Tough hydrogels enabled by transient entanglements

Abstract

Achieving high toughness and strength simultaneously in single-covalent-network hydrogels remains a longstanding challenge. Herein, we report a simple yet effective strategy to resolve this strength-toughness conflict by constructing polyacrylamide (PAAm) networks with abundant dangling chains that form transient entanglements. Unlike permanently trapped entanglements, these transient entanglements can slip and fully disentangle upon loading, enabling highly efficient energy dissipation and stress redistribution over a broad range of strains. Besides, these networks exhibit superior homogeneity compared to other structures, effectively mitigating stress concentration. As a result, our single-covalent-network hydrogels exhibit good mechanical properties, including a fracture strain of 5071%, a fracture strength of 1.06 MPa, a fatigue threshold of 1968 J·m⁻², and a fracture energy of approximately 60,000 J·m⁻². Moreover, these hydrogels feature low friction and high wear-resistance. Such a simple yet robust design paradigm effectively overcomes the longstanding strength–toughness trade-off without the complexity of multi-network architectures, opening avenues for next-generation hydrogels in biomedicine, wearable electronics, and other demanding environments.

Achieving both high strength and toughness in single-covalent-network hydrogels is a challenge. Here, the authors report a transient entanglement strategy which enables energy dissipation and stress redistribution to achieve strong, tough polyacrylamide hydrogels.

Introduction

Mechanically robust hydrogels that combine high toughness and strength are highly desirable for a wide range of applications, such as wearable electronics^1,2, bioengineering^3,4, actuators^5,6, sensors^7,8, and enhanced oil recovery^9–11. However, achieving both properties simultaneously is a significant challenge, as toughness and strength are often seen as mutually exclusive in materials^12–15. This trade-off is particularly pronounced in single-covalent-network hydrogels based on covalently cross-linked polyacrylamide (PAAm). While single-covalent-network PAAm hydrogels are relatively easy to prepare and most widely used, balancing toughness and strength remains difficult. To increase strength, a denser cross-linking network is usually required, but this often leads to embrittlement. According to the Lake-Thomas model^16–18, a higher cross-linking density results in reduced ductility and toughness, making it challenging to achieve both high strength and toughness in a single network.

Over the past few decades, various strategies have been explored to reconcile the inherent strength-toughness trade-off in hydrogels, particularly through the design of multi-network hydrogels. These materials combine covalent cross-links with additional networks based on hydrogen bonding^19–21, electrostatic interactions^22,23, host–guest chemistry^24,25, and crystallization^26–28, providing both deformation resistance and energy dissipation. While multi-network designs can partially address the strength-toughness conflict, their construction is complex and often requires the introduction of functional monomers. A promising recent development involves molecular entanglements, which act as additional topological cross-links, restricting chain mobility and contributing to both elasticity and toughness^29–32. However, tailoring entanglements in hydrogels is challenging due to the competition and repulsion of water molecules, which tend to weaken the entanglement effect³³. Despite these challenges, careful structural design can engineer entanglements to enhance material properties. One such approach is the double-network (DN) design, in which polymer chains in a ductile matrix entangle with a rigid skeleton^34,35. This allows the chains to slip and redistribute stress during deformation, thus improving both strength and toughness. A similar mechanism has been applied to macromolecular microsphere cross-linked hydrogels, where polymer chains interpenetrate and entangle within microspheres^36,37. More recently, single-covalent-network PAAm hydrogels have been developed in which permanently trapped entanglements significantly outnumber covalent cross-links¹². In these hydrogels, entanglements function as slip links, reinforcing the material without causing embrittlement. During deformation, entanglements slip first, and then transmit tension to other chains, dissipating energy when covalent bonds break. This approach has significantly improved the mechanical properties of PAAm hydrogels, achieving strengths of 400 kPa and fracture energies of 2200 J/m². However, these designs still rely on permanently trapped entanglements, which can only slip but cannot fully disentangle. This limits their energy dissipation capacity and overall toughness.

In contrast, introducing transient entanglements may offer a pathway for creating even tougher and stronger hydrogels. These entanglements can slip and disentangle during deformation, providing enhanced energy dissipation and stress redistribution across the material. In this study, we demonstrate an approach to resolving the strength-toughness conflict by fabricating a single-covalent-network PAAm hydrogel with abundant dangling chains that interlock to form transient entanglements. Unlike permanently trapped entanglements, these transient entanglements can not only slip but also fully disentangle under stress. This mechanism allows for superior energy dissipation and stress redistribution across a wide range of strain. As a result, our transiently entangled PAAm hydrogels exhibit good properties, including a fracture strain of 5071%, a fracture strength of 1.06 MPa, a fatigue threshold of 1968 J·m⁻², and a fracture energy of ~60,000 J·m⁻². Furthermore, these hydrogels also feature a more homogeneous network, with low friction and high wear-resistance, making them highly promising for biomedical applications and beyond. This simple yet effective approach provides a significant step forward in the development of tough, strong, and durable hydrogels.

Results

Synthesis of transiently entangled hydrogels

We synthesized three types of single-covalent-network PAAm hydrogels: a regular hydrogel, a permanently entangled hydrogel, and a transiently entangled hydrogel (Fig. 1a, b). The regular hydrogel is cross-linked using N,N′-methylenebis(acrylamide) (MBA), where covalent cross-links dominate the network structure. The permanently entangled hydrogel features an ultra-high polymer concentration with fewer covalent cross-links, allowing the permanently trapped entanglements to outnumber the cross-links. During deformation, the entangled chains in the permanently entangled hydrogel slip to redistribute stress, and the covalently cross-linked chains eventually break, leading to significant damage to the hydrogel network (Fig. 1c, d). In contrast, the transiently entangled hydrogel allows its entangled chains to undergo both slip and disentanglement during deformation, effectively dissipating energy and enhancing the toughness (Fig. 1e).

Fig. 1. Schematic design of three types of hydrogels.

a Network internal polymer chain arrangement styles of regular hydrogel, permanently entangled hydrogel and transiently entangled hydrogel. b Localized structural formulae of the transiently entangled hydrogel networks, with macromolecular cross-linkers in red, dangling PAAm chains in blue, and cross-linked PAAm chains in green. Different fracture behaviors exhibited by regular hydrogel (c), permanently entangled hydrogel (d) and transiently entangled hydrogel (e) at small and large strains upon deformation. f Ashby plot of toughness (Γ, i.e. fracture energy) versus tensile strain of transiently entangled hydrogel and other reported tough hydrogels. g Photographs of a notched LMC-4.8-AAm-4.2 transiently entangled hydrogel (notch width ≈ 80% of sample width) stretched to 3500%.

To construct the transiently entangled hydrogel, a linear macromolecular cross-linker (LMC) is used to construct a network rich in dangling chains but with minimal chemical cross-linking. The LMC is polyallylamine (PAA) with double bonds on its side chains (Supplementary Fig. 1), synthesized by capping the amino groups with allyl glycidyl ether (AGE). The chemical structures of the reactants and final products are confirmed by proton nuclear magnetic resonance (¹H NMR) spectroscopy (Supplementary Fig. 2), revealing an average of 0.9 unsaturated double bonds per PAA molecular repeat unit. The transiently entangled PAAm hydrogels are fabricated through in situ free radical polymerization of acrylamide (AAm) monomer in the presence of LMC. During polymerization, the PAAm chain radicals react with the double bonds of LMC to form dangling chains and LMC radicals. The LMC radicals can polymerize with AAm monomers to form reactive dangling chains, which either react with double bonds on other LMC molecules, creating chemical cross-links, or terminate via reaction with other PAAm chain radicals, forming additional dangling chains (Supplementary Fig. 3). The resulting network consists of numerous dangling chains that intertwine with each other or couple with the backbone chains, forming transient entanglements. Thus, both chemical cross-links and transient entanglements coexist within the PAAm hydrogel (Fig. 1b and Supplementary Fig. 3). The hydrogels are designated as LMC-m-AAm-n, where m‰ and n represent the mass ratio of LMC cross-linker to AAm monomer and the molar concentration of AAm monomer, respectively. The compositions of the various hydrogels covered in this work are shown in Supplementary Table 1.

We summarize the fracture energy and ultimate strain on an Ashby plot to compare our materials with existing tough hydrogels and ionogels (Fig. 1f)^{12,27,31,37–43}. The transiently entangled hydrogels exhibit superior toughness and ductility compared to most current tough hydrogels and ionogels. Remarkably, even when a rectangular sample is cut to 80% of its original width, the hydrogel still demonstrates exceptional tensile strength, stretching over 35 times its original length (Fig. 1g and Supplementary Fig. 4).

Verification of transient entanglement structure

To investigate the structure of transient entanglements in the hydrogels, we first conducted a stress relaxation experiment. Upon applying a step strain to the entangled PAAm hydrogels, all samples exhibit rapid relaxation behavior (Fig. 2a). Over time, the relaxation modulus, G(t), decrease and eventually reach a plateau, showing clear time dependence. The plateau value (G(t)/G₀) reflects the ratio between permanent cross-links and total cross-links. Due to the high water content, hydrogen bonds between PAAm chains are greatly weakened and contribute only marginally to the elasticity, so the dynamic cross-links are primarily attributed to transient entanglements, while the permanent cross-links involve covalent cross-links and permanently trapped entanglements (Supplementary Fig. 5). The G(t)/G₀ values range from 0.3 to 0.5, indicating that transient entanglements outnumber permanent cross-links. An increase in the cross-linker content results in higher G(t)/G₀ values, indicating a higher permanent cross-link density. In contrast, the regular hydrogel deforms immediately, with G(t) showing minimal variation over time (Supplementary Fig. 6), and the final G(t)/G₀ value is as high as 0.93, indicating high elasticity of network.

Fig. 2. Characterization of network structure.

a Normalized stress relaxation curves, b swelling ratio and cross-linking density of the LMC-m-AAm-4.2 transiently entangled hydrogels with various cross-linker contents, m. Data with error bars in (b) are presented as mean values ± standard deviations of n = 3 replicates. DQ build-up curves of transiently entangled hydrogels with c different swelling ratio and d m. e The reduced stress (${\sigma }_M$) as a function of reciprocal extension ratio ($1/\alpha$), and f, g corresponding Mooney–Rivlin constants ($G_c,G_e$) of the three types of hydrogels. Data with error bars are presented in (f) as mean values ± standard deviations of n = 3 replicates. h The static inhomogeneity, ${⟨I_C\left(p\right)⟩}_E$, and i the decay time (Γ⁻¹) distribution functions, G(Γ), of the three types of hydrogels. Data with error bars in (h) are presented as mean values ± standard deviations of n = 3 replicates.

To determine the permanent cross-linking density of the hydrogels, we performed equilibrium swelling experiments. In the presence of a high amount of water, the transient entanglements disentangle, while the chemical cross-linkers and some permanently entangled chains become the primary components resisting swelling. As shown in Fig. 2b and Supplementary Table 2, increasing the amounts of chemical cross-linkers results in a decrease in the equilibrium swelling ratio, indicating a corresponding increase in the chemical cross-linking density. This result aligns with the findings from the stress relaxation behavior.

To confirm that the dynamic cross-links arise from entanglements of dangling chains, we performed ¹H Double-Quantum NMR (¹H DQ NMR) experiments on hydrogels with varying swelling ratios (Fig. 2c) and cross-linking densities (Fig. 2d). The ¹H DQ NMR results reveal that the DQ excitation curves of all samples show two distinct maxima at different excitation times (τ). These maxima suggest the presence of two networks within the hydrogels: chemical cross-linking network and transient entanglement network⁴⁴. To distinguish the origins of these two maxima, we gradually increase the swelling ratio of the hydrogel, which progressively disentangles the transient entanglements. In Fig. 2c, the normalized DQ (nDQ) buildup curves of LMC-4.8-AAm-4.2 hydrogel at different swelling ratios show a sharp first maximum at a shorter τ_max (~2.2 ms). This maximum, corresponding to stronger dynamic constraints, does not shift with swelling. In contrast, the second maximum at a longer τ_max (~7.8 ms), corresponding to weaker confinement, broadens and shifts to longer τ as the swelling ratio increases. Herein, the polymer chain concentration-sensitive second peak can reasonably be attributed to the confinement effect induced by transient entanglements. Swelling decreases the polymer strand density ($\rho$) and increases the tube diameter (α), following the relationship $\alpha \propto {\rho }^{-1/2}$³³. This ultimately reduces the number density of entanglements in the dangling chains, resulting in a disentangled-like phenomenon. Since the covalent cross-links remain unchanged during swelling, the first maximum, associated with shorter τ_max, corresponds to the rapid Rouse relaxation of chemically cross-linked network strands between adjacent covalent cross-linking points. Moreover, increasing the cross-linker content results in a negligible shift in the first maximum, while the second maximum shifts to a shorter τ_max (Fig. 2d). This suggests that with the increase of chemical cross-linking, the entanglements between dangling chains are enhanced, leading to more constrained segmental dynamics. This is further supported by spin echo experiments, which reveal an increase in the rigid fraction with increasing the content of cross-linkers (Supplementary Fig. 7 and Supplementary Table 3). In summary, these results demonstrate that the PAAm hydrogels possess numerous transient entanglements.

To further distinguish the contribution of entanglements and chemical cross-links to the elasticity of these hydrogels, we use the Mooney–Rivlin model

$${\sigma }_M=\frac{\mathrm{\sigma }}{\mathrm{\alpha }-{\alpha }^{-2}}=G_c+G_{e}f\left(\alpha \right),f\left(\alpha \right)=1/\alpha$$ 1

where $\alpha$ and ${\sigma }_M$ represent the elongation and the Mooney stress of the sample, respectively; $G_c$ and $G_e$ are the Mooney–Rivlin constants. Usually, $G_c$ and $G_e$ correspond to the contributions of permanent cross-linking and entanglements to the elastic modulus, respectively⁴⁵. The $G_c$ and $G_e$ values of these transient hydrogels are shown in Fig. 2e, f and Supplementary Table 4. The higher $G_e$ values of all these hydrogels indicate that entanglements play a dominant role in the stretching process, outweighing the contribution of covalent cross-linking. This suggests that while chemical cross-linking leads to stronger confinement, its contribution to elastic modulus is comparatively smaller. As the cross-linker content increases, $G_c$ also increases, reflecting a higher density of chemical cross-links. On the other hand, $G_e$ first increases and then decreases, indicating the number of branched chains in the network first increases and then decreases. This trend can be explained as follows: as more cross-linkers are added, the number of branching sites increases, leading to more dangling chains attached to the macromolecules. However, beyond a certain threshold, the increased number of growing chain radicals reduces the distance between free radicals, raising the probability of bimolecular coupling and forming more chemical cross-links instead of dangling chains⁴⁶. Furthermore, we calculate the $G_e$/($G_c$ + $G_e$) values of transiently entangled hydrogel, regular hydrogel, and permanently entangled hydrogel to evaluate the contribution of entanglement to the elastic modulus. The $G_e$/($G_c$ + $G_e$) value of transiently entangled hydrogel is 96%, while the corresponding values for the regular hydrogel and permanently entangled hydrogel are 53% and 77%, respectively. This result indicates entanglement plays a more critical role in the transiently entangled hydrogel than that of regular and permanently entangled hydrogel (Fig. 2g, Supplementary Fig. 8 and Supplementary Table 4).

Transient entanglements also contribute to a highly homogeneous polymer network on scales larger than the mesh size, making the network optically transparent (Supplementary Fig. 9) and minimizing the inhomogeneity caused by the random distribution of chemical cross-links. Dynamic light scattering (DLS) tests, shown in Supplementary Fig. 10, provide typical scatter plots for the hydrogels. The time-average scattered intensity, ${⟨I⟩}_T$, of the regular hydrogel fluctuates significantly, whereas the permanently entangled and transiently entangled hydrogels exhibit much smaller fluctuations. The time-fluctuating intensity component, ${⟨I_F⟩}_T$, for these three types of PAAm hydrogels are nearly identical, as ${⟨I_F⟩}_T$ originates from excess scattering beyond the static scattering of the frozen structure and is only temperature dependent⁴⁷. The static inhomogeneity, ${⟨I_C\left(p\right)⟩}_E$, can be calculated by the equation:

$${⟨I_C\left(p\right)⟩}_E={⟨I⟩}_E-{⟨I_F⟩}_T$$ 2

where ${⟨I⟩}_E$ is the ensemble-averaged scattered intensity. We summarize the ${⟨I_C\left(p\right)⟩}_E$ values in Fig. 2h. It can be seen that the entanglements can effectively reduce the network inhomogeneity.

To investigate the dynamics of physical entanglements formed by dangling chains, we measured the intensity time-averaged correlation functions (ICFs) (Fig. 2i and Supplementary Fig. 11). The ICFs of the regular hydrogel shows a single relaxation event around 20 μs. In contrast, both the transiently entangled and permanently entangled hydrogels exhibit an additional, slower relaxation mode. According to the work of Amis et al.⁴⁸, these relaxation modes, i.e., the fast and slow modes, are attributed to the collective motion of entangled chains and the translational diffusion of gel clusters, respectively. Figure 2i presents the G(Γ) of these three hydrogels, obtained by performing an inverse Laplace transform of ICFs. The slower relaxation mode is more clearly resolved in this figure. The single relaxation with a relaxation time of ~27 μs for the regular hydrogel reflects local fluctuations in polymer chain concentration, a behavior commonly observed in many hydrogels⁴⁹. For the entangled hydrogels, a broader and weaker peak appears, corresponding to a slow mode with a relaxation time of ~1250 μs, which is associated with the diffusion of entangled dangling chains. This diffusion is hindered by factors such as entanglement and friction between molecular chains, as noted by Gong et al.⁵⁰. Additionally, the relaxation time of slow mode is sensitive to the covalent cross-link density; in the permanently entangled hydrogel, the movement of dangling chains is even more restricted, leading to a slight increase in the relaxation time.

Characterization of mechanical properties

Transient entanglements significantly enhance both the strength and elongation of hydrogels, as shown by comparing the transiently entangled hydrogel with the permanently entangled hydrogel and the regular hydrogel (Fig. 3a). The ultimate strength and strain for the transiently entangled hydrogel are 1.06 MPa and 3184%, respectively, compared to 0.39 MPa and 404% for the permanently entangled hydrogel, and 0.22 MPa and 1215% for the regular hydrogel. To investigate the fracture mechanics of these hydrogels, pure shear tests were performed. All transiently entangled hydrogels demonstrated high fracture energy and resistance to notch propagation (Fig. 3b, c). Notably, the fracture energy of LMC-4.8-AAm-4.2 hydrogel is 6.02 × 10⁵J·m⁻², which is higher than that of many reported single-covalent-network PAAm hydrogels^12,31,51,52. This value is only slightly lower than those of some ionogels and hierarchical poly(vinyl alcohol) hydrogels^27,53. Furthermore, the fracture energy is proportional to n and inversely proportional to m, scaling as $\Gamma ~80-2.3m$ when n = 5.0.

Fig. 3. Characterization of mechanical properties.

a Tensile stress–strain curves of the three types of hydrogels, the transiently entangled hydrogel is LMC-9.6-AAm-4.2. b Force–extension curves of the notched and un-notched hydrogels (LMC-4.8-AAm-4.2) and c plot of fracture energy versus various cross-linker contents (m) and monomer concentration (n). d Two families of the three types of hydrogels plotted in the strength (σ) versus toughness (fracture energy, Γ) plane. e Fatigue threshold (Γ₀) of LMC-4.8-AAm-4.2 hydrogel determined from plot of crack propagation per loading cycle (dc/dn) versus increasing energy release rate (G) and f an Ashby plot of fatigue threshold versus toughness of transiently entangled hydrogel, other reported tough hydrogels and natural rubber. Tensile stress–strain curves (g) and corresponding tensile strength and modulus (h), fracture strain and fracture work (i) parameters of the LMC-m-AAm-4.2 transiently entangled hydrogels with different cross-linker contents, m. Data with error bars in (h) and (i) are presented as mean values ± standard deviations of n = 3 replicates.

The superior fracture energy indicates that transiently entangled hydrogels possess overall better performance compared to both permanently entangled hydrogels and regular hydrogels. For hydrogels, increasing cross-link density usually raises Young’s modulus and ultimate tensile strength, but it also restricts reptation and disentanglement, curtails sacrificial dissipation pathways, and suppresses crack-tip blunting. These effects elevate the elastic energy-release rate while reducing available inelastic dissipation, so toughness declines even as modulus and strength increase. Therefore, regular hydrogels often face a strength-toughness trade-off, where toughness (represented by Γ, or fracture energy) is inversely related to strength (represented by σ or yield strength σ_y, which reflects cross-link density). This relationship is described as $\Gamma ~{\sigma }^{-2.63}$, indicating that chemical cross-links increase strength but also make the hydrogels more brittle. In contrast, permanently entangled hydrogels partially overcome this issue. The permanently trapped entanglements stiffen the hydrogels but do not significantly compromise their toughness, following the scaling $\Gamma ~{\sigma }^{-1.52}$. Transiently entangled hydrogels, however, show minimal strength-toughness conflict, with a scaling of $\Gamma ~{\sigma }^{-0.61}$. Notably, even with nearly an order of magnitude increase in the σ value, the Γ value only changes slightly, indicating that transient entanglements effectively enhance strength without leading to brittleness (Fig. 3d and Supplementary Fig. 12).

To capture genuine fatigue durability under successive tensile loading-unloading rather than single-shot fracture behavior, we determined the fatigue threshold (Γ₀) by measuring the crack expansion of the transiently entangled hydrogel at different energy release rate amplitudes (Fig. 3e). Linear regression of the data yields a Γ₀ of 1968 J·m⁻², which is much higher than that of pure natural rubber (~50 J·m⁻²), the permanently entangled hydrogel (~200 J·m⁻²), and the regular hydrogel (~10 J·m⁻²). Plotted on an Ashby map of fatigue threshold versus fracture energy, the transiently entangled hydrogel outperforms most existing tough hydrogels (Fig. 3f)^{12,14,27,28,31,54–58}. Notably, even materials with appreciable fracture energies, such as natural rubber (~10⁴J·m⁻²) and the permanently entangled hydrogel (~1460 J·m⁻²), exhibit much lower fatigue thresholds, underscoring the exceptional fatigue resistance enabled by the transiently entangled network.

To optimize the mechanical properties, transiently entangled hydrogels were developed with varying values of m and n. The ultimate stress is maximized at m = 9.6‰ and n = 4.2 M (Supplementary Table 5). When m is varied from 1.2‰ to 14.4‰, both tensile strength and fracture work first increase and then decrease, reaching maximum values of 1.06 MPa and 10.65 MJ·m⁻³ at m = 9.6‰, while the strain decreases from 5071% to 2666% (Fig. 3g–i). While the modulus typically correlates with cross-linker content, in this case, it first increases and then decreases with higher cross-linker concentrations. As noted, both chemical cross-links and dynamic cross-links contribute to the modulus. Cross-linker content below the threshold favors the formation of dangling chains, while above the threshold, it results in more chemical cross-links. This interplay leads to the unique modulus behavior observed in these hydrogels. These hydrogels also exhibit distinct yielding behavior. At m = 4.8‰ and n = 4.2 M, yielding is most pronounced, with a maximum yield strength of 0.07 MPa. At m = 14.4‰, the yielding phenomenon becomes much less noticeable, suggesting that disentanglement of transient entanglements is the main origin of the yielding behavior. Moreover, as the strain rate increases, transient entanglements do not have sufficient time to disentangle and therefore behave as effective load-bearing constraints, increasing the modulus and strength while reducing the fracture strain (Supplementary Fig. 13). The effect of n on the mechanical properties shows analogous non-monotonic trend. The modulus increases while stretchability decreases, with strength and fracture work peaking near n = 5.0 M and then falling due to network densification at high solids (Supplementary Fig. 14).

As far as we know, pure PAAm hydrogels typically cannot achieve a tensile strength above the megapascal level, nor do they display strains at break exceeding 5000%. The transiently entangled hydrogels in this study represent a significant advancement for PAAm-based hydrogels, offering both ultra-high strength and strain. Additionally, all transiently entangled hydrogels exhibit high compression strength (~1.0 MPa at 90% strain) and quickly return to their original shape once the compressive stress is removed (Supplementary Fig. 15). At small strains, the transient entanglements dissipate energy by sliding and uncoiling twisted dangling chains, while the network chains absorb energy as they straighten from a curved to a linear configuration. At larger strains, more transient entanglements are untangled, and the chemical cross-links begin to rupture. These unique behaviors contribute to the hydrogels’ superior mechanical properties. Moreover, these hydrogels’ storage modulus exhibits an order of magnitude higher than its loss modulus (Supplementary Fig. 16), and it is accompanied by linear viscoelasticity and a low apparent activation energy of only 34 kJ·mol⁻¹ (Supplementary Fig. 17). Some single-covalent-network polyelectrolyte hydrogels can also be constructed using this strategy to introduce transient entanglements, thereby improving comprehensive mechanical properties (Supplementary Fig. 18).

Tough transiently entangled hydrogels with multifunctions

Tough transiently entangled hydrogels with tunable mechanical properties and numerous hydrophilic dangling polymer chains can serve as an effective lubricating coating that lubricate the hydrogel surface and reduce friction. Owing to their much more dangling chains compared to a regular hydrogel, the transiently entangled hydrogel achieves a significantly lower friction coefficient of 0.0025 (Fig. 4a). This value is approximately 2.7 times lower than that of the permanently entangled hydrogel, 9.6 times lower than that of the regular hydrogel, and even 16 times lower than some common plastics⁵⁹ (Supplementary Figs. 19, 20). The combination of low friction, high toughness, and fatigue resistance allows this hydrogel to exhibit good resistance to wear (Fig. 4b, c). While the regular hydrogel fails after just 3.5 h of shearing in a rheometer test, the transiently entangled hydrogel remains intact after 6 h, demonstrating a superior wear resistance compared to most elastomers and hydrogels¹². These make the transiently entangled hydrogel ideal for modifying the surfaces of medical devices. To evaluate its lubrication performance, we tested the hydrogel films under various conditions, including sliding frequency and applied load (Fig. 4d, e). Both factors affect the friction coefficient, but it consistently remains low. Continuous friction tests at 1 N and 1 Hz show that the hydrogel maintains lubricity for at least 480 cycles (Fig. 4f), with a friction coefficient of approximately 0.0025, demonstrating the long-term stability of the hydrogel as a lubricant.

Fig. 4. Demonstration for the potential application of the transiently entangled hydrogel as friction and lubrication coatings.

a The friction coefficients versus time upon successive shearing of the three types of hydrogels, the F_N represents the normal force. Micrograph of b a transiently entangled hydrogel and c a regular hydrogel before b1, c1) and after shearing b2, c2) and stretch b3, c3), respectively. Plot of the friction coefficients versus d various applied loads (frequency: 1 Hz) and e various sliding frequencies (load:1 N). Data with error bars are presented as mean values ± standard deviations of n = 3 replicates. f The long-term friction coefficient curve, maintaining lubrication for at least 480 cycles. g–k Photographs (g, j) and corresponding micrographs of cross-section (h, k) of PVC catheter (g, h) and latex catheter (j, k) during the application of the lubrication coating. The friction coefficient curves of bare catheters and catheters modified with hydrogel coating: i PVC catheter, l latex catheter. m Schematic illustration of the lubrication mechanism. The intensity of the hydroxyl peak (3300 cm⁻¹) on the infrared spectra of the transiently entangled hydrogel (n) and regular hydrogel (o) after friction testing for different times. All transiently entangled hydrogels used in this section are LMC-4.8-AAm-3.0 samples.

To further assess the lubrication properties, the hydrogel was used to modify the surfaces of catheters. Using water as a lubricant to analyze the lubricating properties of two hydrogel-modified catheter. The average friction coefficients of the bare PVC catheter (Fig. 4g–i) and the latex catheter without gel coating (Fig. 4j–l) were 3.21 and 1.73, respectively. After applying the hydrogel coatings, the friction coefficients were reduced to 0.037 for the PVC catheter and 0.041 for the latex catheter, demonstrating the good lubrication performance of the modified catheters. Both catheters exhibit conformal and continuous contact with the coatings at the interface. A direct comparison between the transiently entangled hydrogel and a composition-matched regular hydrogel underscores the factors governing lubrication. In the transiently entangled hydrogel, densely distributed and hydrophilic dangling chains stabilize a water-rich interfacial film, providing fluid-film lubrication and yielding an ultra-low friction coefficient (Fig. 4m). Crucially, the transiently entangled hydrogel retains this hydration layer: after a continuous 4 h sliding test, it preserved 90.5% of its initial water content, whereas the regular hydrogel retained only 70.7% (Fig. 4n, o). These results confirm that a persistent surface-hydration layer is the key to the transiently entangled hydrogel’s remarkable, long-lived lubricity.

Discussion

In summary, we report a simple and general method for the fabrication of a single-covalent-network PAAm hydrogels with numerous transient entanglements that can slip and disentangle to dissipate energy upon deformation. This design markedly differs from regular hydrogels where chemical cross-links have nearly the same number as trapped entanglements, and permanently entangled hydrogels where trapped entangled chains can only slip until fracture occurs. Our quantitative analyses elucidate the individual contributions of both transient entanglements and chemical cross-links to the physicochemical properties of these hydrogels, demonstrating that different topological networks yield distinct material characteristics. The resultant transiently entangled hydrogels effectively resolve the longstanding strength–toughness conflict, achieving high toughness and strength alongside good fatigue resistance, ductility, transparency, low friction and high wear-resistance. These properties make them promising candidates for lubricants. This simple design strategy is potentially applicable to other monomers that meet similar criteria described herein. We anticipate that employing transient entanglements will enable the transformation of otherwise fragile materials into strong, tough, and fatigue-resistant systems suitable for applications such as adhesives, low-friction coatings, and wearable devices.

Methods

Materials

Polyallylamine (PAA, Mn = 15,000, 30 wt%) was purchased from Zouping Mingyuan Import & Export Trading Co., China. Acrylamide (AAm, ≥99%) was provided by Chengdu Huaxia Chemical Reagent Co., Ltd, China. Allyl glycidyl ether (AGE, 99%), potassium persulfate (KPS, ≥99.5%), tetramethylenediamine (TMEDA, 99%), N,N′-methylenebis(acrylamide) (MBA, ≥99%), deuterium oxide (D₂O, 99.9%) were purchased from Adamas, China. Deionized water (H₂O) was obtained from Chengdu Chron Chemical Co., China. Hydroxy silicone oil was received from Wacker Chemical Co., Ltd, Germany. All the chemicals and solvents were used as received without further purification.

Preparation of linear macromolecular cross-linker (LMC)

LMC was prepared in three steps. Firstly, 3.39 g of PAA and 100 g of H₂O were added into a 250 mL three-necked bottle and stirred for 10 min. Then, 2.0 g of AGE was added into the mixed solution and the solution was heated to 60 °C for 6 h with mechanical stirring. Lastly, the ay-prepared dispersion was dialyzed in a dialysis bag with a molecular weight cutoff of 3000 Da for two weeks with refreshing deionized water once a day for subsequent use. The final mass concentration of LMC solution is 1.38%.

Preparation of hydrogels

The typical transiently entangled hydrogels were prepared by free in situ radical polymerization. Prescribed amounts of AAm, LMC solution, TMEDA were dissolved in deionized water. After dissolving completely, a vacuum pump was used to remove the dissolved oxygen for 15 min. After that, a given amount of KPS aqueous solution as initiator was added into the precursor solution and stirred for another 2 min. At last, the reaction solution was transferred to a reaction box consisting of the organic glass mold with 1 mm spacing and kept at 25 °C for 12 h. The regular hydrogel and permanently entangled hydrogel were prepared with reference to Suo et al. The synthesis steps are consistent with those of the transiently entangled hydrogels. The detail compositions of these hydrogels covered in this work are shown in Supplementary Table 1.

Nuclear magnetic resonance spectra (¹H NMR) test

The ¹H NMR of PAA, AGE and LMC were recorded on a spectrometer operating (AV III HD 400 MHz, Bruker, Germany). Before test, the LMC solution was completely lyophilized in a vacuum freeze dryer and then dissolved in D₂O.

¹H double-quantum NMR (¹H DQ NMR) test

The ¹H DQ NMR experiments were performed at 9.4 T magnetic field on a JEOL JNM-ECZR400R/M1 spectrometer, equipped with a 3.2 mm HXY MAS probe (JEOL RESONANCE Inc., Japan). The ¹H π/2 pulse length is about 2.2 μs. A well-established offset-compensated Baum-Pine pulse sequence was adapted to excite double-quantum (DQ) coherences. For ¹H Hahn Echo (spin echo) mode, using 90° pulse width of 2.3 μs, 180° pulse width of 4.6us, the upper and lower limits of the sampling time (τ) are 0.002 ms and 500 ms, respectively, and take 39 sampling points.

Rheological test

Rheological behaviors of the hydrogels were analyzed by a rotational rheometer (AR2000ex, TA Instrument, USA) with a suit of 40 mm diameter parallel-plates. The thickness of the samples was 1 mm. To keep hydrogels hydrated, the samples were smeared with a thin layer of low-viscosity silicone oil during the testing. Stress relaxation experiments were carried out with a constant strain of 5% for 60 min at 25 °C. Frequency sweeps with variable temperature ranging from 25 to 75 °C were performed to the sample at an immobile strain of 1%. The apparent activation energies were calculated from the slope of the master curves according to Arrhenius’ plot.

Swelling ratio and cross-linking density

The swelling ratio and cross-linking density of the hydrogel was calculated based on the equilibrium swelling method. The sample with a thickness of 1 mm was cut into small cubes of about 10 × 10 mm and then placed in deionized water for swelling, during which the change in the mass of the sample was recorded at regular intervals. Until the quality of the hydrogel no longer changes, to reach the equilibrium swelling. The swelling ratio ($Q_w$) is calculated as $Q_w=\left(W_S/W_d\right)\times 100\%$, where $Q_w$ is the equilibrium swelling ratio, $W_s$ is the mass of the hydrogel after equilibrium swelling, and $W_d$ is the mass of the dry gel before swelling test. The cross-linking density can be calculated according to the Flory-Rehner equation. Details are show in Supplementary material.

Dynamic light scattering (DLS) test

DLS experiments were performed using a BI-200SM instrument manufactured by Brookhaven, USA, using a He-Ne laser (2 mW, wavelength λ of 633 nm) with a detection angle of 90°, and the instrument was fitted with a rotating stage. The time-averaged scattering intensity 〈I〉_T and the time-averaged correlation function (ICF) were tested by rotating the rotating stage arbitrarily at 100 different positions for 30 s for each run, and the temperature was kept constant during the test. The overall average scattering intensity 〈I〉_E, defined as the average of the time-averaged scattering intensity 〈I〉_T measured at 100 points, consists of two components, which can be expressed by the following equation: ${⟨I_C\left(p\right)⟩}_E={⟨I⟩}_E-{⟨I_F⟩}_T$, 〈I_F〉_T is the time-averaged scattering intensity caused by the dynamically fluctuating component, similar to the concentration fluctuation of the liquid component, which is ergodic in nature, and ${⟨I_C\left(p\right)⟩}_E$ is the static time-averaged scattering intensity caused by spatial inhomogeneities arising from cross-linking processes. Details are show in Supplementary material.

Mechanical properties tests

All the mechanical properties of the hydrogels were measured by a commercial tensile tester (5567, Instron, USA). For the uniaxial tensile test, the as-prepared dumbbell shape hydrogel samples with a length of 20 mm, a width of 2 mm, and a thickness of 1 mm were tested at room temperature with a given crosshead velocity. The tensile rate is 100 mm/min when not otherwise specified.

For pure shear test, the samples were cut into a rectangle shape with a cross-section of 35×1 mm (i.e. a₀ × b₀), the gauge length is 7 mm and the tensile rate is 100 mm/min. The fracture energy was calculated via equation: $\Gamma =\mathrm{U}\left(\mathrm{Lc}\right)/\left(a_0\times b_0\right)$, where $U\left(Lc\right)$ denotes the work done to stretch the unnotched sample to $Lc$, $Lc$ denotes the tensile displacement of the notched sample (notch length: 17.5 mm) at the onset of crack expansion, a₀ and b₀ denote the width and thickness of the sample, respectively.

For fatigue resistance test, fatigue thresholds were tested via the single-notch method in a customized temperature and humidity controlled closed acrylic plate. The size of the rectangular gel samples used was consistent with the pure shear test. The sample was prepped with a 7-mm notch and subjected to continuous loading-unloading cyclic tests at 600% strain (ε), and the value of the change in notch size was recorded (c). This value was then divided by the number of cycles N to give the crack extension value (dc/dn) for each cycle. The energy release rate (G) was calculated according to the equation G = 2kcW. where k is a function of strain, empirically determined to be $k=3/{\left(\epsilon +1\right)}^{1/2}$, c is the crack length, and W is the work of fracture of an unnotched sample of the same size stretched to the same strain ε. The critical energy release rate (G), i.e., the inflection point of the dc/dn vs. G curve is the fatigue threshold.

Friction coefficient test

The friction coefficient of the hydrogel was measured using a rheometer consistent with the rheological tests. The torque was measured at a specific applied load and frequency. The friction coefficient (µ) was calculated by the following equation: $\mu =8T/3DP$, where T is the torque, D is the diameter of the parallel-plates, and P is the applied axial force.

Lubrication coating test

Before testing the friction coefficient, PVC catheter and latex catheter were washed with alcohol and deionized water under ultrasound conditions for 40 min, which were followed by drying with N₂ flow for 2 min. Subsequently, both catheters were cut into 8 mm diameter discs using a scalpel. The friction coefficients of the outer surfaces of the samples were tested using a rheometer. In order to modify the surface of the catheter via hydrogel coating, we placed the catheter in a customized ring column consisting of 2 layers of cylindrical glass. Then, the reaction precursor solution was transferred into the ring column and left at 25 °C for 12 h. After the reaction was completed, the modified catheter with gel coating was cut into 8 mm diameter discs and tested the friction coefficient using the same method.

Other tests

Ultraviolet absorption spectroscopy (UV 3600, Shimadzu, Japan) in diffuse reflection mode was used to characterize the optical transparency of hydrogel. Attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR, Nicolet 6700, Thermo Scientific, USA) was used to test the water content on the surface of the hydrogel. The micrograph of the sample was observed by Scanning electron microscopy (SEM, ThermoFisherScientific, Apreo S HiVov, USA), atomic force microscope (AFM, Smart SPM, AIST-NT, USA) and Polarizing microscope (Leica, DM4P, Germany).

Author contributions

Z.Y. and J.W. conceived the idea. Z.C. and Z.Y. designed the research. Z.Y. performed most of the experiments. H.W. (H. Wang) and R.Z. performed the ¹H DQ NMR test. H.W. (H. Wu) participated in the NMR analysis. J.Z. performed the lubrication test. Z.Y., J.Z. and J.W. analyzed the data and wrote the manuscript with the input from all authors.

Peer review

Peer review information

Nature Communications thanks Jinhwan Yoon, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

All data supporting the findings of this study are available within this article and the Supplementary Information. All data are available from the corresponding author upon request. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-70194-9.