Lignin-dynamic networks break the strength–toughness trade-off in thermoplastic elastomers Authors: Peng Zhao, Tianwei Wu, Jinsong Sun, Meng Liu, Qi Tang, Zechun Ren, Yongzhuang Liu, Haipeng Yu Published: 2026 ABSTRACT AbstractThe long-standing trade-off between strength and toughness has constrained the performance of thermoplastic elastomers. Here, we address this challenge by constructing a three-dimensional supramolecular network in a thermoplastic polyurethane elastomer via dynamic chain extension with lignin polyols. Unlike conventional linear or branched extenders, lignin’s rigid aromatic backbone and abundant hydroxyl groups serve as multifunctional nodes that simultaneously enhance microphase separation and establish a hierarchical, dynamic hydrogen-bonding network. At an optimal lignin loading of 10 mol%, the resulting material achieves a tensile strength of 127.7 ± 9.52 MPa and a toughness of 1.43 ± 0.11 GJ m⁻³—representing the highest combination reported for polyurethane elastomers to date. In situ spectroscopy and molecular dynamics simulations reveal that lignin transforms the hydrogen-bond landscape from a static binary system into a dynamic, spatially organized architecture, enabling multi-scale energy dissipation through sacrificial bond rupture and subsequent network reorganization. This lignin-engineered network also imparts superior damage tolerance and high elastic recovery to fiber-reinforced composites. This work unveils a general strategy for employing bio-aromatic motifs to effectively decouple the strength-toughness conflict in thermoplastic elastomers, paving the way for high-performance composites.Subject terms: Organic molecules in materials science, Mechanical properties, Polymers Introduction Achieving a synthetic elastomer that combines high strength with high toughness remains a premier challenge in polymer science1–4. This need has become acute for emerging applications in flexible armor, aerospace structures, and biomedical implants5–9. Thermoplastic polyurethane elastomers (TPUEs) are promising candidates due to their tunable microphase-separated structures and excellent processability10–12. Yet the essentially linear or weakly branched architectures of normal TPUEs enforce a strength–toughness trade-off: strategies that raise strength almost invariably diminish toughness, and vice versa13,14. Conventional strategies for addressing this dilemma typically modify intermolecular interactions using linear chain extenders (e.g., dihydrazides) that generate dense hydrogen-bonding networks15–18. Although the resulting materials can reach tensile strengths exceeding 75 MPa, their toughness remains limited (~390 MJ m⁻³)19,20. Branched extenders introduce additional topological constraints; however, they often compromise overall mechanical performance21,22. The incorporation of dynamic covalent bonds enhances toughness and imparts self-healing capability, but the tensile strength rarely exceeds 40 MPa23–27. Bio-derived extenders (e.g., castor oil and tannic acid) improve sustainability and introduce dynamic interactions, yet their mechanical performance remains modest, with strengths below 50 MPa and toughness below 120 MJ m⁻³ 28–33. Critically, none of these strategies enables the deliberate construction of a three-dimensional network required for efficient multiscale energy dissipation; consequently, the strength–toughness trade-off persists. Lignin, the most abundant aromatic biopolymer on Earth, represents a promising alternative. Its rigid three-dimensional architecture, high density of hydroxyl groups, and π-conjugated motifs make it a multifunctional modifier that can simultaneously serve as a multi-arm cross-linker, rigid segment, and supramolecular interaction node34–40. Nevertheless, in lignin-based polyurethanes, these attributes have thus far been exploited primarily for UV-shielding or photothermal applications41, while the resulting materials exhibit mechanical strengths below 100 MPa and toughness below 1.0 GJ m⁻³ 42,43. Fully harnessing lignin’s structural advantages to simultaneously advance both properties to unprecedented levels remains an important objective44. Here, we address this challenge by incorporating hydroxyl-rich lignin polyols as spatially distributed, dynamic chain extenders into a TPUE matrix. Rather than serving as a passive biofiller, lignin functions as an integral network architect, constructing a reprocessable three-dimensional supramolecular network that integrates dense yet dynamic hydrogen bonding, enhanced microphase separation, and π–π interactions (Fig. 1). As a result, the lignin-based TPUE (Lig-TPUE) achieves a tensile strength of 127.7 ± 9.52 MPa and a toughness of 1.43 ± 0.11 GJ m⁻³, setting a new benchmark for bio-based elastomers. Furthermore, Lig-TPUE serves as a high-performance matrix for fiber-reinforced composites, delivering excellent damage tolerance while enabling closed-loop recyclability. These findings establish a generalizable bio-aromatic network-engineering strategy for effectively overcoming the strength–toughness trade-off in thermoplastic elastomers, thereby opening a sustainable pathway toward next-generation protective and structural materials. Recyclable bio-aromatic polyol dynamic networks that overcome the strength–toughness trade-off in thermoplastic elastomers. The schematic was conceived by the authors and created using ChemDraw Professional (Version 18.1.2.18; PerkinElmer Informatics), and Adobe Illustrator 2025 (Version 29.5; Adobe Inc.). Results A hierarchical and dynamic hydrogen-bonding network engineered by lignin polyols To overcome the intrinsic strength–toughness trade-off, we re-engineered the TPUE by introducing lignin polyols. Lignin polyols were isolated from poplar wood using a tailored, recyclable ternary deep eutectic solvent (DES)45,46. This environmentally benign, single-step process enables both efficient extraction and in situ graft modification of lignin. Specifically, glycerol molecules proactively graft onto the α-position of the lignin side chains and partially substitute specific phenolic hydroxyl groups on the guaiacyl units (Supplementary Figs. 1–3). As a result, the native lignin is effectively converted into highly reactive lignin polyols. Compared to the poplar milled wood lignin (MWL), which possesses a limited total hydroxyl content of 6.61 mmol g⁻¹ (aliphatic OH: 5.17 mmol g⁻¹, phenolic OH: 1.44 mmol g⁻¹), the DES-extracted lignin exhibits an enhanced total hydroxyl content of 10.92 mmol g⁻¹ (aliphatic OH: 8.62 mmol g⁻¹, phenolic OH: 2.30 mmol g⁻¹; Supplementary Fig. 4 and Supplementary Table 1). Furthermore, the lignin polyols exhibit an ideal oligomeric architecture with a moderate molecular weight (Mn = 971 g mol⁻¹, Mw = 3497 g mol⁻¹, PDI = 3.601; Supplementary Fig. 5 and Supplementary Table 2). The combination of this moderate molecular weight and the dense reactive aliphatic hydroxyls confers excellent solubility and high reactivity, effectively preventing phase heterogeneity during polymerization. Simultaneously, this approach allows for large-scale production (Supplementary Fig. 6). Consequently, these polyols were successfully employed as multifunctional, spatially dynamic chain extenders to re-engineer the TPUE, as illustrated in Supplementary Figs. 7–9. DESH was selected as a dual-functional chain extender to construct high-strength and high-toughness elastomers, overcoming the limitations of conventional extenders that typically provide only a single type of dynamic interaction. Structurally, DESH integrates dynamic covalent disulfide bonds, which act as sacrificial elements to efficiently dissipate mechanical energy under stress. Concurrently, its terminal acylhydrazide groups react with isocyanates to introduce a high density of N–H and C=O functionalities. This promotes the spontaneous formation of a robust, dynamically reversible supramolecular hydrogen-bond network. Ultimately, this synergistic dual-network design maximizes mechanical energy dissipation and provides robust structural reinforcement. This bio-aromatic building block does not merely graft onto the polymer backbone; instead, it constructs a three-dimensional supramolecular framework with a hierarchical, dynamic hydrogen-bonding network (Fig. 2a). Spectroscopic analyses confirm the successful synthesis and covalent incorporation of lignin polyols. In the 1H NMR spectrum of Lig-TPUE, the lignin methoxy resonance at 3.7–3.8 ppm and a broad signal at 11.7 ppm, attributable to protonated aromatic moieties, verify lignin’s role as a chain extender (Fig. 2b and Supplementary Fig. 10)47,48. Complementary FTIR spectra show complete consumption of isocyanate groups, evidenced by the absence of the –NCO band at 2200–2300 cm⁻¹ (Supplementary Fig. 11). Lignin polyols engineer a dense, dynamic, and hierarchical hydrogen-bonding network. a Schematic of Lig-TPUE synthesis. b 1H NMR spectra comparing covalent incorporation of Control-TPUE and Lig-TPUE. c, d FTIR spectra showing increased disordered hydrogen bonds (c) and enhanced hydroxyl hydrogen bonding (d) in Lig-TPUE. e, h Temperature-dependent FTIR spectra of the carbonyl (e) and hydroxyl (h) regions revealing dynamic bond behavior. f, g, i, j Synchronous (f, i) and asynchronous (g, j) 2D-COS spectra establishing the sequential dissociation-reorganization mechanism of hydrogen bonds. Lignin polyols fundamentally reconfigure the hydrogen-bonding landscape (Fig. 2c, d). FTIR further reveals that lignin substantially reorganizes the hydrogen-bonding environment. In the carbonyl stretching region (1800–1600 cm⁻¹), Lig-TPUE exhibits a higher fraction of disordered hydrogen bonds centered at 1640 cm⁻¹ compared with neat TPUE (Fig. 2c; Supplementary Fig. 12). This indicates that the rigid, three-dimensional lignin polyols increase segmental spacing and disrupt crystalline H-bond domains. Concurrently, the hydroxyl stretching region (3700–3100 cm⁻¹) shows an intensified band at 3300 cm⁻¹ in Lig-TPUE (Fig. 2d; Supplementary Fig. 13), signifying a substantial rise in both the density and strength of hydrogen bonds contributed by lignin’s abundant hydroxyl groups. In-situ variable-temperature FTIR elucidates the network’s dynamic behavior (Fig. 2e). Upon heating, weak, disordered hydrogen bonds at 1640 cm⁻¹ dissociate first, enhancing chain mobility. This liberation promotes hard-segment aggregation and the subsequent formation of stronger, ordered hydrogen bonds at 1680 cm⁻¹, as indicated by the growth of this band. Simultaneously, the O–H stretching band at 3300 cm⁻¹ in Lig-TPUE (Fig. 2h) exhibits a pronounced red shift and broadening, a response scarcely observed in the control, indicating that the hydrogen-bond network remains robust and reorganizes rather than collapses under thermal stress. Two-dimensional correlation spectroscopy (2D-COS) further clarifies the hierarchical, stimulus-responsive character of the network. Application of Noda’s rule to the C=O region reveals the sequential order 1640 cm⁻¹ (disordered H-bonds) → 1720 cm⁻¹ (free C=O) → 1680 cm⁻¹ (ordered H-bonds) (Fig. 2f, g). This progression confirms that weak, disordered H-bonds function as sacrificial links, dissipating energy first and triggering reorganization that reinforces the remaining network. Analysis of the N–H/O–H region (Fig. 2i, j) shows a similar sequence: free N–H groups respond earliest, followed by reconfiguration of the strong, lignin-enhanced hydrogen-bond network at 3300 cm⁻¹. Collectively, these spectroscopic insights establish a central design principle: integrating three-dimensional lignin polyols transforms TPUE’s hydrogen-bonding motif from a static, binary arrangement into a dynamic, hierarchical, and spatially organized network. This architecture creates multiscale energy-dissipation pathways: weak bonds rupture sacrificially to prevent catastrophic crack propagation while simultaneously facilitating the reinforcement of stronger bonds, thereby decoupling the coupled properties of strength and toughness. The formation of a hierarchical, dynamic hydrogen-bonding network within the elastomer is fundamentally governed by the synergistic assembly of specific molecular segments. Its hierarchical character arises from the ordered distribution of hydrogen-bond strength and density among different structural units. Specifically, the employment of DESH as a chain extender is critical; its acylhydrazide groups, alongside the resulting urea linkages, provide abundant N–H donors and C=O acceptors. This facilitates the formation of a high-density, strong hydrogen-bonding array within the rigid-segment domains. Concurrently, the polyphenolic and polyether architectures of the lignin polyols introduce a secondary, relatively weak hydrogen-bonding network. The extensive interactions between the hydroxyl groups of the lignin polyols and the polar groups in both soft and hard segments allow these two network levels to interpenetrate and couple. Ultimately, a continuous, hierarchical supramolecular network is established throughout the material matrix. Lignin-induced network constriction and enhanced microphase separation The spectroscopic analyses above establish that lignin polyols reconfigure the hydrogen-bonding landscape. We now show how this molecular redesign propagates to the mesoscale, reshaping both the thermomechanical behavior and the phase morphology of the elastomer. The data demonstrate that lignin functions as a multifunctional junction point that (i) constricts the network, (ii) elevates the hard-segment glass-transition temperature, and (iii) amplifies microphase separation, thereby generating a highly robust, multiscale cross-linking architecture. Thermal analysis directly reflects the constrained network dynamics. Differential scanning calorimetry (DSC) reveals that the soft-segment glass transition (Tg) is essentially unchanged, whereas the hard-segment Tg rises sharply from 59.8 °C in Control-TPUE to 79.4 °C in Lig-TPUE (Fig. 3a). This 20 °C shift indicates pronounced immobilization of the hard segments, consistent with their integration into the three-dimensional lignin-mediated network. DMA tanδ curves (Fig. 3b) further corroborated the thermomechanical behavior and phase separation. Both the blank sample and Lig-TPUE display two discrete relaxation transitions, symptomatic of a well-developed microphase-separated architecture. The low-temperature glass transition (Tg1), corresponding to the soft segment matrix, remains virtually unperturbed at approximately −57 °C, signifying that the lignin polyol is selectively partitioned into the hard phase without compromising soft-segment mobility. In stark contrast, the hard-segment transition (Tg2) exhibits a remarkable upward shift from 22.1 °C to 96.4 °C, demonstrating the powerful restricting effect of the rigid lignin skeleton and the formation of a dense, crosslinked network within the hard domains. Notably, the transition peak of Lig-TPUE retains a sharp and symmetric profile without discernible broadening, indicating a highly cooperative relaxation process and a homogeneous integration of the bio-polyol into the host polymer network. Lignin-induced network constriction enhances microphase separation. a DSC thermograms showing elevated glass transition (Tg) and melting (Tm) behaviors. b DMA curves indicating reduced segmental mobility and improved phase separation. c Crosslink density quantified by swelling, showing an order-of-magnitude increase. d WAXS patterns confirming amorphous nature of both materials. e, f 1D (e) and 2D (f) SAXS profiles demonstrating increased long-period spacing and enhanced microphase separation. g SEM, AFM, and TEM images visualizing hierarchical hard-domain morphology in Lig-TPUE. Rubber elasticity analysis of the plateau modulus reveals an order-of-magnitude increase in the effective crosslinking density, from 4.79 × 10² mol m⁻³ for Control-TPUE to 4.82 × 10³ mol m⁻³ for Lig-TPUE (Fig. 3c). This substantial increase confirms that lignin, functioning as a spatially trifunctional or higher-order chain extender, transforms a loosely entangled linear architecture into a partially cross-linked three-dimensional network. It should be noted that the crosslinking density calculated from DMA represents an effective network density rather than a direct measure of chemical crosslinking alone. In Lig-TPUE, lignin can react with isocyanate groups through its hydroxyl groups to form covalent urethane linkages, while its aromatic and hydroxyl-rich structure also promotes reversible physical associations, including hydrogen bonding, π − π interactions, and hard-domain aggregation. Therefore, the enhanced network density should be attributed to the combined contribution of covalent incorporation and physical association. Solvent solubility experiments further supported this interpretation (Supplementary Fig. 14). Lig-TPUE was highly soluble in DMF and THF, with soluble fractions of 99.93% and 99.90%, respectively, indicating that no continuous covalent network had formed. These results suggest that the enhanced mechanical properties primarily stem from lignin-mediated covalent incorporation and reversible physical associations, rather than from the formation of a fully crosslinked thermoset structure. This distinction explains the retention of excellent thermoplastic processability at elevated temperatures. Critically, this crosslinking does not induce undesired crystallization but redirects the phase-separation paradigm. Wide-angle X-ray scattering (WAXS) shows only broad halos for both materials (Fig. 3d), excluding crystallization as a major contributor to the observed stiffening49. To elucidate the nanoscale phase morphology, SAXS profiles were evaluated (Fig. 3e, f). Upon the incorporation of lignin polyol, the primary scattering peak (qmax) exhibited a subtle leftward shift from 0.6 to 0.5 nm⁻¹. According to the Bragg equation, this leftward shift corresponds to an expansion in the interdomain spacing (d) from ~10.5 to 12.6 nm. The structural expansion is primarily attributed to the substantial steric hindrance imposed by the bulky, rigid aromatic skeleton of lignin molecules that are chemically or physically incorporated into the hard domains. Concurrently, the abundant hydroxyl groups within the lignin structure promote the reorganization of local hydrogen-bonding networks, thereby maintaining a well-defined yet spatially expanded microphase-separated morphology. Multiscale microscopy substantiates this structural transition (Fig. 3g). Scanning electron microscopy (SEM) reveals island-like hard domains dispersed within the soft matrix—features absent in the control. Atomic force microscopy (AFM) registers a higher surface roughness (Ra = 1.29 nm) and a distinct alternation of hard and soft regions, attesting to minimal interfacial mixing. Transmission electron microscopy (TEM) resolves multi-layered nanodomains of the hard phase, evidencing a hierarchical architecture produced by lignin-enhanced hydrogen bonding and hard-segment aggregation. In summary, the incorporation of rigid, three-dimensional lignin polyols topologically constrains the TPUE network. This confinement elevates the hard-segment Tg, increases the effective cross-link density, and thermodynamically favors a more pronounced and hierarchical microphase separation. Together, these structural evolutions underpin the emergence of a mechanically robust, toughened elastomeric material. Atomic-scale visualization of the lignin polyols-reinforced network Spectroscopic and scattering measurements indicated extensive network reconfiguration. To resolve this architecture at atomic resolution and quantify the intermolecular forces underlying the mechanical enhancement, we performed all-atom molecular dynamics (MD) simulations. The simulations provide an unambiguous picture: lignin polyols behave as three-dimensional junctions that markedly reinforce the matrix and homogenize its hydrogen-bond network. Models were built and equilibrated using standard protocols (Supplementary Methods). The Lig-TPUE shows a substantially lower total potential energy (−7665 kcal mol⁻¹) than the Control-TPUE (−7350 kcal mol⁻¹; Fig. 4a), confirming its superior thermodynamic stability. As illustrated in Fig. 4b and Supplementary Fig. 15, Lig-TPUE exhibits a significantly denser spatial distribution of oxygen atoms compared to the Control-TPUE. This enriched oxygen density provides an abundant array of hydrogen-bond acceptors within the hard-segment domains. The spatial distribution of nitrogen atoms, which are integral components of the urethane linkages, also shows highly coordinated dense packing within the hard segments (Supplementary Fig. 16). This localized and cohesive clustering of both N and O atoms confirms that lignin incorporation strongly restricts the mobility of the entire hard-segment microdomain, thereby firmly anchoring the urethane bonds and suppressing their dissociation under stress. Molecular dynamics simulations reveal the atomic-scale origins of network reinforcement. a Total binding energy showing enhanced thermodynamic stability of Lig-TPUE. b Spatial distribution maps of oxygen atoms, indicating homogeneous enrichment from lignin hydroxyl groups. c Binding energy contribution identifying lignin domains as primary structural anchors. d Number and spatial distribution of hydrogen bonds. e Distribution of hydrogen-bond lengths, demonstrating stronger and more uniform interactions in Lig-TPUE. f, g Cohesive energy density (f) and total cohesive energy (g), quantifying superior bulk cohesion. To quantitatively elucidate the thermodynamic driving forces underlying this network densification, DFT calculations were performed on representative hard-segment fragments (Fig. 4c). The binding energy between conventional DESH chain extenders is −7.20 kcal mol⁻¹. Notably, the incorporation of lignin enhances the thermodynamic affinity: the binding energy of the DESH–lignin pair increases to −8.74 kcal mol⁻¹, while that of the lignin–lignin pair reaches −12.28 kcal mol⁻¹. To ensure that these results accurately reflect the physicochemical nature of technical lignin, we explicitly considered its inherent structural heterogeneity. Because steric hindrance often prevents complete hydroxyl conversion during practical synthesis, we calculated binding energies for a range of structurally distinct chain-extender configurations, including models containing unreacted (exposed) hydroxyl groups (Supplementary Fig. 17). These models consistently exhibited high binding energies, confirming that the strong thermodynamic anchoring effect is retained despite local steric variations. Collectively, the rigid aromatic backbones and the diverse modes of hydroxyl participation synergistically promote more compact and thermodynamically stable hard-segment packing50. This enhanced thermodynamic affinity directly translates into a more robust and dynamic physical crosslinking network, as further validated by MD trajectory analysis. Figure 4d quantitatively tracks the number of intermolecular hydrogen bonds over a 100-frame production run. Whereas Control-TPUE fluctuates between 250 and 280 hydrogen bonds, Lig-TPUE sustains a consistently higher and more stable population of ~300–310 bonds. Furthermore, the probability distribution of hydrogen-bond lengths (Fig. 4e) reveals a critical structural transition. For Control-TPUE, the peak bond length is centered at 2.0–2.1 Å. By contrast, the distribution for Lig-TPUE is clearly shifted toward shorter distances, with a sharp peak at 1.8–1.9 Å and a broader probability contour in the short-range interaction region. Given that hydrogen-bond strength is inversely related to bond length, this leftward shift demonstrates that lignin incorporation not only increases the number of noncovalent crosslinks but also strengthens their specific intermolecular interactions1. The cumulative effect of these intensified inter- and intrachain interactions governs the macroscopic mechanical robustness of the elastomer. As quantified computationally in Fig. 4f, g, the cohesive energy density (CED) increases from ~410 J cm⁻³ in Control-TPUE to more than 520 J cm⁻³ in Lig-TPUE. At the same time, the total cohesive energy (Ecoh) rises from ~2000 to 2500 kJ mol⁻¹, corresponding to an enhancement of nearly 25%. Cohesive energy provides a direct theoretical indicator of macroscopic resistance to chain pullout and structural yielding. The substantial increase in Ecoh therefore clarifies the fundamental strengthening mechanism of Lig-TPUE: the dense, lignin-anchored hydrogen-bonding matrix functions as an efficient sacrificial energy-dissipation network. Under external deformation, considerably greater mechanical work is required to disrupt these strong, cooperative noncovalent interactions, thereby endowing the material with superior tensile strength and toughness51. Our multi-model simulations reveal a crucial structural advantage that unreacted, exposed –OH groups are not inert defects. Instead, they act as potent secondary physical cross-linking nodes. This heterogeneous physical entrapment promotes synergistic stress dissipation and contributes significantly to the high cohesive energy and mechanical toughness observed in Lig-TPUE. Tunable mechanics and record-high toughness by this lignin polyols incorporation strategy The molecular-level redesign described above translates directly into macroscopic behavior that transcends the classical strength–toughness trade-off. Mechanical performance is highly tunable and reaches a maximum at a critical lignin loading. Systematic variation of lignin content shows that both strength and toughness peak at 10 mol%; lower contents fail to establish a continuous reinforcing network, whereas higher contents introduce excessive heterogeneity that degrades performance (Fig. 5a). When 10 mol% lignin polyols are incorporated, the elastomer (Lig 10%-TPUE) attains a tensile strength of 127.7 ± 9.52 MPa and a toughness of 1.43 ± 0.11 GJ m⁻³ (Fig. 5b; Supplementary Tables 3 and 4). Such high strength-to-weight ratio can be illustrated by a simple demonstration: a 0.1 g filament can hoist a 5 kg mass—50,000 times its own weight (Fig. 5c). This combination reaches the highest record for polyurethane elastomers (Fig. 5d and Supplementary Table 5). Record-high strength and toughness with maintained elasticity and processability. a, b Stress-strain curves (a) and corresponding toughness values (b) for TPUEs with varying lignin content, showing optimal performance at 10 mol%. c Photograph of a 0.1 g Lig-TPUE filament lifting a 5 kg weight (50,000× self-weight). d Ashby plot comparing Lig-TPUE with reported polyurethanes, establishing a new performance benchmark. e, f Cyclic tensile tests at 100% (e) and up to 500% strain (f) exhibiting elastic recovery. g True stress–strain curve revealing intrinsic material strength. h Force-displacement curve from tensile testing. i Notched vs. unnotched stress-strain curves showing notch-insensitive toughness. j, k Demonstrations of processability: solution-casting (j) and 3D-printed geometries (k). To elucidate the mechanical deterioration observed at higher lignin polyol contents, we performed morphological characterization of Lig 15%-TPUE samples, which exhibited reduced performance (Supplementary Fig. 18). AFM analysis revealed a high surface roughness (Ra = 11.8 nm), while TEM images confirmed the presence of extensive, irregular hard-segment aggregates within the matrix. This morphological evolution suggests that, at higher loadings, excessive lignin-induced hydrogen-bonding interactions exceed the dispersion limit and lead to uncontrolled hard-segment aggregation. These heterogeneous domains disrupt the continuity of the elastomeric matrix and act as stress-concentration sites, thereby directly causing premature failure and the observed deterioration in tensile properties. Therefore, 10 mol% represents an optimal composition window in which lignin functions most effectively as a multifunctional network hub. To decouple the synergistic reinforcement effect of lignin polyols from the concentration effect of the chain extender, control experiments were conducted using different DESH dosages (0.358 g or 1.50 mmol; 0.834 g or 3.50 mmol) while maintaining a constant NCO/OH ratio (Supplementary Fig. 19). The deterioration in mechanical properties observed in both divergent groups unambiguously demonstrates that the pronounced strengthening and toughening effects arise from the precisely reconstructed hierarchical network enabled by the three-dimensional architecture of lignin, rather than from simple variations in hard-segment density. Notably, this high toughness is achieved without sacrificing elasticity. The first loading–unloading cycle shows moderate hysteresis, which can be attributed to the disruption of weak, disordered hydrogen bonds. From the second cycle onward, the hysteresis becomes negligible, indicating that the network reorganizes into a more stable configuration (Fig. 5e and Supplementary Fig. 20). This unique combination of high strength, exceptional toughness, and high elastic recovery after the initial conditioning cycle confirms the successful decoupling of network parameters: dense and dynamic cross-links impart strength and tear resistance, whereas hierarchical hydrogen bonding and enhanced microphase separation enable large, reversible energy dissipation. Although the material exhibits excellent resilience with minimal hysteresis at low strains, a pronounced hysteresis loop appears at high strains (e.g., 1500%). The time-dependent recovery of the hysteresis loop further confirms the dynamic and reversible nature of these non-covalent interactions. Collectively, these results demonstrate that internal energy-dissipation mechanisms play a critical role in endowing Lig-TPUE with its outstanding mechanical properties (Supplementary Fig. 21). To further evaluate the elastic resilience and energy-dissipation mechanism of Lig-10%-TPUE under extreme deformation, cyclic tensile tests were performed at ultrahigh strains of up to 2200% (Fig. 5f and Supplementary Fig. 22). At a moderate strain of 500%, the material displayed excellent rubber-like elasticity, retaining 84.5% of its maximum stress and achieving an elastic recovery ratio of 97.4% after five cycles (Supplementary Table 6). At strains of 1000% and 1500%, Lig-TPUE likewise maintained strong elastic recovery. However, as the applied strain increased to an extreme 2200%, the stress retention and elastic recovery ratio decreased to 38.1% and 61.0%, accompanied by a markedly enlarged hysteresis loop. This strain-dependent cyclic behavior is a hallmark of the Mullins effect and is fundamentally governed by the dynamic sacrificial-bond mechanism. To accommodate such a large macroscopic deformation, reaching 22 times its original length, without rupture, a substantial number of dynamic non-covalent interactions must dissociate to dissipate the applied energy. This extensive energy dissipation prevents catastrophic scission of the polymer backbone and thus explains the material’s extraordinary elongation at break of ~3000%. Despite undergoing substantial network reconfiguration and energy dissipation, Lig-10%-TPUE withstood cyclic stretching to 2200% strain without failure and recovered 61.0% of the imposed deformation, clearly demonstrating the robustness of its primary cross-linked elastic network. Even when stretched to near its breaking elongation, the rejoined fractured pieces showed only a small residual strain of approximately 20% (Supplementary Fig. 23). Beyond absolute strength, the elastomer exhibits comprehensive damage tolerance. It sustains a true rupture stress of 3.72 GPa (Fig. 5g) together with a puncture force of 63.862 N, and it also exhibits a tear energy as high as 790.85 ± 37.75 kJ m⁻² (Fig. 5h; Supplementary Fig. 24; Supplementary Table 7). Moreover, Lig-TPUE is markedly notch-insensitive; growing cracks are efficiently blunted by the dynamic network (Fig. 5i; Supplementary Figs. 25 and 26). Lig-TPUE also defies the conventional trade-off between performance and processability. The material remains fully thermoplastic: it can be solution-cast, melt-processed, and melt-spun into uniform fibers (Fig. 5j). These features facilitate versatile 3D printing of complex, load-bearing geometries (Fig. 5k), seamlessly linking laboratory synthesis with advanced manufacturing. To elucidate the excellent spinnability and thermal adaptability of Lig-TPUE during processing, temperature-dependent stress relaxation tests were conducted (Supplementary Fig. 27). At the melting zone temperature (120 °C), the rapid stress relaxation confirms the swift structural rearrangement of the dynamic network, ensuring excellent melt flowability. Notably, even at the cooler spinneret exit temperature (95 °C), Lig-TPUE can still dissipate ~80% of its internal stress within 1 minute and fully relax within 5 minutes. The thermalplastic processability of Lig-TPUE does not originate from the absence of covalent bonds, but from the fact that the material does not form a fully permanent covalent network under the present formulation. At elevated temperatures, reversible physical associations and hard domains can be disrupted and reorganized, allowing chain mobility and melt processing. This robust dynamic adaptability efficiently eliminates the residual shear stress accumulated during extrusion, effectively preventing melt fracture while providing adequate melt strength for continuous fiber drawing. To explicitly decouple the structural contribution of lignin’s three-dimensional aromatic framework from the concentration effect of its hydroxyl groups, control polyurethane elastomers crosslinked by purely aliphatic polyols (erythritol and sorbitol) were synthesized and evaluated (Supplementary Fig. 28). The theoretical hydroxyl densities of erythritol (32.80 mmol g⁻¹) and sorbitol (32.95 mmol g⁻¹) are ~three times higher than that of the lignin polyol (10.92 mmol g⁻¹). However, despite possessing a denser theoretical hydrogen-bonding network, these aliphatic-polyol-based elastomers exhibited significantly lower mechanical properties (e.g., maximum tensile strengths of 37.35 MPa and 74.48 MPa, respectively) compared to Lig-TPUE. This sharp contrast confirms that the remarkable mechanical enhancement observed in Lig-TPUE is not attributable solely to the increased number of hydroxyl groups. Rather, the marked improvement in performance arises intrinsically from the rigid three-dimensional aromatic structure of lignin. This architecture effectively disperses macroscopic stress while allowing the surrounding dynamic hydrogen bonds to undergo sacrificial rupture and reorganization, thereby demonstrating the indispensable synergistic effect of the aromatic framework and the polyhydroxy structure. Collectively, these results demonstrate that lignin-engineered TPUE redefines the property space of thermoplastic elastomers. It delivers record strength and toughness, excellent damage tolerance, and robust elasticity while retaining full thermoplastic processability, thus establishing a new paradigm for sustainable, high-performance polymer design. Integrated performance and closed-loop recyclability in carbon fiber composites The superior mechanical profile and intrinsic processability of Lig-TPUE render it a sustainable matrix for high-performance composites. To verify this capability, carbon-fiber (CF) fabric laminates were fabricated by solution casting. The resulting Lig-TPUE/CF composite attains a tensile strength of 913.7 MPa—nearly three times that of the pristine CF fabric and ~400 MPa higher than the Control-TPUE/CF analogue (Fig. 6a). Its Young’s modulus likewise rises to 3.68 GPa (Fig. 6b). This pronounced reinforcement stems from the lignin-engineered network, which supplies a stiff yet tough matrix that transfers load to the fibers, as confirmed by complete impregnation and smooth interfacial morphology observed by SEM (Fig. 6c). The composite amplifies the matrix’s damage tolerance. It exhibits a maximum tear load of 130.2 N and a puncture resistance of 147.3 N, both far exceeding those of the neat CF fabric and the control composite (Fig. 6d, e; Supplementary Figs. 29 and 30). Lignin-engineered TPUE enables high-performance, fully recyclable carbon fiber (CF) composites. a Tensile stress-strain curves of pristine CF fabric, Control-TPUE/CF, and Lig-TPUE/CF composites, showing nearly threefold reinforcement. b Young’s modulus comparison. c SEM images of Lig-TPUE/CF surface and cross-section, confirming excellent fiber impregnation. d, e Tear resistance (d) and puncture resistance (e) tests demonstrating superior damage tolerance. f Mechanical property retention over multiple recycling cycles (>90% recovery). g Schematic of the closed-loop recycling process enabling recovery of both CF fabric and Lig-TPUE matrix (DMF as solvent, 80 °C for 4h). Beyond its mechanical performance, the system was deliberately designed with circularity in mind. The supramolecular network enables solvent-assisted reversible crosslinking, allowing both the CF fabric and the Lig-TPUE matrix to be recovered through a closed-loop recycling process (Fig. 6g). Specifically, the elastomer was recycled via a solvent-assisted separation strategy using N,N-dimethylformamide (DMF) under stirring at 80 °C for 4 h. The resulting elastomer solution was then cast into a polytetrafluoroethylene (PTFE) mold and vacuum-dried at 75 °C for 12 h to completely remove the solvent, yielding the regenerated elastomer. Notably, the reclaimed elastomer retained more than 90% of its initial mechanical properties after multiple recycling cycles (Fig. 6f). Infrared spectroscopy revealed that the characteristic absorption peaks of the recovered elastomer were highly consistent with those of the original sample, confirming that its chemical structure remained essentially unchanged during the dissolution–regeneration process. Collectively, these results demonstrate that the recovery method does not compromise the polymer backbone and is therefore well suited for closed-loop recycling and repeated reprocessing, offering a practical strategy for the development of sustainable, high-performance composites that integrate durability with end-of-life recoverability. Lig-TPUE also exhibits strong versatility in reinforcing both inorganic rigid glass-fiber (GF) fabrics and organic rigid Kevlar-fiber (KF) fabrics. Specifically, the tensile strength of the Lig-TPUE-GF composite increases markedly from 108 MPa for neat GF to 604 MPa, while that of the Lig-TPUE-KF composite rises from 301 MPa for neat KF to 1249 MPa. These results further underscore its substantial potential as a matrix for rigid-fiber reinforcement (Supplementary Fig. 31). To assess the thermal stability of the elastomer for composite applications, thermogravimetric analysis (TGA) was performed (Supplementary Fig. 32). The DTG curve shows a typical multi-step degradation process: an initial weight loss at 200–250 °C, assigned to the dissociation of urethane linkages and weak dynamic non-covalent bonds; a broader degradation stage at 300–350 °C, mainly associated with decomposition of the polyol soft segments and aliphatic chains; and a final stage at 400–450 °C, corresponding to degradation of aromatic-rich structures and char formation. Notably, the degradation onset temperature (Tonset) of Lig-TPUE remains well above 200 °C, exceeding the 100–180 °C range typically used in the processing and service of structural carbon-fiber composites. These results indicate a sufficiently broad thermal processing window, enabling composite fabrication without premature degradation or loss of structural integrity. Discussion In this work, we demonstrate a lignin-enabled molecular design strategy for thermoplastic polyurethane elastomers that simultaneously enhances strength, toughness and elastic recovery while preserving thermal processability. The central feature of this design is the use of bio-aromatic lignin polyols as multifunctional chain extenders, which convert lignin from a conventional passive filler into an active structural component of the polyurethane network. Through covalent incorporation into the polyurethane backbone and the formation of reversible supramolecular associations, lignin contributes to a dynamic three-dimensional network that promotes efficient energy dissipation and stress transfer. The improvement in mechanical performance should not be attributed solely to the polyhydroxyl functionality of lignin. Rather, it arises from a balanced combination of covalent anchoring and reversible physical interactions. The hydroxyl groups of lignin react with isocyanate groups to form urethane linkages, thereby integrating lignin into the polyurethane matrix and improving interfacial compatibility. At the same time, the rigid aromatic structure of lignin restricts chain mobility, increases the effective hard-segment content and promotes microphase separation. In addition, hydrogen bonding between lignin hydroxyl groups and urethane groups, together with aromatic π–π interactions among lignin units, generates reversible physical associations that can dissipate energy during deformation. These interactions collectively reinforce the elastomer while allowing partial dissociation and reorganization of the physical network at elevated temperatures. Thus, the thermal processability of Lig-TPUE is not based on the absence of covalent urethane bonds, but on the absence of a densely crosslinked permanent covalent network and the presence of dynamic supramolecular interactions. The resulting Lig-TPUE exhibits a tensile strength of 127.7 ± 9.52 MPa and a toughness of 1.43 ± 0.11 GJ m − 3, together with excellent elastic recovery. This combination of properties places Lig-TPUE among the high-performance thermoplastic elastomers reported to date. Beyond its performance as a standalone elastomer, Lig-TPUE also serves as an effective matrix for carbon-fiber composites, providing mechanical reinforcement, enhanced tear and puncture resistance, and recyclability through thermally assisted reprocessing. These features suggest a promising route for integrating high mechanical performance with circular material use. More broadly, this study highlights the potential of rigid aromatic bio-derived units for constructing dynamic polymer networks in which covalent connectivity and reversible physical associations are deliberately balanced. Such a design principle provides a pathway for decoupling mechanical strength and toughness in sustainable polymer systems, while enabling the high-value utilization of lignin, an abundant and underused biomass-derived aromatic resource. Looking forward, this strategy could be extended to other polymer platforms, including polyurethanes, epoxies and polyesters, through appropriate control of functional group reactivity, network density and supramolecular interactions. These advances may contribute to the development of next-generation materials that combine performance, processability and sustainability. Methods Materials Choline chloride (ChCl), oxalic acid dihydrate, and glycerol were purchased from Tianjin Fuyu Fine Chemical Co., Ltd. (China). Poly(tetramethylene ether) glycol (PTMEG, Mn = 2000 g mol⁻¹), dicyclohexylmethane-4,4’-diisocyanate (HMDI, 99%), and N,N-dimethylformamide (DMF, 99.8%) were obtained from Shanghai Rhawn Reagent Co., Ltd. (China). Dibutyltin dilaurate (DBTDL, 95%) was supplied by Shanghai Macklin Biochemical Co., Ltd. (China). Dimethyl 3,3’-dithiodipropionate (DESA) was procured from Aladdin Reagent Co., Ltd. (China). Hydrazine monohydrate (80%) was sourced from Fuchen (Tianjin) Chemical Reagent Co., Ltd. (China). All chemicals were used as received without further purification. Milled poplar wood powder with a nominal particle size of 100 mesh (aperture size ≈ 150 μm) was purchased from a commercial supplier via an online platform in China and used as received without further purification. Isolation of lignin polyols A deep eutectic solvent was prepared by combining ChCl (0.10 mol), oxalic acid dihydrate (0.10 mol), and glycerol (0.10 mol) in a 250 mL round-bottom flask. The mixture was heated to 120 °C and stirred at 500 rpm until a transparent liquid formed. Milled poplar wood powder (3 g) was then added, and the suspension was kept at 120 °C for 4 h to achieve biomass fractionation. After cooling to room temperature, the reaction mixture was poured into a 500 mL beaker containing a ten-fold excess of counter-solvent (ethanol/water, 7:3 v/v; ~400 mL) and stirred for an additional 2 h. The resulting slurry was vacuum-filtered, and the filtrate was washed three times with fresh counter-solvent. Removal of solvent by rotary evaporation afforded an aqueous slurry that was precipitated in distilled water. The precipitate was collected by filtration and dried at 60 °C overnight to yield lignin polyols. Synthesis of 3,3’-Dithiobis(propionohydrazide) (DESH) DESA (0.02 mol) and hydrazine monohydrate (0.10 mol) were dissolved in methanol (100 mL) and stirred at 45 °C for 24 h. The resulting suspension was filtered, the solid was washed three times with methanol, and the product was dried under vacuum at 100 °C to give DESH as a white powder (yield: 87 %). Synthesis of control-TPUE and Lig-TPUE PTMEG (5.0 g) was dried at 120 °C under vacuum for 2 h and equilibrated at 80 °C. For Control-TPUE, HMDI (1.31 g) and DBTDL (0.02 g) in DMF (10 mL) were added to PTMEG and reacted at 80 °C for 4 h to obtain an NCO-terminated prepolymer. A solution of DESH (0.596 g) in DMF (50 mL) was then introduced, and the mixture was stirred under N2 at 80 °C for 15 h, affording a viscous solution. For Lig-TPUE, HMDI (1.5, 1.7, or 1.9 g) and DBTDL (0.02 g) in DMF (10 mL) were reacted with PTMEG (5.0 g) at 80 °C for 4 h. Afterward, DESH (0.596 g in 50 mL DMF) was added and stirred for 2 h, followed by lignin polyols (0.121, 0.243, or 0.364 g in 30 mL DMF). The reaction proceeded at 80 °C under N2 for 15 h to yield a homogeneous brown solution. The product was cast into a PTFE mold, cured at 85 °C for 24 h, and dried under vacuum at 80 °C for another 24 h to obtain the final elastomeric films. Characterizations 1H NMR spectra were recorded on a Bruker Avance III 500 MHz spectrometer. Gel permeation chromatography (Waters 1515, THF eluent, 1.0 mL min⁻¹) determined the molecular weights, using 2–3 mg mL⁻¹ solutions and 60 µL injection volumes. FTIR spectra were collected on a Bruker VERTEX 80 V (400–4000 cm⁻¹, 32 scans), and XRD patterns on a Rigaku Ultima IV diffractometer (Cu Kα, λ = 0.15418 nm, 40 kV, 30 mA). Dynamic mechanical analysis (DMA) was performed on a TA Instruments Q800 in tension mode (−100 to 150 °C, 5 °C min⁻¹, 1 Hz). Small-angle X-ray scattering (SAXS) was conducted on a Xeuss 3.0 system (GeniX3D Cu ULD source, λ = 0.154 nm) with a 1035 mm sample–detector distance and a Pilatus 100 K detector. Two-dimensional patterns collected over 900 s were azimuthally integrated (Fit2D), and the domain spacing (L) was calculated from Bragg’s law, L = 2π/q, where q is the scattering vector at the primary peak maximum. Calculation of apparent crosslink density The apparent crosslink density (ve) of all samples was calculated from the storage modulus (Er) of the polyurethane film in the rubbery plateau region (Tg + 40 °C), by using the following equation: where R refers to the universal constant and T is the absolute temperature. Mechanical performance determination The tensile tests were implemented at room temperature using a universal testing machine (Gotech AI-7000S, 5 KN). Rectangle-shaped samples (15.0 × 2.0 × 0.2 mm) were prepared, and each reported result was the average of measurements from at least three samples at a stretching speed of 50 mm min⁻¹. The toughness (T) was defined as the area under the engineering stress (σ)–strain (ε) curve, and calculated by the following equation: True stress (σt) and true strain (εt) were calculated based on the engineering stress-strain curves, according to the following equations: The fracture energy test was measured during the tensile test using a sample notched on one edge at the stretching speed of 50 mm min−1. Both the notched and unnotched specimens (gauge length of 10.0 mm, width of 5.0 mm, and thickness of 0.2 mm) were tested. The fracture energy (Gc) was given by the following equation: where c is the length of the notch (1.0 mm), λc is the value of (elongation at break + 1) of the notched sample, and W is the strain energy calculated by integrating the stress-strain curve of the unnotched specimen until elongation at break. Supplementary information Supplementary Information Transparent Peer Review file Acknowledgements The authors acknowledge the support by the National Natural Science Foundation of China (No. 32330072; No. 32571981) and the Fundamental Research Funds for the Central Universities (No. 2572024AW40). Author contributions H.Y. and Y.L. supervised this project. P.Z. performed most of the experiments and analyzed the data. T.W., J.S. and M.L. performed the experiments of lignin polyols extraction and characterization. Q.T. participated in the characterization of the polyurethane elastomers. Z.R. and P.Z. performed the 3D printing experiments. P.Z., Y.L. and H.Y. analyzed the results and co-wrote the paper. All authors commented on the manuscript. Peer review Peer review information Nature Communications thanks Zhanhui Yuan and the other anonymous reviewers for their contribution to the peer review of this work. [A peer review file is available.] Data availability The data that support the findings of this study are available within the article and the Supplementary Information. Detailed information, including the structural characterization of lignin polyols (HSQC and ³¹P NMR spectra, and GPC molecular weight data), control experiments with varied chain extender dosages and aliphatic polyol crosslinkers, cyclic tensile and stress‑relaxation tests under extreme strains, thermal stability and composite reinforcement with glass fibers and Kevlar fibers, as well as the molecular dynamics simulation protocols and binding energy calculations, is available from the corresponding authors upon request. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Yongzhuang Liu, Email: lyz@nefu.edu.cn. Haipeng Yu, Email: yuhaipeng20000@nefu.edu.cn. Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41467-026-77396-1. References Associated Data Supplementary Materials Supplementary Information Transparent Peer Review file Data Availability Statement The data that support the findings of this study are available within the article and the Supplementary Information. Detailed information, including the structural characterization of lignin polyols (HSQC and ³¹P NMR spectra, and GPC molecular weight data), control experiments with varied chain extender dosages and aliphatic polyol crosslinkers, cyclic tensile and stress‑relaxation tests under extreme strains, thermal stability and composite reinforcement with glass fibers and Kevlar fibers, as well as the molecular dynamics simulation protocols and binding energy calculations, is available from the corresponding authors upon request.