functions into consideration, a major challenge of constructing structural color hydrogels for injured joint rehabilitation is to achieve good mechanical properties, antibacterial properties, bioactivity, and an adjustable colorimetric mechanical response via a facile strategy.^24–26

In this work, a structural color hydrogel with a hybrid double-network is developed for accelerated wound healing and rehabilitation exercise monitoring (Fig. 1). The first network of the hydrogel matrix is established based on covalent crosslinking of acrylamide and vinylimidazole to form poly(acrylamide-co-vinylimidazole) (PAV), in which a non-close packed photonic crystal structure is embedded to provide a broad range of color change. Acrylamide endows the hydrogel with good optical properties and tunable mechanical properties, while the imidazole groups of vinylimidazole provide sites for subsequent functionalization.²⁷ The second network is formed by the coordination of the imidazole group from PAV with Zn²⁺ by simply soaking the first network in an aqueous solution of Zn²⁺. This strong but still dynamic non-covalent crosslinked network can significantly toughen the hydrogel, making it applicable to joint sites. Due to the gradual release of Zn²⁺, this PAV-Zn structural color hydrogel exhibits excellent antibacterial, anti-inflammatory and pro-angiogenic properties, thus accelerating wound healing. To demonstrate the ability of rehabilitation training monitoring, the structural color hydrogel is applied to fingers, which are the most flexible parts of extremities and demand fine movements, to monitor joint ROM. By adjusting the crosslink density of the first network, the Poisson's ratio of the hydrogel can be tuned from 0.36 to 0.43, resulting in sensibility ranging from 3.6 nm to 6.2 nm photonic-bandgap shift per 1% strain, thus allowing for adaptive monitoring of minimal ROM (2° bending can be naked-eye distinguished) or wide-range ROM monitoring (0–90°), depending on the stage of rehabilitation. By combining the capabilities of enhanced wound healing and adjustable rehabilitation exercise monitoring, this structural color hydrogel offers valuable clinical insights into injured joint rehabilitation and health management.

Fig. 1 Structural color hydrogels for accelerated wound healing and rehabilitation exercise monitoring. The hydrogel contains a covalently-crosslinked network and a Zn²⁺ coordination network, providing antibacterial, anti-inflammatory, and pro-angiogenic properties for wound healing and adjustable sensibility for monitoring joint motions.

2. Results and discussion

2.1. Preparation and mechanical properties of the PAV-Zn hydrogel

A covalently-crosslinked PAV network is firstly constructed by polymerization of acrylamide and vinylimidazole with an N,N′-methylene bisacrylamide (MBAA) crosslinker. The as-prepared PAV hydrogel can be further crosslinked with Zn²⁺ through metal coordination, and the resulting PAV-Zn hydrogel has a stable hybrid double-network. As shown in the scanning electron microscopy (SEM) images in Fig. 2a, compared with the porous structure of PAV, a denser network is formed after introducing Zn²⁺. Energy dispersive spectroscopy (EDS) and X-ray photoelectron spectroscopy (XPS) analyses (Fig. S1, ESI†) confirm the presence and uniform distribution of Zn²⁺ in the network. The N 1s spectra of PAV and PAV-Zn are shown in Fig. 2b. The characteristic peaks at 400.84 eV and 398.94 eV of PAV are attributed to the amine group and imine group of the imidazole ring. After the introduction of Zn²⁺, these peaks shift to 400.90 eV and 399.12 eV, respectively, indicating the formation of Zn²⁺-imidazole complexation.^28,29 The coordination between Zn²⁺ and the imidazole group is further characterized by Fourier transform infrared (FTIR) spectra (Fig. 2c). The absorption peak corresponding to the C=N stretching vibration of the imidazole ring shifted from 1498 cm⁻¹ to 1517 cm⁻¹. At the same time, the intensity of the PVI ring mode at 915 cm⁻¹ decreased, while a new absorption band appeared at 954 cm⁻¹, suggesting the inductive effect of electron-deficient zinc ions and their interaction with the imidazole group.

Fig. 2 Supramolecular crosslinking formed by introducing Zn²⁺ into PAV. (a) SEM images of PAV and PAV-Zn, with EDS mapping of characteristic elements in PAV-Zn. (b) XPS spectra of N 1s of PAV and PAV-Zn. (c) FTIR spectra of PAV and PAV-Zn. (d) and (e) Stress–strain curves of PAV and PAV-Zn hydrogels with different times of Zn²⁺ crosslinking (d) and PAV-Zn hydrogels with different concentrations of Zn²⁺ crosslinking (e).

The metal coordination results in a significant improvement in the mechanical properties of the hydrogel. As shown in Fig. 2d, when the chemical crosslinker MBAA for the first network is fixed at 0.5 mol% of the total monomers and Zn²⁺ concentration fixed at 100 mM, the PAV hydrogel treated with Zn²⁺ for 30 min exhibits increased tensile strain from 106.4% to 352.3% and increased stress from 58.3 kPa to 161.6 kPa. This enhancement of mechanical properties can significantly benefit the use of PAV-Zn hydrogels for joint sites. A longer crosslinking time (60 min) is of limited help in improving the mechanical properties of the hydrogels, confirming the rapidity of the metal coordination equilibrium. The mechanical properties of the PAV-Zn hydrogel depend on the concentration of Zn²⁺ (Fig. 2e). When the Zn²⁺ concentration is 50 mM, the tensile strain can reach 376.9% and the stress increases to 365.4 kPa. Higher Zn²⁺ concentrations cause significantly reduced strength of the hydrogel. When the Zn²⁺ concentration increases to 400 mM, the crosslinking of the second network is drastically inhibited. This is because higher concentrations of Zn²⁺ result in lower pH values of the aqueous solutions, which promotes the protonation of the imine group at the imidazole ring and thus prevents the formation of Zn²⁺-imidazole complexation. Therefore, in the following experiments, we choose the treatment of the PAV hydrogel with 50 mM Zn²⁺ for 30 min to construct the second network. In addition, the hydrogel exhibits a relatively low swelling rate (Fig. S2, ESI†). This is attributed to the increased cross-linking density caused by imidazole-metal ion complexation, which makes the hydrogel network dense and inhibits water absorption. At the same time, the addition of Zn²⁺ also increases the mechanical stiffness of the material while reducing the adhesive strength.²⁸ The wet porcine skin adhesion test confirms that the adhesion is negligible, and the hydrogel can be completely removed from the tissue surface without the residue (Fig. S3a, ESI†). Unlike traditional hydrogel dressings that adhere to the skin and swell after absorbing wound exudate, resulting in wound expansion, pain, and increased infection risk, the low adhesion and low swelling properties of the PAV-Zn hydrogel prevent wound stretching, promote healing, and minimize complications. To prevent displacement during joint motion monitoring, the hydrogel is covered with a commercial 3M dressing to ensure firm contact with the skin while maintaining structural integrity and monitoring accuracy (Fig. S3b, ESI†). These results confirm the successful synthesis of the hybrid double-network PAV-Zn hydrogel with good mechanical properties.

2.2. In vitro biocompatibility assessment and antibacterial properties

Whether it is for trauma or postoperative wounds, the PAV-Zn hydrogel should first protect the injured area from infection and promote tissue regeneration, which requires good biocompatibility and antibacterial efficacy. Since Zn²⁺ is contained in the hydrogel, we first quantify the cumulative release concentration of Zn²⁺ using an inductively coupled plasma optical emission spectrometer (ICP-OES) (Fig. 3a). After immersing the PAV-Zn hydrogel in PBS for 12 h, the release of Zn²⁺ reaches equilibrium, at a concentration of 3.25 mg L⁻¹, which is within the safe concentration range. This concentration is particularly significant as previous studies have demonstrated the concentration-dependent bioactivity of Zn²⁺. Specifically, within the range of 16–80 μmol L⁻¹ (equivalent to 1.056–5.23 mg L⁻¹), Zn²⁺ has been shown to enhance cellular functions by promoting both proliferation and migration.^30,31 However, it should be noted that at concentrations exceeding 65.13 ppm (ca. 65.13 mg L⁻¹), Zn²⁺ exhibits inhibitory effects on macrophage polarization, particularly in suppressing the M2 phenotype, which can consequently impede wound healing processes.³² Despite the release of Zn²⁺, the mechanical properties and the color of the corresponding structural color hydrogel remain stable (Fig. S4, ESI†), indicating that Zn²⁺-imidazole crosslinking is unaffected and the release originates from excess Zn²⁺ in the hydrogel network. To verify the cytocompatibility of the hydrogel, mouse fibroblasts (L929) are co-cultured in the hydrogel extract for three days. L929 live/dead staining shows that the cells in the PAV-Zn group and the control group have a normal morphology, and no obvious cell apoptosis or necrosis is observed (Fig. 3b). In addition, the CCK-8 assay shows that the cells in each group continue to proliferate within three days, and there is no statistical difference in cell viability in the PAV-Zn group compared with the control group (Fig. S5, ESI†). In addition, the hemocompatibility of PAV-Zn is tested by a commonly used in vitro hemolysis assay. As shown in Fig. 3c, unlike the hemolysis of the positive control group caused by water treatment, the PAV-Zn hydrogel shows no obvious signs of hemolysis after one hour of incubation. Statistical analysis of the hemolysis ratios shows no significant difference between the PAV-Zn and PBS negative control groups. The above results indicate the good biocompatibility of the PAV-Zn hydrogel.

Fig. 3 Zinc ion release, biocompatibility and antimicrobial efficacy of the PAV-Zn hydrogel. (a) The cumulative release curve of zinc ions from the PAV-Zn hydrogel. (b) Representative fluorescence images of live (green) and dead (red) staining L929 cells cultured on various samples at different time points. (c) Photographs from the hemolytic activity assay using hydrogel dispersion as well as PBS (negative control) and ultrapure water (positive control) and the corresponding hemolysis ratios (n = 3, mean ± SD, ***p < 0.001, and n.s. means not significant). (d) Colonization and (e) inhibition ratios of the S. aureus and E. coli in control, PAV and PAV-Zn groups (n = 3, mean ± SD, *p < 0.05, **p < 0.01, and ***p < 0.001). (f) SEM images of S. aureus and E. coli in control, PAV, and PAV-Zn groups.

The PAV-Zn hydrogel has a broad-spectrum antibacterial ability, attributed to the presence of imidazole groups and Zn²⁺. The antibacterial ability can be directly and qualitatively illustrated by the results of co-culture with Escherichia coli (E. coli, ATCC 6538, Gram-negative bacteria) and Staphylococcus aureus (S. aureus, ATCC 25922, Gram-positive bacteria) (Fig. 3d and e). The results show that the number of colonies in the PAV group is reduced compared with the dense blank group, and the inhibition rates are 18.4% and 12.7% for S. aureus and E. coli, respectively. The imidazole groups in PAV hydrogels have a mild antibacterial effect and can cause bacterial death by destroying the integrity of bacterial membranes.^33,34 After further crosslinking by Zn²⁺, the hydrogel shows significantly improved antibacterial properties.³⁵ The bacterial population in the PAV-Zn group remarkably reduces. The inhibition rate of PAV-Zn on E. coli exceeds 99.8%, and the inhibition rate on S. aureus reaches more than 95.3%. Morphological changes of S. aureus and E. coli after different treatments are characterized. As shown in Fig. 3f, both S. aureus and E. coli in the control group show complete and smooth cell membranes. In contrast, bacteria treated with the PAV hydrogel show some shrinkage and damage, indicating that the hydrogel has certain antibacterial abilities. Despite this, some well-shaped bacteria are still observed in the PAV group, indicating that the antibacterial properties are not sufficient to kill all bacteria. However, in the PAV-Zn group, the bacterial cell membranes show severe damage, and the bacteria are deformed or even ruptured. These results confirm that the PAV-Zn hydrogel has a significant bactericidal effect and can induce the destruction of bacterial cell membranes.

2.3. In vivo accelerated wound healing ability

Since the PAV-Zn hydrogel demonstrates good mechanical properties, in vitro biocompatibility, and in vitro antibacterial ability, its ability to accelerate wound healing is evaluated by establishing a S. aureus infected full-thickness skin defect model. We speculate that the inherent anti-inflammatory and pro-angiogenic properties of zinc ions make the hydrogel suitable for participating in the wound healing process.^36,37 The wounds are divided into three groups: control, PAV, and PAV-Zn. The wound healing process is photographed and recorded on days 0, 3, 6, and 10, respectively. Fig. 4a shows that the wound healing rate in the control group is the lowest, and a large scab is visible at the wound site. The PAV hydrogel can promote the healing process to some extent since the imidazole group can inhibit the growth of bacteria in the infected wound. In comparison, the wounds treated with the PAV-Zn hydrogel exhibit the fastest healing rate and more intact morphology, leaving a wound area of 0.9% on day 10 (Fig. S6, ESI†). Histopathological analysis is conducted on day 10 through hematoxylin and eosin (H&E) staining. As shown in Fig. 3b, the surface epithelium of the control group shows a large amount of inflammatory response. Compared with the control and PAV groups, the regenerated epithelium of the PAV-Zn group is better integrated with the surrounding normal tissue, and the epidermis and dermis of the lesion site exhibit a normal arrangement morphology. Given that collagen can support functional cells for wound contraction and healing, we further investigate the effect of the hydrogels on collagen deposition by Masson staining. The results show that collagen deposition is accelerated in all treatment groups compared with the control group (Fig. 4b). The PAV-Zn hydrogel group shows the heaviest blue, proving that there is more collagen deposition. Next, we analyze the proinflammatory cytokine interleukin-6 (IL-6) by immunofluorescence staining (Fig. 4c). The green fluorescent labeled protein significantly decreases in treatment groups. The corresponding quantitative analysis shows that the expression of IL-6 in hydrogel-treated groups is much lower than that in the control group (Fig. 4d). The PAV-Zn group demonstrates the lowest IL-6 expression. Angiogenesis is considered to be a key factor in wound repair, and therefore, CD31 (red) immunofluorescence staining is performed to characterize the newly formed blood vessels (Fig. 4c and d). While there are minimal CD31-positive areas in the control group, the PAV-Zn group shows the highest level of angiogenesis among all groups. In addition, H&E staining of the main organs (heart, liver, spleen, lungs, and kidneys) of each group of mice shows no significant tissue damage or pathological changes (Fig. S7, ESI†). Additionally, Zn is known to degrade naturally in physiological environments without producing gas or harmful byproducts,^38,39 ensuring long-term biocompatibility. Compared with traditional hydrogels, PAV-Zn hydrogels can easily achieve multifunctionality without the addition of drugs or growth factors.^40,41 These results suggest that the PAV-Zn hydrogels can be used as biocompatible and effective wound dressings, providing antimicrobial ability, inhibiting inflammatory responses, promoting collagen deposition and angiogenesis, and accelerating wound healing, which is crucial for joint rehabilitation in the initial stage.

Fig. 4 In vivo treatment efficacy of the PAV-Zn hydrogel. (a) Representative images of mice skin wounds and wound traces over 10 days in the control group, PAV group, and PAV-Zn group, respectively. (b) Corresponding H&E and Masson staining of the wound tissue on day 10. (c) Immunofluorescence assay of IL-6 (green), and CD31 (red), and DAPI (blue) of the wounds for each group on day 10. (d) Quantification of the IL-6 expression and CD31 expression for each group on day 10 (n = 3, mean ± SD, **p < 0.01, and ***p < 0.001).

2.4. Adjustable mechanical response of the structural color hydrogel

As the injured joints heal relatively, rehabilitation training can be gradually introduced. To achieve naked-eye monitoring of rehabilitation exercises, PAV-Zn structural color hydrogels are constructed by incorporating non-close packed photonic crystal structures into the PAV hydrogel network, followed by Zn²⁺ treatment. Compared with the porous structure and relatively poor mechanical properties of the PAV photonic crystal hydrogel, the PAV-Zn photonic crystal hydrogel shows more compact morphology and improved toughness, which is consistent with the changes of pure hydrogels (Fig. S8 and S9, ESI†). The color of the hydrogel corresponds to the location of the photonic band gap, which can be calculated by Bragg's law⁴²

mλ = 2nd cos θ

where λ is the peak wavelength, m is the order of diffraction, n is the refractive index, d is the lattice spacing, and θ is the angle of incident light. When stretching is applied to the structural color hydrogel, the lattice spacing decreases, leading to a blue shift of the photonic band-gap and thus a color change (Fig. 5a). The mechanochromic sensibility of PAV-Zn can be tuned by simply varying the content of the MBAA crosslinker. As shown in Fig. 5b, when the crosslinker content is 0.5 mol% of the total monomers, the color of the PAV-Zn hydrogel changes from violet-red to green as it is stretched to 96%. An increase of the MBAA content to 1 mol% leads to a wide-range color change from orange-red to blue when the hydrogel is subjected to 60% strain. More sensitively, when the MBAA content is 2 mol%, the hydrogel exhibits a color change from pink to cyan when it is pulled from 0 to strain of 42%. The color changes are then quantitatively analyzed by calculating ΔE (ΔE = [(ΔL*)² + (Δa*)² + (Δb*)²]^1/2, where L*, a*, and b* denote the lightness, red/green value, and yellow/blue value, respectively (Table S1, ESI†)) based on the Commission Internationale de L’Eclairage Lab (CIE L*a*b*) system.⁴³ As shown in Fig. 5c, an increase of the MBAA content results in larger ΔE upon specific strain. For example, when the strain is 12%, the hydrogels with 0.5 mol%, 1 mol%, and 2 mol% MBAA exhibit ΔE values of 7.3, 10.1, and 18.9, respectively, indicating a more sensitive mechanochromic response of the hydrogel with a higher content of MBAA.

Fig. 5 (a) Schematic representation of the change in lattice spacing and color of a non-close-packed photonic crystal hydrogel during stretching. (b) Photographs of the structural color hydrogels with different MBAA contents at different tensile strains. (c) Quantitative analysis of color changes of hydrogels with different MBAA contents. (d)–(f) Reflectance spectra of structural color hydrogels with MBAA contents of 0.5 mol% (d), 1 mol% (e) and 2 mol% (f) in a tensile strain range of 0–7%. (g) Poisson's ratios and sensitivity for each group of the samples.

The sensibility of PAV-Zn structural color hydrogels depends on Poisson's ratios of the materials. A higher Poisson's ratio suggests the material is more incompressible, and thus transverse deformation is more pronounced when longitudinal strain is applied to the material. As a result, the corresponding structural color hydrogel exhibits a decrease of the lattice spacing and color change to a greater extent. Adjusting the MBAA content can induce the Poisson's ratio change of the PAV-Zn hydrogel (Fig. S10, ESI†). When the MBAA contents are 0.5 mol%, 1 mol%, and 2 mol%, the Poisson's ratios of the hydrogels are 0.36, 0.38, and 0.43, respectively. The possible reason for this Poisson's ratio change is that a higher MBAA content results in a higher crosslink density of the first covalent network, and inhibits the diffusion of Zn²⁺ into the network, which leads to a lower metal coordination based dynamic crosslink density. This is evidenced by lower fracture strain and stress of the PAV-Zn hydrogel with a higher content of MBAA (Fig. S11, ESI†). Since Zn²⁺-imidazole complexation is easily broken under an applied force, the hydrogel with a higher fraction of Zn²⁺-imidazole crosslinks can give more spacing to be compressed and thus shows a lower Poisson's ratio. We then investigate the photonic-band gap shift (Δλ) of the hydrogels within a narrow range of strain (Δε), which corresponds to that applied to characterizing Poisson's ratio. As shown in Fig. 5d–f, with an increase of strain to 7%, the hydrogels with MBAA contents of 0.5 mol%, 1 mol%, and 2 mol% show Δλ values of 24.4 nm, 31.5 nm, and 42.7 nm, demonstrating sensitivity (photonic-band gap shift per 1% strain, Δλ/Δε, nm %⁻¹) of 3.6 nm %⁻¹, 4.5 nm %⁻¹ and 6.2 nm %⁻¹ by linear fitting analysis, respectively (Fig. S12, ESI†). Therefore, an increase of the MBAA content results in higher sensitivity of the mechanochromic response, which is consistent with the variation trend of the Poisson's ratio (Fig. 5g). Cyclic mechanical tests demonstrate that the hydrogels exhibit robust mechanical stability and resilience (Fig. S13, ESI†), which are attributed to the dynamic metal–ligand coordination within the hydrogel network.^44,45 In addition, the structural color change of the hydrogel is stable during stretching–releasing cycles (Fig. S14, ESI†). The adjustable sensibility and stability endow the hydrogel with adaptive motion monitoring in different stages of the rehabilitation process.

2.5. ROM monitoring of fingers for rehabilitation

ROM is used to evaluate the extent or limit that a joint can move, which is a basic indicator to assess the motor functions of injured joints. While the ROM measurement is usually conducted by a physical therapist using goniometry, it is inconvenient for patients to perform continuous monitoring during ROM exercises and assess the joint condition. Compared with electrical sensors that rely on external devices and most current photonic crystal-based sensors with limited sensitivity,^46,47 the PAV-Zn structured color hydrogel provides a highly sensitive and continuous visualization method for ROM monitoring, which can significantly guide clinical treatment and training.

Since fingers are the most flexible parts of extremities and demand fine movements, we demonstrate the ROM monitoring ability of the structural color hydrogels by attaching them to the finger joints. We firstly use a human hand model to simulate the bending in a small angle range, where the proximal interphalangeal joint is modeled as a spherical structure with a measured curvature radius of 10 mm, and a high-sensitivity hydrogel (2 mol% MBAA) is attached to the interphalangeal joint of the index finger (Fig. 6a). The structural color hydrogel can achieve a pink to yellow-green transition at a small angle of 2°. In contrast, the bending from 0° to 6° is hard to be observed without colorimetric indication. A wide range of color changes from pink to violet-blue is demonstrated to indicate a ROM of 0–25°, consistent with the initial phase of rehabilitation training outlined in the clinically applied Brunnstrom standard.⁴⁸ This suggests the hydrogel's applicability in early-stage rehabilitation, where narrow-range ROM exercises but high-sensitivity monitoring are required. Using structural color hydrogel with a lower content of MBAA can achieve wider-range ROM monitoring, and thus it can be applied to the subsequent stages of rehabilitation training (Fig. S15, ESI†).

Fig. 6 ROM monitoring by the structural color hydrogel. (a) Detection of 0–25° bending of a finger model. Minimal bending can be observed by a change in the structural color. (b)–(d) Photographs of gently holding a ball with a radius of 4.5 cm (b) or grasping it with force (c), as well as gently holding a ball with a radius of 4.0 cm (d). (e)–(g) Digital photographs corresponding to the structural color hydrogel sensor array at finger joints. (h)–(j) Corresponding RGB values of the sensor array.

The structural color hydrogels are then attached to all finger joints to establish a sensor array for pattern-based ROM exercise monitoring. For convenience, the metacarpophalangeal joint and interphalangeal joint of the thumb are named T1 and T2, respectively. Similarly, the metacarpophalangeal joint, proximal interphalangeal joint and distal interphalangeal joint of the index finger, middle finger, ring finger and little finger are named I1–I3, M1–M3, R1–R3 and L1–L3, respectively. The hydrogels with 2 mol% MBAA are chosen to monitor the ROM exercises. When a ball with a radius of 4.5 cm is held gently (Fig. 6b), the sensor array composed of 14 hydrogels shows a unique color pattern (Fig. 6e). Further ROM exercise of grasping the ball with force leads to a change of the color pattern (Fig. 6c and f). Although there is no obvious difference between the gestures of “grasping” and gentle “holding” by naked-eye observation, the sensor array can give distinguished color patterns. Taking T1 and R1 as an example, the red color of T1 becomes weaker and the blue color of R1 becomes darker. This visualization of minimal motion is of great importance for patients with nerve damage, which can guide them to conduct appropriate ROM exercises without excessive bending. When a smaller ball with a radius of 4.0 cm is gently held, a clear blue shift in the color of the sensor array is observed (Fig. 6d and g). For example, T1 changes from red to light green, and R1 changes from light blue to dark blue. In order to achieve quantitative analysis of the color change, we extract the RGB values of the sensor array. As shown in Fig. 5h–j, there are obvious differences in the responses to various ROM exercises, which confirms the feasibility of the structural color hydrogels for rehabilitation training monitoring. The production of structural color hydrogels demonstrates scalability and cost-effectiveness, leveraging commercially available materials, straightforward synthesis methods, and efficient fabrication techniques, while their customizable properties, such as curvature radius and mechanical sensitivity, make them highly suitable for diverse clinical applications, including personalized rehabilitation and therapeutic interventions.

3. Conclusions

In summary, a structural color hydrogel capable of accelerating wound healing and monitoring rehabilitation exercises is developed to promote full recovery of the injured joints. The hydrogel is composed of a hybrid double-network, in which the first network is established based on covalent crosslinking of acrylamide and vinylimidazole to form PAV, followed by the second network with Zn²⁺-imidazole crosslinking. This PAV-Zn structural color hydrogel shows good mechanical properties, benefiting its application on movable joint sites. Due to the release of Zn²⁺, the biocompatible hydrogel exhibits antibacterial, anti-inflammatory, and pro-angiogenic properties, and thus can significantly accelerate the wound healing process, which is crucial for joint rehabilitation in the initial stage. Besides, the hydrogel shows adjustable structural color responses to mechanical deformations by tuning the Poisson's ratio, conferring the ability of adaptive monitoring of rehabilitation exercises with high sensitivity, which is of great significance for the subsequent stages of joint rehabilitation. The sensibility can be tuned from 3.6 nm to 6.2 nm photonic-bandgap shift per 1% strain, thus achieving the monitoring of minimal ROM (2° bending can be naked-eye distinguished) and wide-range ROM monitoring (0–90°), depending on the stage of rehabilitation. This structural color hydrogel provides a strategy for the multi-stage management of joint injuries and promotes the injured joint rehabilitation.

4. Experimental

4.1. Materials

Tetraethyl orthosilicate (TEOS), concentrated ammonia, ethanol (EtOH), acrylamide (AAm), 1-vinylimidazole (VI), N,N-methylenebisacrylamide (MBAA), 2-hydroxy-2-methyl-1-[4-(2-hydroxyethoxy)phenyl]-1-propanone (Darocur 2959) and zinc chloride were purchased from Macklin. Ultrapure water with a resistivity of 18.2 MΩ cm was used in all experiments. All chemicals were used as received. Escherichia coli (E. coli, ATCC 25922) was purchased from Huan Kai Microbial Sci. & Tech. Co., Ltd (Guangdong, China). Staphylococcus aureus (S. aureus, CMCC 26003) was purchased from Huan Kai Microbial Sci. & Tech. Co., Ltd (Guangdong, China). Alpha-modified Eagle's medium (α-MEM), fetal bovine serum (FBS), and penicillin–streptomycin solution were purchased from Bosco Biotechnology Co., Ltd (Chengdu, China). Propidium iodide (PI) and fluorescein diacetate (FDA) were purchased from Beijing Solarbio Science & Technology Co., Ltd (Beijing, China).

4.2. Synthesis of monodispersed SiO₂ nanoparticles of different sizes

Monodispersed SiO₂ NPs were synthesized by a modified stöber method.⁴⁹ In a typical procedure to synthesize 140 nm SiO₂ NPs, 65 g of ethanol, 7.2 g of aqueous ammonia and 3 g of ultrapure water were added to the flask with vigorous mechanical stirring at a speed of about 500 rpm at 60 °C. Then, 6 mL of TEOS was added to the flask. The reaction continued for 2 h and stopped. The resulting SiO₂ NPs were separated by three cycles of washing with ethanol and centrifugation at 8000 rpm for 5 min, followed by drying under vacuum overnight. Similarly, monodispersed SiO₂ NPs with different sizes were also prepared by adjusting the amount of ethanol.

4.3. Preparation of supramolecular hydrogels

To synthesize PAV hydrogels, AAm and VI were first dissolved in ultrapure water to a final molar ratio of 5 : 1. The total monomer concentration of the hydrogel is controlled at 2 M. Crosslinker MBAA (0.5–2 mol% of monomer) and Darocur 2959 (1 wt% of the solution) were also added, followed by polymerization via UV light irradiation for 10 min. For PAV-Zn, the as-prepared PAV hydrogel was soaked into ZnCl₂ solution (50 mM) for 30 min at room temperature to form the second network. The hydrogels prepared with different crosslinker amounts were used to change the mechanical sensing sensitivity. Unless otherwise stated, the cross-linking degree of all experiments and characterized hydrogels in this article is 0.5%.

4.4. Fabrication of supramolecular photonic hydrogels

The vortex oscillation method was used to obtain non-close-packed photonic crystals. 200–300 mg of monodisperse SiO₂ NPs with a particle size of about 140 nm were dispersed in water or prepolymer solution so that the total mass of the suspension was 1 g (i.e., the mass fraction of SiO₂ NPs was 20–30 wt%). Finally, appropriate ion exchange resin was added to it and vortexed in a vortexer for 2 h to obtain a non-close-packed photonic crystal with a structural color. Two pieces of clean, hydrophilic glass serve as templates, with 300 μm space in between. Then, the precursor solution of the hydrogel was filled into the space between two glass slides due to capillary force. After UV (365 nm) irradiation for 10 min, the two glass slides were carefully separated and the photonic hydrogel was peeled off to make them freestanding. Finally, the photonic hydrogel was soaked in ZnCl₂ solution (50 mM, 30 min) to obtain a photonic hydrogel dressing.

4.5. Characterization

The prepared hydrogel or structural color hydrogel was stored in PBS solution, and all tests were performed after its swelling equilibrium. The scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) images of hydrogels were obtained using a Apreo S HiVoc microscope (Thermo Fisher Scientific). Fourier-transform infrared (FT-IR) spectroscopy was performed using a Nicolet is50 (Thermo Fisher Scientific). X-ray photoelectron spectroscopy (XPS) was performed using an ESCALAB 250Xi. (Thermo Fisher Scientific). Stress–strain tests were performed using TA.XTC-20 (BosinTech). Inductively coupled plasma optical emission spectrometry (ICP-OES) was used to measure the concentration of Zn²⁺ in the hydrogel dispersion (Element XR, Thermo Fisher Scientific). The reflectance spectra were recorded using a fiber spectrometer (FX2000). The optical images were obtained using a cell phone camera. The RGB values of the images were analyzed using ImageJ (Fiji). A non-contact full-field strain measurement system (DIC) (VIC-3D) was used to determine the Poisson's ratio of the hydrogel.

4.6. Cell compatibility studies

Mouse fibroblasts (L929) were used to verify the cytotoxicity of the material. The L929 cells were cultured in an α-MEM medium containing a 1% penicillin–streptomycin solution and 10% FBS. L929 cells were seeded at a density of 5000 cells per well into 96 wells and cultured for 24 hours to allow cells to adhere to the wall. The cell culture medium was then replaced with 12.5 mg mL⁻¹ hydrogel dispersion and cultured in a 37 °C, 5% CO₂ incubator for 24 h, 48 h, and 72 h. Then, the old culture medium was replaced with fresh culture medium containing CCK-8 solution. After 2 h of reaction, the media were transferred to wells of a 96-well plate, and the absorbance at 450 nm was measured with a microplate reader (KHB ST-360, China). In addition, L929 cells were cocultured with 12.5 mg mL⁻¹ hydrogel dispersion for 24 h, 48 h, and 72 h. Live-dead cell staining was then carried out using fluorescein diacetate (FDA) and propidium iodide (PI) staining solution, respectively. Fluorescence images were captured with a fluorescence microscope (IX-71, Olympus, Japan).

4.7. In vitro hemolytic test

Fine pieces of PAV-Zn were dispersed in 0.01 M phosphate buffered saline (PBS) to a final concentration of up to 12.5 mg mL⁻¹. To every 1 mL of the above dispersion, 20 μL of mouse whole blood was added. Three replicates were prepared for each concentration studied. After thorough mixing, the samples were incubated at 37 °C for 1 h. Blood samples treated with ultrapure water and PBS were used as positive and negative controls, respectively. At the end of the incubation, these samples were centrifuged at 1000 rpm for 5 min. Finally, the absorbance of supernatants was measured at 540 nm using a microplate reader. The hemolysis ratio was calculated using the following equation:

Hemolysis ratio (%) = [(OD_S − OD_P)/(OD_W − OD_P)] × 100%

where OD_S is the absorbance of supernatant from the PAV-Zn test group, and OD_W and OD_P correspond to the absorbance of water and PBS treated controls, respectively.

4.8. In vitro antibacterial assay

The antibacterial activity of PAV and PAV-Zn against Staphylococcus aureus (S. aureus) and Escherichia coli (E. coli) was investigated in vitro using the spread plate method. Typically, hydrogel pieces with a diameter of 1 cm were washed with PBS three times. After 30 min of sterilization with a UV lamp, they were immersed in a bacterial suspension (1 mL, 1.0 × 10⁶ CFU mL⁻¹). After co-incubating at 37 °C for 24 h, the bacterial suspension of each group was diluted 1.0 × 10⁴ times with PBS. Then, 100 μL of the diluted bacterial solution was uniformly spread onto fresh Luria Bertani broth agar plates. The bacterial suspension treated with PBS was considered the control group. After overnight culture, the bacterial colonies on the agar plates were photographed and counted using ImageJ. In addition, the collected bacteria of each group were washed with PBS and then fixed with glutaraldehyde at 4 °C, washed with PBS, and completely air-dried after gradient dehydration with alcohol. After Au coating to increase conductivity, the morphology of bacteria was examined using SEM.

4.9. In vivo experiments

Male Kunming mice weighing 30–50 g were randomly divided into 3 groups. The backs of the mice were shaved, and a 0.8-cm-diameter whole layer of skin was cut on the back of the mice to create a skin wound model, and then S. aureus (20 μL, 1 × 10⁹ CFU mL⁻¹) was added to the circular wound bed. Control mice were treated with PBS solution only. In the patch group, a 1 cm diameter patch was applied to the wound. Wound changes were measured and photographed on days 0, 3, 6, and 10 and counted using ImageJ.

The wound area (%) was calculated from the equation below:

Wound healing (%) = [(A₀ − A_t)/A₀] × 100%

where A₀ is the initial area and A_t is the wound area obtained on day n (n = 0, 3, 6, and 10).

All mice were executed after 10 days. Organ and wound bed tissues were harvested, then treated with 4% paraformaldehyde, embedded in paraffin and sectioned. Subsequently, histological analyses were performed including hematoxylin–eosin staining (H&E), Masson staining and immunofluorescence staining for IL-6 and CD31. Animal protocols were approved by the Animal Ethics Committee of Chengdu Dossy Experimental Animals Co., Ltd, China (No. IACUC-SWLAB-2024-03-15-01).

4.10. Statistical analyses

Data were presented as mean ± standard deviation and the sample size (n) for each statistical analysis is provided in the figure legend. Student's t-test was applied to assess significant differences. *p < 0.05, **p < 0.01, ***p < 0.001, and n.s.: not significant. Origin 2020 and ImageJ were applied for statistical analysis.

Data availability

Data are available from the authors upon request.

Conflicts of interest

The authors declare no conflicts of interest.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant No. 52103176 and U22A20158), the Natural Science Foundation of Sichuan Province (2024NSFSC0241), the Chengdu Science and Technology Program (2024-YF05-00530-SN), and the Aier Eye Hospital-Sichuan University Research Fund (No. 23JZH045).

References

1. J. C. Flood, BMJ [Br. Med. J.], 1940, 1, 689–690.
2. M. J. Simons, N. H. Amin and G. R. Scuderi, J. Am. Acad. Orthop. Surg., 2017, 25, 547–555.
3. R. J. Nudo, B. M. Wise, F. SiFuentes and G. W. Milliken, Science, 1996, 272, 1791–1794.
4. R. L. Gajdosik and R. W. Bohannon, Phys. Ther., 1987, 67, 1867–1872.
5. Y. Luo, M. R. Abidian, J.-H. Ahn, D. Akinwande, A. M. Andrews, M. Antonietti, Z. Bao, M. Berggren, C. A. Berkey, C. J. Bettinger, J. Chen, P. Chen, W. Cheng, X. Cheng, S.-J. Choi, A. Chortos, C. Dagdeviren, R. H. Dauskardt, C.-A. Di, M. D. Dickey, X. Duan, A. Facchetti, Z. Fan, Y. Fang, J. Feng, X. Feng, H. Gao, W. Gao, X. Gong, C. F. Guo, X. Guo, M. C. Hartel, Z. He, J. S. Ho, Y. Hu, Q. Huang, Y. Huang, F. Huo, M. M. Hussain, A. Javey, U. Jeong, C. Jiang, X. Jiang, J. Kang, D. Karnaushenko, A. Khademhosseini, D.-H. Kim, I.-D. Kim, D. Kireev, L. Kong, C. Lee, N.-E. Lee, P. S. Lee, T.-W. Lee, F. Li, J. Li, C. Liang, C. T. Lim, Y. Lin, D. J. Lipomi, J. Liu, K. Liu, N. Liu, R. Liu, Y. Liu, Y. Liu, Z. Liu, Z. Liu, X. J. Loh, N. Lu, Z. Lv, S. Magdassi, G. G. Malliaras, N. Matsuhisa, A. Nathan, S. Niu, J. Pan, C. Pang, Q. Pei, H. Peng, D. Qi, H. Ren, J. A. Rogers, A. Rowe, O. G. Schmidt, T. Sekitani, D.-G. Seo, G. Shen, X. Sheng, Q. Shi, T. Someya, Y. Song, E. Stavrinidou, M. Su, X. Sun, K. Takei, X.-M. Tao, B. C. K. Tee, A. V.-Y. Thean, T. Q. Trung, C. Wan, H. Wang, J. Wang, M. Wang, S. Wang, T. Wang, Z. L. Wang, P. S. Weiss, H. Wen, S. Xu, T. Xu, H. Yan, X. Yan, H. Yang, L. Yang, S. Yang, L. Yin, C. Yu, G. Yu, J. Yu, S.-H. Yu, X. Yu, E. Zamburg, H. Zhang, X. Zhang, X. Zhang, X. Zhang, Y. Zhang, Y. Zhang, S. Zhao, X. Zhao, Y. Zheng, Y.-Q. Zheng, Z. Zheng, T. Zhou, B. Zhu, M. Zhu, R. Zhu, Y. Zhu, Y. Zhu, G. Zou and X. Chen, ACS Nano, 2023, 17, 5211–5295.
6. K.-Y. Chun, S. Seo and C.-S. Han, Adv. Mater., 2022, 34, 2110082.
7. H. Wang, Q. Ding, Y. Luo, Z. Wu, J. Yu, H. Chen, Y. Zhou, H. Zhang, K. Tao, X. Chen, J. Fu and J. Wu, Adv. Mater., 2023, 36, 2309868.
8. C. Wang, X. Qu, Q. Zheng, Y. Liu, P. Tan, B. Shi, H. Ouyang, S. Chao, Y. Zou, C. Zhao, Z. Liu, Y. Li and Z. Li, ACS Nano, 2021, 15, 10130–10140.
9. Y. Q. Liang, Z. L. Li, Y. Huang, R. Yu and B. L. Guo, ACS Nano, 2021, 15, 7078–7093.
10. X. N. Sun, C. M. Ding, M. Qin and J. S. Li, Small, 2024, 20, 2306960.
11. N. Xu, Y. Yuan, L. Ding, J. Li, J. Jia, Z. Li, D. He and Y. Yu, Burns Trauma, 2022, 10, tkac019.
12. Q. Lei, D. He, L. Ding, F. Kong, P. He, J. Huang, J. Guo, C. J. Brinker, G. Luo, W. Zhu and Y. Yu, Adv. Funct. Mater., 2022, 32, 2113269.
13. L. Wang, L. Y. Sun, F. K. Bian, Y. Wang and Y. J. Zhao, ACS Nano, 2022, 16, 2640–2650.
14. Y. Wang, L. Y. Sun, G. P. Chen, H. X. Chen and Y. J. Zhao, ACS Nano, 2023, 17, 1437–1447.
15. Q. Zhou, K. Y. Dong, M. Wei, B. F. He and B. B. Gao, Adv. Funct. Mater., 2024, 34, 2316565.
16. Z. X. Shu, X. N. Sun, X. Y. Xu, M. Qin and J. S. Li, J. Mater. Chem. B, 2024, 12, 8488–8504.
17. L. Wang, X. Y. Ding, L. Fan, A. M. Filppula, Q. Y. Li, H. B. Zhang, Y. J. Zhao and L. R. Shang, Nano-Micro Lett., 2024, 16, 232.
18. H. Liu, X. Hu, W. Li, M. Zhu, J. Tian, L. Li, B. Luo, C. Zhou and L. Lu, Chem. Eng. J., 2023, 452, 139368.
19. Y. G. Kim, S. Park, Y. H. Choi, S. H. Han and S.-H. Kim, ACS Nano, 2021, 15, 12438–12448.
20. J. B. Kim, C. Chae, S. H. Han, S. Y. Lee and S.-H. Kim, Sci. Adv., 2021, 7, eabj8780.
21. C. E. Finlayson, P. Spahn, D. R. E. Snoswell, G. Yates, A. Kontogeorgos, A. I. Haines, G. P. Hellmann and J. J. Baumberg, Adv. Mater., 2011, 23, 1540–1544.
22. Y. Wang, X. Cao, J. Cheng, B. Yao, Y. Zhao, S. Wu, B. Ju, S. Zhang, X. He and W. Niu, ACS Nano, 2021, 15, 3509–3521.
23. R. Lin, D. Kou, L. Gao, S. Li, Z. Gao, X. Li, W. Ma and S. Zhang, ACS Nano, 2024, 18, 13346–13360.
24. Y. Liang, J. He and B. Guo, ACS Nano, 2021, 15, 12687–12722.
25. Y. Zhao, Z. Li, S. Song, K. Yang, H. Liu, Z. Yang, J. Wang, B. Yang and Q. Lin, Adv. Funct. Mater., 2019, 29, 1901474.
26. M. Li, Y. Zhang, L. Lian, K. Liu, M. Lu, Y. Chen, L. Zhang, X. Zhang and P. Wan, Adv. Funct. Mater., 2022, 32, 2208141.
27. Q. Xu, Z. Zheng, B. Wang, H. Mao and F. Yan, ACS Appl. Mater. Interfaces, 2017, 9, 14656–14664.
28. T. Xie, J. Ding, X. Han, H. Jia, Y. Yang, S. Liang, W. Wang, W. Liu and W. Wang, Mater. Horiz., 2020, 7, 605–614.
29. X. Luo, S. H. Goh, S. Y. Lee and C. H. A. Huan, Macromol. Chem. Phys., 1999, 200, 874–880.
30. B. Tao, C. Lin, X. Qin, Y. Yu, A. Guo, K. Li, H. Tian, W. Yi, D. Lei, Y. Chen and L. Chen, Mater. Today Bio, 2022, 13, 100216.
31. F. Yang, Y. Xue, F. Wang, D. Guo, Y. He, X. Zhao, F. Yan, Y. Xu, D. Xia and Y. Liu, Bioact. Mater., 2023, 26, 88–101.
32. S. Huo, S. Liu, Q. Liu, E. Xie, L. Miao, X. Meng, Z. Xu, C. Zhou, X. Liu and G. Xu, Adv. Sci., 2024, 11, 2302674.
33. S. N. Riduan and Y. G. Zhang, Chem. Soc. Rev., 2013, 42, 9055–9070.
34. D. Demberelnyamba, K.-S. Kim, S. Choi, S.-Y. Park, H. Lee, C.-J. Kim and I. D. Yoo, Bioorg. Med. Chem., 2004, 12, 853–857.
35. S. A. S. Chatha, M. Asgher, R. Asgher, A. I. Hussain, Y. Iqbal, S. M. Hussain, M. Bilal, F. Saleem and H. M. N. Iqbal, Sci. Total Environ., 2019, 690, 667–682.
36. P. Lin, M. Sermersheim, H. Li, P. H. U. Lee, S. M. Steinberg and J. Ma, Nutrients, 2018, 10, 16.
37. Y. Wang, T. Ying, J. Li, Y. Xu, R. Wang, Q. Ke, S. G. F. Shen, H. Xu and K. Lin, Chem. Eng. J., 2020, 402, 126273.
38. Y. Su, H. Yang, J. Gao, Y.-X. Qin, Y. Zheng and D. Zhu, Adv. Sci., 2019, 6, 1900112.
39. Y. F. Zheng, X. N. Gu and F. Witte, Mater. Sci. Eng., R, 2014, 77, 1–34.
40. Y. Gong, P. Wang, R. Cao, J. Wu, H. Ji, M. Wang, C. Hu, P. Huang and X. Wang, ACS Nano, 2023, 17, 22355–22370.
41. K. Kim, J. Yang, C. Li, C. Y. Yang, P. Hu, Y. Liu, Y. y Huang, X. Sun, M. Chi, C. Huang, X. Sun, L. Zhao and X. Wang, Bioact. Mater., 2025, 47, 64–82.
42. M. Qin, J. S. Li and Y. L. Song, Anal. Chem., 2022, 94, 9497–9507.
43. D. H. Kim, J. H. Cha, J. Y. Lim, J. Bae, W. Lee, K. R. Yoon, C. Kim, J. S. Jang, W. Hwang and I. D. Kim, ACS Nano, 2020, 14, 16907–16918.
44. N. Xu, Y. Gao, Z. Li, Y. Chen, M. Liu, J. Jia, R. Zeng, G. Luo, J. Li and Y. Yu, Chem. Eng. J., 2023, 466, 143173.
45. Y. Yu, P. Li, C. Zhu, N. Ning, S. Zhang and G. J. Vancso, Adv. Funct. Mater., 2019, 29, 1904402.
46. J. Yang, K. Yang, X. An, Z. Fan, Y. Li, L. Yin, Y. Long, G. Pan, H. Liu and Y. Ni, Adv. Funct. Mater., 2025, 2416916.
47. H. Liu, Y. Wang, Z. Shi, D. Tan, X. Yang, L. Xiong, G. Li, Y. Lei and L. Xue, Small Methods, 2022, 6, 2200461.
48. Y. Liu, J. Liu, L. Ai, Q. Wei, Q. Liu and S. Xie, IEEE ASME Int. Conf. Adv. Intellig. Mechatron. AIM, 2019, 441–446.
49. W. Stöber, A. Fink and E. Bohn, J. Colloid Interface Sci., 1968, 26, 62–69.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4tb02673j

‡ Xiaoning Sun and Dengfeng Lu contributed equally to this work.

---