Hybrid nanocomposite hydrogels with high strength and excellent self-recovery performance

Abstract

Polymer nanocomposite hydrogels (NC gels) have been intensively investigated for applications in academic and industrial fields. In the present work, hybrid cross-linked polyacrylamide (PAAm)/LAPONITE® hydrogels were synthesized by in situ free radical polymerization using LAPONITE® nanoplatelets and N,N-methylenebisacrylamide (MBA) as physical and chemical crosslinkers, respectively. We aimed to make clear the synergistic effect of physical and chemical crosslinking on the mechanical properties, self-recovery, fatigue resistance and self-healing performances of hybrid PAAm/LAPONITE® NC gels. The effects of LAPONITE® and chemical cross-linker concentration on the mechanical properties of the physical and hybrid NC gels were investigated in details. It was found that the fracture stress and fracture energies of hybrid NC gels decreased sharply as the chemical crosslinker (MBA) concentration became larger than 0.05 mol% of monomer. With a very small amount of chemical crosslinker, the hybrid NC gels exhibited improved tensile properties, large hysteresis and good self-recovery performances. The best tensile properties of hybrid NC gels in this study were σ_f of 0.45 MPa, ε_f of 50.63 mm mm⁻¹, E of 72.64 kPa and W of 9.81 MJ m⁻³, which were much better than those of physical NC gels and pure chemical gels. However, it was also revealed that there was a negative effect from chemical crosslinking on the fatigue resistance and self-healing properties of NC gels. A deformation and recovery mechanism was also proposed. We believe this work will provide better understanding about the relationship of network structure and properties for hybrid NC gels.

---

1. Introduction

Polymer hydrogels, which are three-dimensional polymer networks with a large amount of water, have attracted much attention owing to their potential applications in various fields including tissue engineering,^1–3 drug delivery,^4,5 sensors & actuators^6–8 and superabsorbents.^9,10 However, conventional hydrogels are often chemical crosslinked network structures and exhibit extremely brittle or weak mechanical properties, which is one of the biggest drawbacks for their applications in load-bearing fields. The poor mechanical properties of conventional hydrogels are mainly attributed to their heterogeneous network structure and lack of effective energy dissipation mechanisms.¹¹ In recent years, significant efforts have been made to design high strength and tough hydrogels with novel microstructures and energy dissipation mechanisms including nanocomposite hydrogels (NC gels),¹² double network hydrogels,^13,14 hydrophobically associated hydrogels,^15,16 ionically cross-linked hydrogels,¹⁷ hydrogen bonding or dipole–dipole enhanced hydrogels^18–20 and macromolecular microparticle composite hydrogels.²¹

Polymer/clay hydrogels are important NC gels crosslinked by clay nanoplatelets (e.g. LAPONITE® platelets) that have been extensively studied due to their extraordinary mechanical, optical and swelling/de-swelling properties.^22–24 The excellent mechanical properties, network structure and deformation mechanisms of NC gels have been characterized by various technologies such as tensile tests,^20,21 transmission electron microscopy,²³ small angle X-ray scattering (SAXS) and small angle neutron scattering (SANS).²⁴ It was well accepted that the clay nanoplatelets were uniformly dispersed in the NC gels and acted as multifunctional crosslinkers in the formation of polymer/clay networks. The hydrogen bonding and electrostatic interactions between polymer chains and clay platelets are the main driving force for these unique structures.^20,21 The measurements of SAXS and SANS revealed that there was a process of peeling-off of adsorbed polymer chains on the clay platelets during the deformation of NC gels while the polymer chains tended to be re-adsorbed to the surface of clay platelets after the removal of the external force.²⁴ The peeling-off and re-adsorption phenomena resulted in energy dissipation and hysteresis during loading–unloading of the gel specimen. The self-healing properties and mechanism of the NC gels were attributed to the reconstruction of the interactions between polymer chains and clay platelets at the interface of the cut surfaces.²⁵ Although the energy dissipation and self-healing of NC gels have been investigated in depth, the results are mainly based on physically linked NC gels. Little is known about the energy dissipation and recovery mechanisms of hybrid physically and chemically crosslinked NC gels.

Recently, hybrid physical and chemical crosslinking in the hydrogel network has emerged as an effective strategy to design high strength hydrogels. Tuncaboylu et al.²⁶ found that the compressive mechanical properties of hybrid hydrophobically associated hydrogels (HA gels) were significantly improved in the presence of a very small amount of chemical crosslinker N,N-methylenebisacrylamide (MBA), while the self-healing properties of these HA gels were not largely influenced. Recently, hybrid crosslinkings in the NC gels have been emerged as one of effective strategies to design high strength hydrogels.^27–30 Li et al.²⁷ reported PAAm/Na-montmorillonite NC gels in the presence of 1 mol% of MBA could achieve the maximum fracture stress of 150 kPa and elongation of ∼1300%, respectively. Shen et al.²⁸ found PAAm/graphene oxide NC gels with 1 mol% MBA showed a fracture stress of ∼28 kPa and elongation of ∼300%. Rose et al.²⁹ prepared hybrid crosslinked poly(N,N-dimethylacrylamide)/silica NC gels and found that the elastic modulus of the NC gels increased with the increase of MBA concentration; nevertheless, the elongation decreased rapidly. The above reports showed that the elongation of these NC gels was often less than 1500%, and the addition of chemical crosslinker often caused negative effects on the mechanical properties.^27–29 One of reasons may be attributed to the addition of too much chemical crosslinker (>0.1 mol%), which resulted in a heterogeneous network within the NC gels. When the gels are stretched, the stress concentration at the shortest chains of the gels leads to chain breakage, followed by catastrophic and irreversible fracture of the whole gel. Xiong et al.³⁰ found the NC gel prepared by in situ copolymerization of sodium acrylate and acrylamide (AAm) in an aqueous LAPONITE® platelets suspension with a very small amount of MBA (0.03 mol%) exhibited the maximum tensile stress and strain. The values decreased sharply when the MBA concentration was more than 0.03 mol%, which indicated the balance of physical and chemical crosslinking can achieve better mechanical properties in the NC gels. However, the energy dissipation, self-recovery and self-healing of the hybrid NC gels were not characterized.

In the present study, hybrid PAAm/LAPONITE® NC gels with a very small amount of chemical crosslinker (MBA < 0.1 mol%) were synthesized by in situ free radical polymerization of AAm in an aqueous LAPONITE® suspension with MBA. We aimed to make clear the synergistic effect of physical and chemical crosslinking on the mechanical properties, self-recovery, fatigue resistance and self-healing performances of hybrid PAAm/LAPONITE® NC gels. A series of tests (e.g., uniaxial tensile, swelling, cyclic loading and self-healing) were systematically conducted to reveal the effect of LAPONITE® and MBA concentrations on these properties. Compared to the physical NC gels, it was found that the hybrid PAAm/LAPONITE® NC gels with loose chemical crosslinking exhibited high mechanical strength and good self-recovery properties. The best tensile properties of the hybrid NC gel in this study were the fracture stress of 0.45 MPa, elongation of 50.63 mm mm⁻¹, elastic modulus of 72.64 kPa and fracture energy of 9.81 MJ m⁻³, which were much better than those of a physical NC gel and a pure chemical gel. However, it was also revealed that there was a negative effect from chemical crosslinking on the fatigue resistance and self-healing performances of the NC gels. As a complement to physical NC gels, we believe this work will provide better understanding about the relationship of network structure and properties for hybrid NC gels.

2. Materials and methods

2.1. Materials

All chemicals and solvents purchased were of the highest available purity and, unless otherwise stated, they were used as received. Monomer acrylamide (AAm, 98%) was purchased from TCI Shanghai Co. Ltd. Synthetic hectorite clay of sol-forming grade LAPONITE® RDS, as physical cross-linker, modified from gel-forming grade LAPONITE® RD with pyrophosphate ions (P₂O₇⁴⁻), was kindly provided by Rockwood Co. Ltd. N,N-Methylenebisacrylamide (MBA), as a chemical crosslinker, and N,N,N′,N′-tetramethylethylenediamine (TEMED), as an accelerator, were purchased from Aladdin Shanghai Co. Ltd. Potassium persulfate (KPS), as an initiator, was obtained from Shanghai Chemical Reagent Co. Ltd. Deionized water was used in all polymerization reactions.

2.2. Preparation of PAAm/LAPONITE® NC gels

The PAAm/LAPONITE® NC gels were synthesized through in situ free radical polymerization of acrylamide in the aqueous suspension of LAPONITE® and MBA. The LAPONITE® suspension was prepared by dispersing LAPONITE® in pure water (10 mL) at the desired concentration under stirring for at least 8 h. The monomer AAm, catalyst TEMED and the aqueous solution of initiator KPS were subsequently added to the LAPONITE® suspension under stirring. The free radical polymerization was allowed to proceed in a water bath at 30 °C for 72 h. The NC gels were formed in glass tubes for measurements.

For all hydrogel samples, the weight of the monomer AAm was 2 g and the mole ratio of monomer to initiator to catalyst was kept at 100 : 0.263 : 0.453.³¹ The mole ratio of MBA to monomer was varied from 0 to 0.1 mol%, and the LAPONITE® concentration was changed from 0 to 10 w/v% in water. In order to simplify the discussion, the gel samples were designated as LxMy, where L and M stood for LAPONITE® and MBA, respectively; x and y stood for 100 × LAPONITE®/water (w/v) and 100 × MBA/AAm (mol mol⁻¹), respectively. For example, L4M5 means the gel was prepared in 4 w/v% aqueous LAPONITE® suspension and MBA/AAm of 0.05 mol%.

2.3. Mechanical tests

2.3.1 Uniaxial tensile tests. Uniaxial tensile tests of the as-prepared gels (diameter of 8 mm and length of 70 mm) were carried out with a commercial testing machine with a 100 N load cell with a crosshead speed of 100 mm min⁻¹. Elongation (ε_f) was determined as ε_f = Δl/l₀ (mm mm⁻¹), where Δl is the difference of the elongation length (l) and the initial length (l₀). The fracture stress (σ_f) was defined as σ_f = F/A₀, where F is the load force and A₀ is the original specimen cross-sectional area. Elastic modulus (E) of gels was calculated by fitting the initial linear regime of the stress–strain curve. The means and standard deviations of the above data were calculated for each treatment. Statistical analysis was performed using the one-way analysis of variance (ANOVA) and a LSD's multiple comparison post-test with 95% confidence interval among the applied treatments. Error ranges were defined as the standard deviation from the data of three samples.

2.3.2 Cyclic, self-recovery and fatigue resistance tests. For loading–unloading measurements, the gel specimens were first stretched to a maximum extension ratio (λ = 10) at a crosshead speed of 100 mm min⁻¹ and then unloaded at the same speed. The dissipated energies were estimated by the area between the loading–unloading curves. For self-recovery tests, the gel specimen was first performed with one cyclic loading as mentioned above, and then the same specimen was allowed to recover at room temperature at various times. Once the predetermined recovery time was reached, the same specimen was conducted through another cyclic loading with the same process as mentioned above. For fatigue resistance test, the same gel specimen was tested with six cyclic loadings with the same process as mentioned above. In the self-recovery and fatigue resistance tests, the recovery (%) was determined by the ratio between dissipated energies at various recovery times and dissipated energies at the first cyclic loading.

2.3.3 Self-healing tests. For self-healing tests, the as-prepared cylindrical PAAm/LAPONITE® NC gels were first cut into two pieces completely, and then the cut surfaces were kept in close contact to heal for 24 h at 30, 50 and 80 °C. After healing, the uniaxial tensile test was conducted at a crosshead speed of 100 mm min⁻¹.

2.4. Swelling tests

For swelling tests, the as-prepared cylindrical gels were soaked in a large amount of deionized water at room temperature for 10 days. The swollen gel was removed from water and wiped with filter paper to remove water on the surface of the gel. The swelling ratio (SR) was determined by the following equation:where W_s and W_d were the weights of the swollen and corresponding dried hydrogels, respectively. The swelling ratio after 240 h was chosen as the equilibrium swelling ratio.

2.5. Scanning electron microscopy (SEM)

The as-prepared cylindrical gels were frozen and fractured in liquid nitrogen, and then dried in a lyophilizer. The fractured surfaces of the dried gels were coated with a thin layer of gold before the tests, and then the SEM tests were conducted using a MERLIN Compact instrument (Zeiss, Germany) at a voltage of 15 kV.

3. Results and discussion

3.1. Synthesis of hybrid NC gels

As shown in Fig. 1, hybrid PAAm/LAPONITE® NC gels were synthesized by in situ polymerization of AAm in the presence of LAPONITE® nanoplatelets and a small amount of chemical crosslinker (MBA) in an aqueous solution. PAAm chains adsorbed onto the surface of LAPONITE® nanoplatelets acted as physical crosslinkers in the hybrid NC gels. At the same time, PAAm chains were also covalently crosslinked by MBA during the polymerization, forming a loosely chemical crosslinked network. We assumed the synergistic effect of physical and chemical crosslinking would achieve better mechanical properties for the NC gels, which is attributed to the effective energy dissipation with the peeling off of PAAm chains from the surface of the LAPONITE® nanoplatelets and the soft and ductile nature of the loosely chemical crosslinked PAAm network. Although the loosely chemical crosslinked PAAm network cannot bear stress, it can hold the integrity of the gel when the physical interactions between PAAm chains and LAPONITE® are broken. Fatigue resistance and self-healing properties of the NC gels should be also influenced by the loosely chemical crosslinked network. The physical NC gels without MBA (LxM0) and PAAm gels crosslinked only by MBA (L0M5) were also prepared for comparison.

Fig. 1 Schematic of the preparation of hybrid PAAm/LAPONITE® NC gels.

3.2. Tensile properties

The tensile stress–strain curves of the physical PAAm NC gels (L4M0 and L10M0), chemical PAAm gel (L0M5) and hybrid PAAm/LAPONITE® NC gels (L4M5 and L10M5) are shown in Fig. 2a. It could be found that the hybrid PAAm/LAPONITE® NC gels exhibited better fracture stress, elastic modulus and fracture energies than those of the physical PAAm NC gel (Fig. 2b and c and Table S1†). The L10M5 hybrid NC gel demonstrated the best tensile properties (σ_f of 0.45 MPa, E of 72.64 kPa and W of 9.81 MJ m⁻³). In contrast, the L10M0 physical NC gel showed σ_f of 0.19 MPa, E of 47.32 kPa and W of 6.63 MJ m⁻³. Compared with the L10M0 physical NC gel, the hybrid NC gel exhibited small elongation though only a very small amount of chemical crosslinker were introduced in the network (Fig. 2c and Table S1†). A similar trend had been also found in L4M5 hybrid NC gel and L4M0 physical NC gel. Moreover, at the same concentration of chemical crosslinker (MBA was only 0.05 mol% of AAm), the higher LAPONITE® concentration resulted in better tensile behaviors. The L4M5 hybrid NC gel exhibited σ_f of 0.22 MPa, E of 45.51 kPa and W of 4.14 MJ m⁻³, which were much lower than those of the L10M5 hybrid NC gel. Not surprisingly, in the absence of MBA, the mechanical properties of the L10M0 physical NC gel with higher LAPONITE® concentration (σ_f of 0.19 MPa, E of 47.32 kPa and W of 6.63 MJ m⁻³) were also better than those of the L4M0 physical NC gel (σ_f of 0.06 MPa, E of 35.92 kPa and W of 2.85 MJ m⁻³), which was consistent with the reports by other groups.^31,32 Here, we also found the mechanical properties of the L0M5 PAAm chemical gel (σ_f of 0.03 MPa, E of 32.15 kPa, ε_f of 3.55 mm mm⁻¹ and W of 0.08 MJ m⁻³) were much weaker than those of the hybrid NC gels and physical NC gels (Fig. 2b and c and Table S1†). The above results of the hybrid NC gels, physical NC gels and chemical gel indicated that the synergistic effect of physical and chemical crosslinking could lead to better mechanical properties of the NC gels. Moreover, all hydrogel samples were transparent (Fig. S1†), suggesting the addition of MBA did not affect the transparency of the NC gels.³³

Fig. 2 Mechanical properties of physical, chemical and hybrid crosslinked hydrogels: (a) tensile stress–strain curves, (b) fracture stress and fracture energy and (c) elongation and elastic modulus.

We further performed a series of tensile tests to quantitatively evaluate MBA and LAPONITE® concentration on the tensile properties of hybrid PAAm/LAPONITE® NC gels. As shown in Fig. 3a, MBA concentration had a great influence on the tensile properties of hybrid NC gels. There was a peak value for σ_f or W of the hybrid NC gels. As evidenced in Fig. 3b and Table S2,† when MBA increased from 0 to 0.05 mol%, σ_f and W increased from 0.06 to 0.22 MPa and 2.85 to 4.14 MJ m⁻³, respectively. Differently, it was found from Fig. 3c and Table S2† that the E of hybrid NC gels monotonously increased (from 35.92 to 51.05 kPa) with the increase of MBA concentration, while ε_f decreased drastically (from 60.52 to 7.05 mm mm⁻¹) as MBA increased. However, as the MBA concentration further increased (>0.05 mol%), σ_f and W decreased distinctly (Fig. 3c and Table S2†). Specifically, the hybrid NC gel with MBA of 0.1 mol% had a σ_f of 0.09 MPa, ε_f of 7.05 mm mm⁻¹, E of 51.05 kPa and W of 0.38 MJ m⁻³. In contrast, the hybrid NC gel with MBA of 0.05 mol% had a σ_f of 0.22 MPa, ε_f of 31.77 mm mm⁻¹, E of 45.51 kPa and W of 4.14 MJ m⁻³. Consequently, the hybrid PAAM/LAPONITE® NC gels with MBA > 0.05 mol% would become brittle, indicating loose chemical crosslinking was important for achieving balanced strength and toughness in the hybrid NC gels, which is consistent with the report of Xiong et al.³⁰ Therefore, MBA of 0.05 mol% was chosen for further investigation. The effect of LAPONITE® concentration on the tensile properties of the hybrid PAAm/LAPONITE® NC gels is illustrated in Fig. S2 and Table S3. †All the mechanical parameters (σ_f, ε_f, E and W) of hybrid NC gels increased with the increase of LAPONITE® concentration. The hybrid PAAm/LAPONITE® NC gels with LAPONITE® of 10 w/v% and MBA of 0.05 mol% (i.e. L10M5) exhibited the highest tensile properties (σ_f of 0.45 MPa, ε_f of 50.63 mm mm⁻¹, E of 72.64 kPa and W of 9.81 MJ m⁻³), which were much better than those of the pure chemical PAAm gel (L0M5 gel, σ_f of 0.03 MPa, ε_f of 3.55 mm mm⁻¹, E of 32.15 kPa and W of 0.08 MJ m⁻³) and the physical PAAm/LAPONITE® NC gel (L10M0 gel, σ_f of 0.19 MPa, ε_f of 65.04 mm mm⁻¹, E of 47.32 kPa and W of 6.63 MJ m⁻³).

Fig. 3 Effect of MBA concentration on the mechanical properties of PAAm/LAPONITE® hydrogels with a LAPONITE® concentration of 4 w/v%.

3.3. Swelling properties

The swelling properties of the physical, chemical and hybrid crosslinked hydrogels are shown in Fig. 4a. It could be found that the swelling ratio of the gels was significantly influenced by the loose chemical crosslinking, the equilibrium swelling ratio (ESR) of L0M5, L4M5 and L10M5 gels was almost the same and the values were 21.61, 21.81 and 21.54, respectively. In contrast, the ESR of the L4M0 and L10M0 gels were 90.57 and 52.22, respectively, which were much higher than those gels with loose chemical crosslinking. The photos of the as-prepared and swollen hydrogels could be seen in Fig. 4b. The volume change of L4M0 was bigger than that of L10M0, but L10M0 was still bigger than that of L0M5, L4M5 and L10M5.

Fig. 4 (a) Swelling ratio of physical, chemical and hybrid crosslinked hydrogels in water. (b) Image of the physical, chemical and hybrid crosslinked hydrogels before (second line) and after (first line) swelling.

Fig. 5 shows SEM photos of the PAAm/LAPONITE® NC gels. The chemical crosslinking of the L0M5 gel without the LAPONITE® was loosest and could form large porous structures (Fig. 5a and b) that absorbed more water, resulting in the large ESR. The data of swelling and SEM indicated LAPONITE® nanoplatelets acted as multifunctional crosslinkers in the NC gels, which was consistent with the reported literature.³⁴ High LAPONITE® concentration led to larger crosslinking density of the NC gels, resulting in a denser network structure (Fig. 5c, d, g and h) and a smaller ESR (Fig. 4a).^35–37 Moreover, MBA acted as chemical crosslinkers and formed smaller and denser micropores (Fig. 5e, f, i and j), as well as a smaller equilibrium swelling (Fig. 4a). However, based on the similar swelling ability of L0M5, L4M5 and L10M5 (Fig. 4a), it could be concluded that the influence of the LAPONITE® concentration on the swelling properties of the hybrid NC gels was weakened owing to the introduction of the small amount of chemical crosslinker.

Fig. 5 (a) SEM images of (a and b) L0M5 gel, (c and d) L4M0 gel, (e and f) L4M5 gel, (g and h) L10M0 gel and (i and j) L10M5 gel.

3.4. Energy dissipation

Fig. 6 demonstrated the cyclic loading–unloading tensile curves of various gels to reveal the energy dissipation capacity of PAAm/LAPONITE® NC gels. At λ = 10, hybrid PAAm/LAPONITE® NC gels (L4M5 and L10M5) exhibited distinct hysteresis loops, while physical NC gels (L4M0 and L10M5) only showed smaller hysteresis loops. Consistently, the dissipated energies of hybrid L4M5 and L10M5 NC gels were 103.64 kJ m⁻³ and 123.21 kJ m⁻³, respectively. In contrast, the physical L4M0 and L10M0 NC gels were 58.90 and 84.73 kJ m⁻³, respectively (Table S4†). The results indicated hybrid NC gels could dissipate more energy than that of the physical NC gels. Moreover, the dissipated energies of NC gels increased with the increase of LAPONITE® concentration (Fig. 6a and Table S4†). The second loading cycle was also conducted immediately to the same gel specimens after the first loading–unloading cycle. As shown in Fig. 6b, the gels without resting exhibited negligible hysteresis loops. In addition, a softening phenomenon was also detected at the second loading. The E at the second loading of L4M0, L4M5, L10M5 and L10M5 NC gels were 33.75, 36.89, 38.53 and 47.25 kPa, respectively, which were much lower than those of the first loading (Fig. 2c). The results indicated the physical crosslinking caused by adsorption of polymer chains onto the LAPONITE® nanoplatelets also contributed to the effective elastic chain density. The peeling of polymer chains from nanoplatelets decreased the effective elastic chain density, leading to the decrease of E and softening of the NC gels at the second loading.

Fig. 6 (a) The first and (b) second immediate cyclic loading–unloading stress–strain curves of PAAm/LAPONITE® hydrogels at λ = 10.

3.5. Self-recovery and fatigue resistance properties

Polymer–LAPONITE® interactions are physical interactions and the peeled polymer chains from nanoplatelets can be re-adsorbed onto the surface of LAPONITE® nanoplatelets. Therefore, PAAm/LAPONITE® NC gels were expected to show self-recovery performances, though they could not be recovered at the second immediate loading. Herein, the cyclic loading–unloading experiments with various recovery times after the first loading were conducted at room temperature without any external stimuli to further examine the influence of LAPONITE® concentration and the small amount of MBA on the self-recovery properties of NC gels.

Fig. 7a and b illustrate the cyclic loading–unloading stress–strain curves of the physical NC gel (L10M0) and the hybrid NC gel (L10M5) with various recovery times. It was found that the hysteresis loops increased with increasing recovery time, which were larger than the loops without recovery time (t = 0 min), but smaller than the loops of the first loading. The dissipated energies of L4M0 and L4M5 gels also increased as the recovery time increased (Table S4†). The ratio of the dissipated energy of the first loading to the dissipated energy after different recovery times can reflect the self-recovery of the gel structure after tensile deformation. Fig. 7 shows the recovery of dissipated energy at different recovery times of the PAAm/LAPONITE® NC gels at λ = 10. The self-recovery of all the NC gels was found to increase significantly with the increase of recovery time. For physical NC gels at t = 30 min, the recovery of the L4M0 gel was 68.01%, which was better than that of the L10M0 gel (63.90%), indicating higher LAPONITE® concentration of the physical NC gels led to a slower recovery in the absence of chemical crosslinkers. However, for hybrid NC gels at t = 30 min, the recovery of the L4M5 gel was 51.23%, which was lower than that of the L10M5 gel (60.36%), inferring the recovery of hybrid NC gels was better at higher LAPONITE® concentration than in the presence of chemical crosslinkers. Moreover, in the present work, we also found the recovery of physical NC gels was faster than that of hybrid NC gels. In Fig. 8, L4M0 and L10M0 physical NC gels achieved recovery of 65.26% and 46.81% after 3 min recovery, respectively. While the recovery of L4M5 and L10M5 was only 28.48% and 38.62%, respectively. These results revealed that better self-recovery properties of the NC gels could be achieved by the synergistic effect of physical and chemical crosslinking.

Fig. 7 Cyclic loading–unloading stress–strain curves of the first and different recovery times of (a) L10M0 gel and (b) L10M5 gel at λ = 10.

Fig. 8 Recovery of dissipated energy of different recovery times of the PAAm/LAPONITE® NC gels at λ = 10.

Fig. 9 showed the recovery of the PAAm/LAPONITE® NC gels for the same gel specimen in five successive cyclic loadings with 30 min recovery between two tests at λ = 10. It was found that the recovery rate of the L4M0 hydrogel varied from 68.01% (second) to 65.87% (sixth), which was almost unchanged after the first cyclic loading. Similarly, the recovery rate of the L10M0 physical NC gel varied from 63.90% (second) to 53.19% (sixth), which was only a small decrease (∼17% decrease). The results indicated the physical NC gels had excellent fatigue resistance properties. However, the recovery rate of the L4M5 hybrid NC gel varied from 51.23% (second) to 24.10% (sixth), which decreased obviously (∼53% decrease); and the recovery of the L10M5 hybrid NC gel was from 60.36% (second) to 35.93% (sixth), which deceased by ∼40%. These results revealed that there was a negative effect on the fatigue resistance property of the NC gels in the presence of chemical crosslinking. However, the better fatigue resistance of hybrid NC gels could be achieved at a higher LAPONITE® concentration.

Fig. 9 Recovery of dissipated energy of PAAm/LAPONITE® NC gels for the same gel specimen in five successive loadings with 30 min recovery between two tests at λ = 10.

3.6. Self-healing properties

Fig. 10a illustrates the tensile stress–strain curves of the healed PAAm/LAPONITE® NC gels at 80 °C for 24 h. It was found that the fracture stress of the healed NC gels was 21–52 kPa, and the healed L10M0 physical gel exhibited the highest fracture stress of ∼52 kPa (Table S7†). Fig. 10b shows the self-healing efficiency (fracture stress ratio of healed gel to as-prepared gel, %) of PAAm/LAPONITE® NC gels at various healing temperatures. It could be found the healing efficiency of all the NC gels increased with increasing healing temperature, inferring the healing process could be improved by heat treatment. The healing efficiency of the L4M0 and L10M0 physical NC gels were 49.03% and 27.84% at 80 °C, respectively, which were similar to results reported by Haraguchi et al.²⁵ They found the healing efficiency of physical poly(N,N-dimethylacrylamide)/clay NC gels became worse as the LAPONITE® concentration became higher than 6 w/v%. Here, we also found in Fig. 10b that the healing efficiency of the L4M5 and L10M5 hybrid NC gel were only 13.96% and 9.70% at 80 °C, respectively, which were much lower than those of the physical NC gels. These results revealed that the addition of a small amount of chemical crosslinker could strongly suppress the self-healing ability of NC gels, and healing efficiency decreased with the increase of LAPONITE® concentration in the hybrid NC gels.

Fig. 10 (a) Tensile stress–strain curves of healed PAAm/LAPONITE® hydrogels with the healing temperature of 80 °C. (b) Healing efficiency of PAAm/LAPONITE® hydrogels at different healing temperatures.

3.7. Mechanisms

Due to the excellent mechanical properties of the NC gels, the toughening and energy dissipation mechanisms of NC gels have drawn great attention in recent years. As reported by Tang et al.,³⁸ the dissipation mechanisms of NC gels are dependent on the maximum stretch ratio. At small strain, the energy dissipation origins from the orientation of LAPONITE® platelets, and the hysteresis loops are almost overlapped as applied multiple cyclic loading. However, at large strain, the hysteresis loops are much larger than at the small strain, which are caused by the peeling-off of polymer chains from the surface of the LAPONITE® platelets. Similarly, Klein et al.³⁹ also stated that the energy dissipation is due to the formation and rupture of clay–clay and polymer–clay interactions. The extent of hysteresis is influenced both by the strength and density of these interactions. The peeling-off and re-adsorption of polymer chains in NC gels are analogous to carbon black in filled rubbers, in which polymer chains are also physically and reversibly adsorbed onto the carbon black nanoparticles.

The energy dissipation mechanisms of our hybrid NC gels should be similar to the physical NC gels as mentioned above. The sole difference of our hybrid NC gels to physical NC gels was the loose chemical crosslinking. Fig. 11 demonstrates the deformation and recovery mechanisms of the hybrid NC gels. As a large strain is applied, polymer chains peeled-off from the LAPONITE® platelets. The “peeling-off” process was accompanied with energy dissipation, resulting in large hysteresis loops in the cyclic loading. In this process, chemical crosslinking could maintain the integrity of the whole gel network. As the force released, the peeled-off polymer chains re-adsorbed onto the surface of the LAPONITE® platelets. Owing to the hybrid network structure and effective energy dissipation, PAAm/LAPONITE® hybrid NC gels exhibited high strength and large hysteresis, and the improvement of mechanical properties became more effective at higher LAPONITE® concentrations (i.e. stronger polymer–particles interactions). In the PAAm/LAPONITE® hybrid NC gels, the peeling-off of polymer chains from the surface of the LAPONITE® platelets could serve as a “sacrificial bond”. Notably, we also found that there were very limited hysteresis loops for all the NC gels in the immediate second cyclic loading (Fig. 6b and Table S4†). It could be concluded that the re-adsorption process of the peeled-off polymer chains could not occur instantaneously as the force released. The results were consistent with Tang's work.³⁸ They estimated the re-adsorption ratio of polymer chains was approximately 15%, which was independent of maximum strain, indicating a poor recovery of NC gels without resting time.

Fig. 11 Illustration of deformation and recovery mechanisms of hybrid PAAm/LAPONITE® NC gels.

More importantly, it is in the re-adsorption process that the NC gels exhibit self-recovery fatigue resistance and self-healing performances. In this work, we found all the physical and hybrid NC gels demonstrated distinct self-recovery processes. For physical NC gels, L10M0 showed a relatively slower recovery than that of L4M0 at each recovery time (Fig. 8). This phenomenon might be attributed to the orientation of the nanoclay platelets along the stretching direction at higher LAPONITE® concentrations. Based on the analysis of Lian et al.,⁴⁰ as the LAPONITE® concentration becomes larger than 6 w/v%, the distance between nanoclay platelets in the NC gels is equal to the average diameter of nanoclay platelets. As a result, the orientation of nanoclay platelets is pronounced, and the clay–clay interactions are strong, which cannot be destroyed easily as the external force is released. For the L10M0 NC gel (C_clay = 10 w/v% > 6 w/v%), the orientation of the nanoclay platelets in the stretched NC gel would strictly hamper the re-adsorption of polymer chains onto the platelet surface,^24,32 which also limits the relaxation of connected polymer chains.⁴⁰ However, for the L4M0 NC gel (C_clay = 4 w/v% < 6 w/v%), the distance between nanoclay platelets was larger than the average diameter of nanoclay platelets, and the recovery of the network structure was easier. Therefore, the physical NC gels with lower LAPONITE® concentrations exhibited better self-recovery, fatigue resistance and self-healing properties. Nevertheless, the influence of loose chemical crosslinking on the mechanical responsive properties of NC gels was complex. The self-recovery rate, fatigue resistance and self-healing of the hybrid NC gels became worse than those of the physical NC gels. In the presence of chemical crosslinking, the recovery speed of the hybrid NC gels was slower than that of the physical NC gels, but the recovery was similar if the recovery time was long enough (Fig. 8). In contrast, the fatigue resistance and self-healing properties of the hybrid NC gels were significantly influenced by the small amount of chemical crosslinker. For fatigue resistance, the probable reason was attributed to the presence of a heterogeneous network as the introduction of chemical crosslinking in the NC gels causes stress concentration at the shortest polymer chains and irreversible chain breakage of these polymer chains. Moreover, compared with the L10M5 and L4M5 gels (Fig. 9 and Table S5†), the higher LAPONITE® concentration was a benefit to the fatigue resistance of hybrid NC gels. For self-healing, the presence of chemical crosslinking would strongly restrict chain diffusion, resulting in the decrease of chain density at the damage interface. At the same time, the re-adsorbed polymer chains on the nanoclay platelets also decreased. Consequently, the self-healing efficiency of the hybrid NC gels decreased obviously. In addition, in spite of physical and hybrid NC gels, the increase in healing temperature improved the self-healing efficiency, which might be caused by the acceleration of chain diffusion at higher temperatures.

Actually, incorporating LAPONITE® or MBA decreased both the self-healing and the self-recovery properties. However, the self-healing property was influenced much worse than the self-recovery property. The healing efficiency at 80 °C sharply decreased from ∼49% for L4M0 to ∼14% for L4M5 and from ∼28% for L10M0 to ∼10% for L10M5 (Fig. 10). Nevertheless, the self-recovery of the gels with 4 or 10 w/v% LAPONITE® or with/without MBA were similar (∼80%, Fig. 8) if the recovery time was long enough. The difference might also be caused by the different gel states. For both of self-recovery and self-healing, the re-adsorption of polymer chains is very important. But in self-recovery tests, the gel specimen was integrated and the fracture of the network structure during loading was peeling of the adsorbed polymer chains from the surface of clay. In self-healing tests, the gel specimens were cut into separated pieces, which were contacted together for healing tests. It indicated that there was breakage of covalent bonds in the hybrid NC gels. It was difficult to repair such broken chemical networks in the hybrid NC gels. Moreover, it was also found that the self-recovery and self-healing properties were influenced by clay concentration. Therefore, incorporating LAPONITE® has a negative effect on the self-healing property (for the cut gels), but the hybrid NC gels still had excellent self-recovery performances (after tensile deformation).

4. Conclusions

In summary, hybrid crosslinked PAAm/LAPONITE® NC gels were synthesized by in situ free radical polymerization and the effect of LAPONITE® and chemical crosslinker concentration on the tensile properties, energy dissipation, self-recovery, fatigue resistance and self-healing of the hydrogels were investigated by various tests. The fracture stress and fracture energies of the hybrid NC gels increased obviously as MBA increased from 0 to 0.05 mol%; however, these mechanical parameters decreased sharply as the MBA amount rose above 0.05 mol%. At the optimal formulation, the PAAm/LAPONITE® hybrid NC gel with 10 w/v% of LAPONITE® and 0.05 mol% of MBA showed a σ_f of 0.45 MPa, ε_f of 50.63 mm mm⁻¹, E of 72.64 kPa and W of 9.81 MJ m⁻³, which were much better than those of the physical NC gel and the pure chemical gel. With loose chemical crosslinking, the hybrid NC gels exhibited improved tensile properties, large hysteresis and good self-recovery performances. Our results further identify that a combination of physical and chemical crosslinking in the network is an effective strategy to prepare high strength and tough hydrogels. However, it was also revealed there was a negative effect of chemical crosslinking on the fatigue resistance and self-healing performances of NC gels. Compared to hybrid double network hydrogels (DN gels) with two crosslinkings in separated networks, the two crosslinkings in our hybrid PAAm/LAPONITE® NC gels were in a single network, which may be the original reason for the different effects of this loose chemical crosslinking on the mechanical behaviors of the hydrogels. As a complement to physical NC gels, we believe this work will provide better understanding about the relationship of network structure and properties for hybrid NC gels.

Acknowledgements

J. Y. and Q. C. are grateful for financial support from the Joint Fund for Fostering Talents of NSFC-Henan Province (U1304516), National Nature Science Foundation of China (21504022), Henan Province (12B430007,13A430015, 14B430013 and No. 16IRTSTHN005) and Henan Polytechnic University (B2010-6, B2013-016, NSFRF140111 and 72105/001).

References

1. D. Lim, E. Lee, H. Kim, S. Park, S. Baek and J. Yoon, Soft Matter, 2015, 11, 1606–1613.
2. B. Balakrishnan and R. Banerjee, Chem. Rev., 2011, 111, 4453–4474.
3. H. Li, A. M. Koenig, P. Sloan and N. D. Leipzig, Biomaterials, 2014, 35, 9049–9057.
4. N. Huebsch, C. J. Kearney, X. Zhao, J. Kim, C. A. Cezar, Z. Suo and D. J. Mooney, Proc. Natl. Acad. Sci. U. S. A., 2014, 111, 9762–9767.
5. A. Servant, V. Leon, D. Jasim, L. Methven, P. Limousin, E. V. Fernandez-Pacheco, M. Prato and K. Kostarelos, Adv. Healthcare Mater., 2014, 3, 1334–1343.
6. C. Yu, P. Yuan, E. M. Erickson, C. M. Daly, J. A. Rogers and R. G. Nuzzo, Soft Matter, 2015, 11, 7953–7959.
7. M. E. McConney, N. Chen, D. Lu, H. A. Hu, S. Coombs, C. Liu and V. V. Tsukruk, Soft Matter, 2009, 5, 292–295.
8. D. Buenger, F. Topuz and J. Groll, Prog. Polym. Sci., 2012, 37, 1678–1719.
9. H. Chen, M. Zhang, J. Yang, C. Zhao, R. Hu, Q. Chen, Y. Chang and J. Zheng, Langmuir, 2014, 30, 10398–10409.
10. K. Xu, J. Wang, S. Xiang, Q. Chen, Y. Yue, X. Su, C. Song and P. Wang, Compos. Sci. Technol., 2007, 67, 3480–3486.
11. J. Y. Sun, X. Zhao, W. R. Illeperuma, O. Chaudhuri, K. H. Oh, D. J. Mooney, J. J. Vlassak and Z. Suo, Nature, 2012, 489, 133–136.
12. K. Haraguchi and T. Takehisa, Adv. Mater., 2002, 14, 1120–1124.
13. J. P. Gong, Y. Katsuyama, T. Kurokawa and Y. Osada, Adv. Mater., 2003, 15, 1155–1158.
14. Q. Chen, H. Chen, L. Zhu and J. Zheng, J. Mater. Chem. B, 2015, 3, 3654–3676.
15. W. Li, H. An, Y. Tan, C. Lu, C. Liu, P. Li, K. Xu and P. Wang, Soft Matter, 2012, 8, 5078–5086.
16. S. Abdurrahmanoglu, V. Can and O. Okay, Polymer, 2009, 50, 5449–5455.
17. T. L. Sun, T. Kurokawa, S. Kuroda, A. B. Ihsan, T. Akasaki, K. Sato, M. A. Haque, T. Nakajima and J. P. Gong, Nat. Mater., 2013, 12, 932–937.
18. J. Zhang, N. Wang, W. Liu, X. Zhao and W. Lu, Soft Matter, 2013, 9, 6331–6337.
19. T. Bai, P. Zhang, Y. Han, Y. Liu, W. Liu, X. Zhao and W. Lu, Soft Matter, 2011, 7, 2825–2831.
20. M. Guo, L. M. Pitet, H. M. Wyss, M. Vos, P. Y. W. Dankers and E. W. Meijer, J. Am. Chem. Soc., 2014, 136, 6969–6977.
21. T. Huang, H. G. Xu, K. X. Jiao, L. P. Zhu, H. R. Brown and H. L. Wang, Adv. Mater., 2007, 19, 1622–1626.
22. T. Wang, D. Liu, C. Lian, S. Zheng, X. Liu and Z. Tong, Soft Matter, 2012, 8, 774–783.
23. J. Yang, C. R. Han, X. M. Zhang, F. Xu and R. C. Sun, Macromolecules, 2014, 47, 4077–4086.
24. T. Nishida, A. Obayashi, K. Haraguchi and M. Shibayama, Polymer, 2012, 53, 4533–4538.
25. K. Haraguchi, K. Uyama and H. Tanimoto, Macromol. Rapid Commun., 2011, 32, 1253–1258.
26. D. C. Tuncaboylu, A. Argun, M. P. Algi and O. Okay, Polymer, 2013, 54, 6381–6388.
27. T. Li, S. Xiang, P. Ma, H. Bai, W. Dong and M. Chen, J. Polym. Sci., Part B: Polym. Phys., 2015, 53, 1020–1026.
28. J. Shen, B. Yan, T. Li, Y. Long, N. Li and M. Ye, Soft Matter, 2012, 8, 1831–1836.
29. S. Rose, A. Marcellan, T. Narita, F. Boue, F. Cousin and D. Hourdet, Soft Matter, 2015, 11, 5905–5917.
30. L. Xiong, M. Zhu, X. Hu, X. Liu and Z. Tong, Macromolecules, 2009, 42, 3811–3817.
31. L. Xiong, X. Hu, X. Liu and Z. Tong, Polymer, 2008, 49, 5064–5071.
32. G. Gao, G. Du, Y. Sun and J. Fu, ACS Appl. Mater. Interfaces, 2015, 7, 5029–5037.
33. K. Haraguchi and L. Song, Macromolecules, 2007, 40, 5526–5536.
34. C. W. Chang, A. van Spreeuwel, C. Zhang and S. Varghese, Soft Matter, 2010, 6, 5157–5164.
35. S. Zheng, T. Wang, D. Liu, X. Liu, C. Wang and Z. Tong, Polymer, 2013, 54, 1846–1852.
36. M. Liu, W. Li, J. Rong and C. Zhou, Colloid Polym. Sci., 2012, 290, 895–905.
37. E. Zhang, T. Wang, C. Lian, W. Sun, X. Liu and Z. Tong, Carbon, 2013, 62, 117–126.
38. J. Tang, G. Xu, Y. Sun, Y. Pei and D. Fang, J. Appl. Phys., 2014, 116, 244901.
39. A. Klein, P. G. Whitten, K. Resch and G. Pinter, J. Polym. Sci., Part B: Polym. Phys., 2015, 53, 1763–1773.
40. C. X. Lian, Z. M. Lin, T. Wang, W. X. Sun, X. X. Liu and Z. Tong, Macromolecules, 2012, 45, 7220–7227.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra04234a

---