Poly(acrylic acid)/Boron Nitride Composite Hydrogels with Superior Mechanics and Rapid Self‑Healing via Hierarchical Physical Interactions

Background

Hydrogels, prized for their water content, biocompatibility, and ease of fabrication, often suffer from limited mechanical strength, restricting their practical use. Nature offers examples—muscle, ligament, and skin—that combine high toughness with self‑repair, inspiring a new generation of smart hydrogels. Prior efforts have leveraged dynamic covalent bonds, supramolecular networks, and inorganic fillers to enhance mechanical properties, yet many still lack the balance of strength and rapid healing.

Inorganic nanofillers such as graphene oxide, montmorillonite, and boron nitride nanosheets (BNNS) have been shown to reinforce hydrogels by providing additional crosslinking points and reinforcing domains. BNNS, often called “white graphene,” offers exceptional mechanical stiffness, chemical inertness, and low toxicity, making it an attractive candidate for hydrogel reinforcement. Surface modification of BNNS with amino groups (BNNS‑NH₂) further introduces hydrogen‑bonding capability with polymer chains, potentially improving both mechanical robustness and self‑healing.

Here, we present a straightforward in‑situ polymerization strategy that combines PAA with BNNS‑NH₂ and Fe³⁺ ions to form a three‑dimensional network governed by hierarchical physical interactions. This approach yields hydrogels with unprecedented toughness, strength, and rapid self‑healing.

Method/Experimental

Materials

Potassium persulfate (KPS, 99 %) and FeCl₃·6H₂O (99 %) were sourced from J&K Chemical Technology; acrylic acid (AA, 98 %) and Rhodamine B (95 %) were obtained from Sigma‑Aldrich. BNNS‑NH₂ was prepared via our previously reported protocol. Deionized water was used throughout.

BNNS‑NH₂ Dispersion Preparation

Stable BNNS‑NH₂ suspensions were achieved by stirring 100 mg, 80 mg, 50 mg, or 10 mg of BNNS‑NH₂ in 100 mL deionized water (concentrations 1.0, 0.8, 0.5, and 0.1 mg m⁻¹, respectively). The mixtures were magnetically stirred at 1000 rpm for 24 h, followed by 2 h sonication (20 kHz). The resulting dispersions were sealed for subsequent hydrogel synthesis.

Self‑Healing Hydrogel Synthesis

Hydrogels were formed by free‑radical polymerization of AA (10 mL) in the presence of FeCl₃·6H₂O (0.25 g, 1.05 mol % AA) and KPS (0.1 g, 0.25 mol % AA) within the BNNS‑NH₂ dispersions or deionized water. After 10 min stirring, N₂ purging removed dissolved oxygen, and the mixture was heated to 25 °C for 6 h to complete polymerization. Hydrogels derived from BNNS‑NH₂ concentrations of 1.0, 0.8, 0.5, and 0.1 mg m⁻¹ were designated B₁P₉₀, B₀.₈P₉₀, B₀.₅P₉₀, and B₀.₁P₉₀, respectively; the control hydrogel (no BNNS‑NH₂) was B₀P₉₀.

By controlled drying, hydrogels with lower water contents (70 wt %, 50 wt %, 25 wt %) were obtained and labeled BₓP₇₀, BₓP₅₀, and BₓP₂₅, where x indicates BNNS‑NH₂ concentration. Cross‑linking densities were inferred from rheological data; mechanical testing required higher‑density samples due to the softness of the BₓP₉₀ series.

Mechanical Testing

Specimens (50 mm × 2 mm × 2 mm) were tested on a universal testing machine (200 N load cell) at 50 mm min⁻¹, 25 °C, 45 % RH. Stress (σ) was calculated as F/(a × b) and strain (ε) as (l − l₀)/l₀ × 100 %. Young’s modulus was derived from the initial stress–strain slope; toughness was the area under the stress–strain curve. Cyclic loading assessed dissipated energy via the hysteresis area.

Characterization

Fourier‑transform infrared (FTIR) spectra (ATR mode, 400–4000 cm⁻¹) identified key functional group shifts. Scanning electron microscopy (SEM) examined dried morphology. Rheology (HAAKE MARS III) measured storage (G′) and loss (G″) moduli to estimate cross‑link density (N = Gₑ/RT). Tensile data were collected on an Instron 2360.

Results and Discussion

During in‑situ polymerization, AA chains were cross‑linked by two hierarchical interactions: (1) Fe³⁺‑mediated carboxyl coordination at the molecular scale, and (2) hydrogen bonding between –COOH of PAA and –NH₂ of BNNS‑NH₂ at the nanoscale. This dual network architecture is illustrated in Scheme 1.

Scheme 1. Hierarchical assembly of the PAA/BNNS‑NH₂ composite hydrogel.

FTIR analysis confirmed the formation of the two interaction modes. The characteristic C=O stretch of PAA shifted from 1690 cm⁻¹ to 1620 cm⁻¹ in the composite, indicating hydrogen bonding. A new peak at 620 cm⁻¹ appeared, confirming Fe³⁺‑carboxyl coordination. BNNS‑NH₂ signals (1388 cm⁻¹ B–N stretch, 1780 cm⁻¹ B‑N‑B bend) were also present.

Fig. 1. FTIR spectra of (a) composite hydrogel and (b) PAA hydrogel.

SEM images revealed porous networks with pore sizes ranging from several nanometers to tens of micrometers, providing both flexibility and structural integrity. BNNS‑NH₂ nanosheets were clearly visible within the composite matrix.

Fig. 2. SEM images of (a) PAA hydrogel, (b) composite hydrogel, and (c) BNNS‑NH₂ nanosheets.

Mechanical testing highlighted the decisive role of BNNS‑NH₂. Without BNNS‑NH₂ (B₀P₇₀), fracture stress was ~406 kPa. Introducing 0.5 mg m⁻¹ BNNS‑NH₂ (B₀.₅P₇₀) amplified fracture stress to ~1311 kPa—threefold higher—while toughness rose from ~0.5 MJ m⁻³ to ~4.7 MJ m⁻³. Excessive BNNS‑NH₂ (>0.5 mg m⁻¹) reduced performance due to nanosheet aggregation. Lower water content further enhanced strength and toughness, with B₀.₅P₂₅ achieving a Young’s modulus of ~17.9 MPa, tensile strength of ~8.5 MPa, and toughness of ~10.5 MJ m⁻³.

Fig. 3. Stress–strain curves for BₓP₇₀, BₓP₅₀, and BₓP₂₅; photomicrographs of (d) intact, (e) stretched, (f) bent, and (g) knotted hydrogels.

Rheological data corroborated the mechanical findings. Storage modulus (G′) exceeded loss modulus (G″) across 0.1–100 rad s⁻¹, confirming network integrity. G′ peaked for B₀.₅P₉₀, aligning with its superior mechanical metrics. Calculated cross‑link density mirrored this trend, peaking at BNNS‑NH₂ of 0.5 mg m⁻¹ before declining due to aggregation.

Fig. 4. (a) Frequency dependence of G′ and G″ for BₓP₉₀; (b) Cross‑link density versus BNNS‑NH₂ concentration.

Cyclic tensile tests showed pronounced hysteresis, especially for B₀.₅P₅₀, indicating efficient energy dissipation via reversible hydrogen bonds. Dissipated energy peaked at 0.5 mg m⁻¹ BNNS‑NH₂ and decreased at higher loadings due to nanosheet clustering.

Fig. 6. (a) Loading–unloading curves for BₓP₇₀; (b) for BₓP₅₀; (c) dissipated energy of BₓP₇₀; (d) dissipated energy of BₓP₅₀.

Self‑healing was rapid and efficient. Cut B₀.₅P₉₀ healed in 10 min at room temperature, restoring ~81 % of its original fracture stress. Hydrogels with higher BNNS‑NH₂ content exhibited superior healing (up to 94 % for B₁P₅₀), while lower water content slowed healing due to reduced ion mobility.

Fig. 7. Self‑healing process of B₀.₅P₉₀: (a) cut specimen, (b) reassembled, (c) post‑healing, (d) fracture stress comparison for BₓP₇₀, (e) for BₓP₅₀.

Overall, the hierarchical network—combining Fe³⁺‑carboxyl coordination and PAA/BNNS‑NH₂ hydrogen bonding—enables a rare synergy of high stiffness and toughness while maintaining rapid, stimulus‑free self‑repair.

Conclusions

We have introduced a simple, scalable route to PAA/BNNS‑NH₂ composite hydrogels that integrate metal‑coordination and hydrogen‑bonding networks. These materials achieve a Young’s modulus of ~17.9 MPa, toughness of ~10.5 MJ m⁻³, and self‑healing efficiencies exceeding 80 % within 10 min—without external stimuli. The dual‑interaction design offers a versatile platform for high‑performance soft materials in biomedicine, sensing, and actuating applications.

Abbreviations

B_xP_y

Composite hydrogel with BNNS‑NH₂ concentration of x mg m⁻¹ and water content of y wt %

Fe³⁺

Ferric ion

FTIR

Fourier‑transform infrared spectroscopy

PAA/BNNS‑NH₂

Poly(acrylic acid)/surface‑modified boron nitride nanosheets

SEM

Scanning electron microscopy