Spider-silk-inspired self-healing conductive elastomer for joint rehabilitation detection and tactile temperature warning

Kai Yan*^a, Xiaolong Cao^a, Yuhao Tang^a, Qunna Xu^a, Yan Zong^a, Zejian Yang^b and Can Zhou^b
aCollege of Bioresources Chemical and Materials Engineering, Shaanxi University of Science & Technology, Xi’an 710021, PR China. E-mail: yankai@sust.edu.cn
^bBreast Surgery Department, First Affiliated Hospital, Xi’an Jiaotong University, Xi’an 710061, PR China

First published on 11th August 2025

---

---

New concepts In this study, inspired by the structure of spiderweb protein, a health monitoring platform based on a self-healing conductive polyurethane elastomer was constructed. Compared with previous studies, the polyurethane elastomer exhibits excellent mechanical strength (34 MPa) and room-temperature self-healing performance (self-healing efficiency of 91.4% at room temperature) through the construction of multi-hydrogen-bond nanodomains. Additionally, the well-designed ionically/electronically conductive elastomer exhibits bidirectional mechanical detection and dual temperature detection capabilities, allowing it to simultaneously decode biomechanical and thermal characteristics. It can be applied to wearable electronic devices to monitor the rehabilitation status of human joints and external environment for real-time rehabilitation assessment and preemptive injury warning. Therefore, the ion/electron-conductive self-healing polyurethane expands ‘sensory fusion’ in wearable electronics, and transforms multiple physical data into operable health insights through a single material.

1. Introduction

The convergence of artificial intelligence (AI) and internet of things (IoT) technologies has garnered widespread attention for smart wearable devices in the field of healthcare. Self-healing ion-conductive elastomers offer significant advantages in health monitoring (e.g., body temperature, human joint movement),^1–4 electronic skin,^5–7 and human–computer interaction^8–10 due to their excellent mechanical and self-healing properties. Although soft electronic materials have made great strides in health detection, most flexible sensors with non-integrated functions can only detect mechanical deformation or environmental stimuli (such as temperature,¹¹ humidity and photonic signals) due to incompatible sensing mechanisms.^12,13 Therefore, the development of stretchable and self-healing electronics that combine sensing of multiple mechanical and environmental stimuli is essential, but often challenging.

Profiting from the rapid development of supermolecular chemistry, ion-conductive elastomers with specific mechanical properties and self-healing ability can be prepared by introducing synergy among multiple dynamic bonds (e.g., hydrogen bonds,^14,15 ion interactions,^16,17 host–guest interactions^18–20) or regulation of microphase separation structures (such as polyurethane–ionic liquid systems)^21,22 in polymers. However, due to the contradictory nature of mechanical strength and self-healing ability, self-healing ion-conductive elastomers usually exhibit a combination of high mechanical strength and high healing temperature.^23–25 For example, Ye et al.²⁶ prepared a poly(oxime-urethane) elastomer with high tensile strength (22 MPa) and good self-healing ability (self-healing efficiency ∼91.7%) through hydrogen bonding and coordination interactions. The recovery toughness was 75.4 MJ m⁻³ at 80 °C. These requirements significantly increase energy consumption, which limits practical application in complex scenarios. In addition, in order to achieve the synergy of self-healing with conductivity, continuous ion channels (such as ionic liquid enrichment zones) can be formed through microphase separation, thus ensuring that the conductivity (>10⁻³ S cm⁻¹) is not disturbed by the self-healing process.^27,28 For example, Xu et al.²⁹ prepared a self-healing ionic gel with high conductivity (5.28 mS cm⁻¹) and excellent self-healing ability (self-healing efficiency ∼98.8%) by constructing a phase separation structure. The hard phase region undergoes self-healing via reversible fracture and formation, while the hydrogen bonds between the polymer and IL in the soft phase region facilitate ion migration, leading to high conductivity. Therefore, in order to meet the needs of practical applications, the preparation of ion-conductive elastomers with high mechanical strength and excellent room temperature self-healing ability is an urgent problem to be solved.

The emergence of the internet of things (IoT) has stimulated demand for smart sensors with health monitoring and high-temperature monitoring/early warning ability to pre-empt potential health and safety risks. However, traditional health monitoring (which mainly relies on medical methods such as magnetic resonance imaging and X-ray imaging)^30–32 and high-temperature-prevention systems have high monitoring costs and low convenience. Although emerging technologies such as gait analysis methods and wearable physiological monitoring devices show promise, their application is currently constrained by environmental and equipment limitations.^33–36 Furthermore, most of the current thermal monitoring platforms remain predominantly confined to biometric thermometry applications (e.g., wearable fever detection) and exhibiting various limitations for simultaneously sensing elevated ambient thermal regimes and preventing injuries. Lin et al.³⁷ prepared a multifunctional sensor by introducing non-covalently modified carbon nanotubes into carboxylated styrene-butadiene rubber. The prepared sensor has an impressive thermal response (0.01636 °C⁻¹), which enables temperature detection in the human body. Lu et al.³⁸ prepared a NiO/CNTF flexible temperature sensor to prevent fires caused by out-of-control electronic equipment. The sensor has a wide operating range (−15 to 60 °C) and high sensitivity (maximum TCR is 20.2% °C⁻¹). Most current temperature sensors for health monitoring can monitor only the object temperature or ambient temperature, but cannot monitor both to achieve better protection for the human body. Therefore, the realization of integrated flexible wearable systems with concurrent environmental sensing and physiological monitoring capabilities is of considerable significance for enabling continuous health surveillance and proactive risk mitigation in daily living scenarios.

Here, a series of room-temperature self-healing ion-conductive polyurethane elastomers with excellent mechanical properties, high ionic conductivity, bidirectional sensing and thermoelectric properties were prepared via the synergistic use of LiTFSI/COOH-CNTs and multiply hydrogen-bonded polyurethane. The existence of multiply hydrogen-bonded hard phase regions and soft phase regions containing a large number of ether bonds provides the basis for high mechanical strength and self-healing. The hard phase region dissipates energy through the fracture and formation of multiple hydrogen bonds, endowing the material with high mechanical strength. At the same time, multiple hydrogen bonds are broken and recombined to achieve self-healing at room temperature. In addition, Li–O bonds are formed between the polyurethane in the soft phase region and the soft segment of LiTFSI, which promotes ion migration through molecular chain movement. Furthermore, the polyurethane molecular chain can also form hydrogen bonds with the COOH-CNTs, which ensures the high conductivity of the ion-conductive elastomer. Interestingly, based on its excellent synergistic ion/electron-conductive components, synergistic monitoring of strain, temperature, and thermoelectricity can be achieved. Thus, the elastomer shows advantages as a wearable electronic device in joint rehabilitation monitoring systems and prevention/monitoring systems.

2. Results and discussion

2.1. Structure and mechanical performance of self-healing polyurethane elastomers

Most organisms have complex repair mechanisms to reduce damage. For example, the structural proteins of spider silk are characterized by repetitive amino acid sequences that organize into rigid β-sheet crystalline domains (imparting mechanical strength) and flexible amorphous regions (providing elasticity). Under external stress, the β-sheet crystalline domains exhibit dynamic hydrogen bond dissociation and reconfiguration, facilitating energy dissipation while maintaining structural integrity (Fig. 1a).³⁹ Inspired by the multi-level hydrogen bond structure of spider silk, a linear polymer was synthesized via two-step polymerization using polytetrahydrofuran ether glycol (PTMG), isophorone diisocyanate (IPDI) and adipic acid dihydrazide (ADH), and then a polyurethane incorporating numerous hydrogen bond donors and acceptors was prepared through crosslinking with 2-amino-2-methyl-1,3-propanediol (AMPD); the final polyurethane was designated as SPU (Fig. 2a and Fig. S1). ADH was chosen as a chain extender because of its hydrazide groups, which react with the isocyanate groups of IPDI to generate acylsemicarbazide (ASCZ) bonds. These ASCZ bonds feature plentiful hydrogen bond donors (–NH–) and acceptors (–CO, –NH–), facilitating the development of supramolecular networks. Additionally, the successful preparation of SPU was confirmed via Fourier transform infrared (FT-IR) spectroscopy (Fig. S2). The stretching vibration of hydrogen-bonded CO was significantly enhanced at 1660 cm⁻¹ (Fig. 2b), indicating that the incorporation of ADH produced an ASCZ-rich part (–NH and –CO groups), which effectively increased the number and strength of hydrogen bond interactions. Furthermore, the characteristic amide I (CO stretching at 1680 cm⁻¹) and amide II (N–H bending/C–N stretching at 1540 cm⁻¹) bands showed a blue shift and peak broadening, demonstrating that the intermolecular hydrogen bond enhancement (N–H⋯OC) was generated by the ASCZ structure formed during chain extension.⁴⁰ Deconvolution analysis of the FT-IR stretching region at 1750–1600 cm⁻¹ resolved the CO vibration band of SPU₄ into seven Gaussian components (Fig. 2c), which were systematically assigned to distinct carbonyl species: free (1728 cm⁻¹) and hydrogen-bonded (1703 cm⁻¹) carbamate groups, non-associated (1687 cm⁻¹) and H-bonded (1665 cm⁻¹) urea linkages, and three amide-related configurations (1648, 1632, 1615 cm⁻¹), confirming the presence of multiple modes of hydrogen-bonding interactions (N–H⋯OC) within the ASCZ segments that collectively contribute to supramolecular network formation.⁴¹ The hierarchical hydrogen-bonding networks formed between ASCZ segments (N–H⋯OC interactions) fundamentally underpin the enhanced mechanical performance of SPU. The mechanical strength of SPU exhibited progressive enhancement with increasing ADH content. Specifically, when the molar ratio of PTMG to ADH reached 1:1, the resulting SPU exhibited a maximum tensile strength of 34 MPa, which represents an obvious advantage compared with previous studies (Fig. 2d and Fig. S11a). This enhancement mechanism is mainly due to the hydrogen bond interactions between the amino group of ADH and the hard segment of SPU, which effectively limit the mobility of the molecular chain under stress conditions. Additionally, the incorporation of the cross-linking agent AMPD significantly improved the mechanical strength of SPU, achieving a tensile strength of 36 MPa. This enhancement was accompanied by a reduction in tensile ductility, suggesting a trade-off between cross-linking density and the flexibility of the material (Fig. 2e, Fig. S3, S5 and S6). This behavior likely stems from AMPD-induced covalent network formation, which restricts polymer chain mobility while increasing structural rigidity. The synergistic interplay between the hierarchical hydrogen-bonding architecture of SPU (comprising 66% reversible N–H⋯OC interactions) and chemically cross-linked networks endows the material with exceptional damage tolerance. As quantified by puncture resistance tests and trouser tear measurements, the fracture energy of the elastomer was 87.23 kJ m⁻² (Fig. 2f and g), which may be attributed to the synergistic energy dissipation mechanism formed by the reversible fracture/recombination of ASCZ hydrogen bonds and the tortuous crack path caused by the chemical crosslinking structure.⁴²

2.2. Self-healing performances of the self-healing polyurethane elastomers

Through the synergistic effect of multi-scale hydrogen bond networks and dynamic covalent bonds, a balance between mechanical strength and self-healing ability is achieved. This hierarchical hydrogen bond network is mainly composed of strong hydrogen bonds (N–H⋯OC) from the acylsemicarbazide (ASCZ) fragments, moderate interactions from the urethane (–O–CO–NH–) and amide (–CONH–) motifs, and weak H-bonds from the IPDI-derived NH-group. The existence of this synergy allows the hydrogen bonds to break and dissipate energy during the stretching process of SPU. After recycling, the hydrogen bonds are recombined to form a new network structure that guarantees its mechanical properties (Fig. 3a). As a result of its doubly cross-linked structure, SPU shows exceptional load-bearing capacity, being capable of supporting weights up to 18750 times its own mass while maintaining dimensional stability with minimal permanent deformation post-recovery (Fig. 3b). The self-healing ability of SPU was quantitatively studied via tensile testing. As illustrated in Fig. 3c, both the tensile strength and strain of the self-healed SPU demonstrate progressive enhancement with extended healing duration. Following a 48-hour self-healing period, the tensile strength and strain of the material were effectively restored to 32.62 MPa and 512%, respectively, approaching the performance levels of the pristine SPU specimens. This recovery corresponds to a remarkable healing efficiency of 91.4%, calculated as the ratio of healed-to-original tensile strength. To investigate the self-healing mechanism, a high-resolution microscope was utilized for three-dimensional morphological assessments of the SPU surfaces at various recovery stages. Prior to damage, the SPU exhibited distinctive surface roughness, whereas the sheared sample displayed distinct notch marks at the fracture interface. Notably, after 24 hours of recovery at ambient temperature, these gaps show significant attenuation, and the scratches become significantly shallower and eventually disappear. At the same time, it can be observed in the reflection infrared spectrum that the hydrogen bond strength is almost the same as that in the initial state, indicating that the elastomer is almost completely repaired (Fig. 3d and Fig S4). These results indicate that SPU has excellent room-temperature self-healing ability.

2.3. Synthesis and performance of the ionic–electronic conductive elastomer

An ionically conductive self-healing elastomer (SPU₄–LiTFSI_m) was engineered by incorporating lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) into supramolecular polyurethane (SPU). The successful introduction of LiTFSI was studied using FT-IR spectroscopy. The characteristic peaks of N–H (3320–3285 cm⁻¹) and the CO (1730–1715 cm⁻¹) stretching vibration shift were observed (Fig. 4a and b), confirming the coordination of Li⁺ with the polymer soft segments while preserving the hydrogen-bonding crosslinks essential for mechanical resilience. Additionally, the TFSI⁻ ions of the small molecular structure are embedded in the polyurethane, resulting in the plasticization of the polyurethane and a linear decrease in its mechanical properties. When the content reaches 50%, the mechanical strength decreases by about 80 times (Fig. S7). However, as the ion content increases, the conductivity is significantly enhanced (Fig. S14). Based on the balance between conductivity and mechanical properties, SPU–LiTFSI_0.3 was finally selected for further optimization. To solve the inherent problem of the trade-off between conductivity and mechanical properties in the SPU₄–LiTFSI_0.3 system, COOH-CNTs were introduced into SPU. The mechanical properties of the SPU–LiTFSI–CNT elastomer were gradually enhanced with increasing COOH-CNTs content due to the hydrogen bonding between SPU–LiTFSI–CNT and the amino acid groups and the π–π stacking of TFSI⁻. The tensile strength was increased to 14 MPa, and the material exhibited ultra-high ductility (ε ≈ 900%) with an ionic conductivity of 4.2 mS cm⁻¹ (Fig. 4c, Fig. S14 and S15). At the same time, it exhibited considerable fatigue resistance, maintaining 92% initial elastic recovery after 5 deformation cycles (0–300% strain) (Fig. 4d and Fig. S8). Subsequently, the surface structure of the SPU₄–LiTFSI_0.3–CNT_0.3 elastomer was characterized using scanning electron microscopy (SEM). It can be observed that the elements C, O, N, F and S were evenly distributed in the polymer, indicating that LiTFSI and COOH-CNTs were evenly distributed in SPU₄ (Fig. 4e, Fig. S8 and S9). However, it can also be observed that the COOH-CNTs are deposited at the bottom layer of polyurethane under the action of gravity in the cross-sectional SEM image of the SPU–LiTFSI_0.3–CNT_0.3 elastomer, while the LiTFSI is uniformly distributed in the SPU elastomer. This uneven distribution provides a basis for the realization of bidirectional sensing (Fig. S10). Furthermore, due to the presence of mobile ions and hydrogen bonds, SPU₄–LiTFSI_0.3–CNT_0.3 can achieve self-healing while maintaining enhanced ionic conductivity through the permeable ion transport pathway. In order to quantitatively evaluate the self-healing ability of the SPU₄–LiTFSI_0.3–CNT_0.3 ionic elastomer, the elastomer strip was integrated into a series circuit. The SPU₄–LiTFSI_0.3–CNT_0.3 ionic elastomer exhibited ultrafast electrical recovery speed (25 °C, 38% RH), and the volume resistivity returned to the baseline level within 500 ms after contact (Fig. 4f). Concurrently, instantaneous LED reactivation was observed (Fig. 4g), confirming the re-establishment of continuous ion-conductive pathways across the healed interface through the reconstruction of Li⁺ coordination networks and reorientation of COOH-CNTs-assisted electron tunneling junctions (Fig. S17). This rapid dual-channel restoration mechanism to achieve both ionic and electronic continuity, coupled with the tripartite performance characteristics, positions SPU₄–LiTFSI_0.3–CNT_0.3 as a prime candidate for durable flexible electronics.

2.4. Sensing performance of the ionically and electronically conductive elastomers

The electrical properties of SPU₄–LiTFSI_0.3–CNT_0.3 were further characterized to evaluate its potential use as a flexible sensor. It is worth noting that current common flexible conductive materials often face a trade-off between ionic conductivity and mechanical strength.⁴³ In our study, the addition of solid conductive additives (LiTFSI) and COOH-CNTs was found to improve the Young's modulus and toughness, which may overcome the undesirable contradictions in traditional soft conductors. Therefore, SPU₄–LiTFSI_0.3–CNT_0.3 was fabricated as a strain sensor, and its sensing performance was evaluated. The sensitivity of the SPU₄–LiTFSI_0.3–CNT_0.3 sensor was quantitatively assessed through cyclic tensile testing (0–300% strain). The relative resistance change (ΔR/R₀) exhibits a monotonic increase with the applied strain, demonstrating a nonlinear piezoresistive response characteristic of hierarchical conductive networks. This strain-dependent behavior arises from the dynamic interplay between the disruption of the COOH-CNTs-based electron tunneling pathways and the alignment-enhanced ionic conduction through the oriented Li⁺ transport channels. In the low-strain regime (ε < 50%) governed by reversible microcrack formation in the COOH-CNTs percolation network, the GF is ∼0.67. In the intermediate regime (50–200%), the GF increases to 3.64–6.84, reflecting the synergistic effects of COOH-CNTs reorientation and ion-dipole network distortion. In the high-strain regime (200–300%), the GF stabilizes at 5.23 due to the strain-induced crystallization of the PU soft segments (Fig. 5a and Fig. S17). A higher GF indicates a more sensitive resistance response, which is more conducive to detecting small movements. In light of its excellent tensile properties, the relative real-time resistance (ΔR/R₀) of the SPU₄–LiTFSI_0.3–CNT_0.3 sensor was tested under 5–50% strain and bending at different angles (Fig. 5b, c and Fig. S19). It can be seen that the SPU₄–LiTFSI_0.3–CNT_0.3 sensor maintains a regular and stable response under different strains and bending angles, showing excellent stress sensitivity. Subsequently, when loaded onto human joints (fingers, elbows, wrists), the sensor can capture the motions of these joints. Different joint movements give rise to different signals, which can be used to clearly distinguish the movements of the different joints (Fig. 5d–f). Remarkably, the sensor achieves millisecond-scale response time (τ = 163.7 ms) with exceptional cyclic stability (over 10³ cycles) (Fig. 5g and h). In addition, although the sensor exhibits some signal variation at different temperatures and humidity values, the strain-sensing signals showed negligible difference when tested at the same strain and frequency under different temperatures and humidity values, demonstrating that the SPU–LiTFSI–CNT sensor can perform accurate and reliable strain sensing at various temperatures (Fig. S18). In addition, the sensor exhibits good biocompatibility (Fig. S23). This tunable strain-response profile, coupled with a broad sensing range (>300% strain), positions the material as a versatile platform for multiscale biomechanical monitoring ranging from subtle pulse waveforms (ε ∼ 1%) to large joint movements (ε > 200%).

2.5. Joint rehabilitation detection performance of ionically and electronically conductive elastomers

It is well known that the wrist connects the hand to the forearm. If the wrist has inflammation or injury symptoms, the movement of the hand will be limited, the grip strength of the hand will be reduced, and the torque of the wrist will be poor.⁴⁴ Therefore, the recovery of the wrist can be predicted according to the wrist bending torque during rehabilitation. Due to the synergistic effect of the ion/electron-conductive materials, the SPU₄–LiTFSI_0.3–CNT_0.3 sensor exhibits bidirectional responses to bending. When bent upwards, the relative resistance change shows a positive response, while when bent downwards, it shows a negative response (Fig. 6a, Fig. S19 and S20). Based on the bidirectional sensing characteristics of the sensor, a joint detection system was designed based on the four common motion states of the wrist joint (flexion, extension, adduction, and extension; Fig. 6b). According to the relative resistance change, it corresponds to the four motion states of the wrist joint. When the wrist joint is in a healthy state, it shows a standard relative resistance change (Fig. 6c–f). When the wrist joint is under rehabilitation, the rehabilitation state can be detected in real time based on comparison of the observed real-time resistance change to the standard state. On the other hand, in the course of daily exercise, incorrect running may cause knee joint injury, which is likely to have an impact on the daily life of patients. Therefore, early detection of knee joint motion is of great significance for preventing/reducing knee joint injury. Therefore, based on the excellent bidirectional sensing characteristics of SPU₄–LiTFSI_0.3–CNT_0.3, a knee joint detection sensor was prepared. Different ΔR/R₀ states are observed under different motion states of the knee joint. For example, in the bending state, the relative resistance change is 50%, while the normal micro-bending state is 30% (Fig. 5g and h). Therefore, it is possible to detect whether the knee joint can be bent normally in order to detect its health status. In view of the joint detection ability shown by the sensor, it may be used in a human joint detection system in the future for joint rehabilitation status detection and joint injury prevention.

2.6. Temperature-detection performance of polyurethane ion-conductive elastomers

2.7. Thermoelectric-based object recognition and high-temperature warning

Due to the excellent sensing properties, temperature sensing and thermoelectric properties of the SPU₄–LiTFSI_0.3–CNT_0.3 ion-conductive elastomer, it can be integrated onto the hand for high temperature monitoring and human–computer interaction. In order to ensure low cost and wide applicability, the prepared SPU₄–LiTFSI_0.3–CNT_0.3 ion-conductive elastomer was curled into a ring, placed onto the finger joint, and connected using copper wire. Copper foil was used as the electrode to obtain a universal touch-monitoring system (Fig. 10a). The temperature-monitoring system identifies the temperature by monitoring the thermal voltage signal generated by the Seebeck effect. In particular, it is worth noting that the sensor produced different voltage signals at different joints due to changes in distance and movement (e.g., “grip”, stretch, proximity). As shown in Fig. 10b–d, when holding a beaker filled with iced, lukewarm, or hot water, different voltage signals are generated at different joints, allowing for accurate sensing of temperature. In addition, due to the excellent photothermal conversion capability of the sensor, the integrated temperature monitoring system can detect a voltage signal before contact to sense the temperature and prevent burns due to high temperatures (Fig. 10e). The sensor demonstrates significant potential for assisting visually impaired individuals in object recognition and manipulation during daily living activities, as well as in high-temperature sensing applications for robotic–computer interaction systems (Fig. 10f).

3. Conclusion

In this work, a room-temperature self-healing ion-conductive polyurethane elastomer was synthesized by incorporating multiple hydrogen-bond donors and acceptors. This elastomer exhibited superior mechanical properties, high conductivity, bidirectional sensing capabilities, and thermoelectric properties through the synergistic combination of carbon nanotubes (COOH-CNTs) and lithium bis(trifluoromethanesulfonyl)imide (LiTFSI). The presence of multiple hydrogen bonds facilitated energy dissipation through reversible destruction and recombination, leading to high mechanical strength (34 MPa) and exceptional self-healing efficiency at room temperature (91.4%). Moreover, the incorporation of synergistic ion/electron-conductive components in the elastomer enabled the simultaneous monitoring of strain, temperature, and thermoelectricity in a synergistic manner. The strain module exhibited bidirectional monitoring with high sensitivity (GF ∼ 5.23), while the temperature module displayed a resistance temperature coefficient of 2.42. The thermoelectric module demonstrated a power density of 1.59 mW m⁻¹ K⁻². Furthermore, based on the excellent photothermal conversion ability of the COOH-CNTs, non-contact sensing could be realized. This sensor technology presents novel opportunities for applications in human health monitoring, human–computer interaction, and high-temperature warning systems.

4. Experimental section

4.1. Materials

Poly(tetrahydrofuran) (PTMG, M_n ∼ 1000), adipic dihydrazide (ADH, ≥98%), N,N-dimethylacetamide (DMAc, ≥99.8%), February dibutyltin silicate (DBTDL, ≥98%), and lithium bis(trifluoromethanesulphonyl)imide (LiTFSI, ≥98%) were supplied by Anhui Sunrise Technology Co., Ltd. Isophorone diisocyanate (IPDI, ≥99%) was bought from Shanghai Macklin Biochemical Co. 2-Amino-2-methyl-1,3-propanediol (AMPD, ≥99%) was purchased from Shanghai Yi'en Chemical Technology Co., Ltd. Carboxylic carbon nanotubes (COOH-CNTs) were purchased from Jiangsu Xianfeng Nano Co., Ltd.

4.2. Preparation of self-healing polyurethane (SPU)

Using SPU₄ as a representative example, PTMG (5 g), IPDI (2.53 g) and DBTDL (0.08 g) were added to three flasks in turn, and DMAc (10 mL) was used as a solvent. The reaction was carried out at 80 °C for 2 hours. Then, DMAc (20 mL) was added and stirred for 10 min, followed by ADH (0.71 g) at 80 °C for 2 h. Finally, AMPD (0.088 g) was dissolved in DMAc (10 ml) and injected into the reaction system and mechanically stirred for 2 h, after which an appropriate amount of DMAc was added by observing the viscosity change of the system to obtain the final polyurethane elastomer. The samples were named SPU–ADH_x–AMPD_y, abbreviated as SPU, where x is the molar ratio of ADH to PTMG, and y is the mass ratio of AMPD in the system.

4.3. Preparation of SPU–LiTFSI_m–CNT_n

The self-healing polyurethane APU prepared in the previous step was dissolved in THF. After the SPU was completely dissolved, a certain amount of LiTFSI (10%, 20%, 30%) was added. After stirring for 30 min, COOH-CNTs (10%, 20%, 30%) with different polyurethane mass fractions were added, respectively. After stirring for 2 h, it was completely dispersed and introduced into a polytetrafluoroethylene plate to form a film at room temperature.

4.4. Characterization

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the SI. See DOI: https://doi.org/10.1039/d5mh01126d

The data supporting this article have been included as part of the SI.

Acknowledgements

This work was financially supported by the Natural Science Foundation of Shaanxi Province (2024JC-YBMS-412), Scientific Research Program Funded by Shaanxi Provincial Education Department (Program No. 24JR036).

References

1. A. Roy, S. Zenker, S. Jain, R. Afshari, Y. Oz, Y. Zheng and N. Annabi, Adv. Mater., 2024, 36, 2404225 CrossRef CAS .
2. X. Chen, Z. Wan, R. Zhang, L. Ma, Z. Yang and X. Xiao, Chem. Eng. J., 2024, 499, 156431 CrossRef CAS .
3. J. Li, Z. Fang, D. Wei and Y. Liu, Adv. Healthcare Mater., 2024, 13, 2401532 CrossRef .
4. M. Zhong, L. Zhang, X. Liu, Y. Zhou, M. Zhang, Y. Wang, L. Yang and D. Wei, Chem. Eng. J., 2021, 412, 128649 CrossRef .
5. J. C. Yang, J. Mun, S. Y. Kwon, S. Park, Z. Bao and S. Park, Adv. Mater., 2019, 31, 1904765 CrossRef PubMed .
6. S. Xiang, D. Liu, C. Jiang, W. Zhou, D. Ling, W. Zheng, X. Sun, X. Li, Y. Mao and C. Shan, Adv. Funct. Mater., 2021, 31, 2100940 CrossRef .
7. A. Chen, J. Zhang, J. Zhu, Z. Yan, Q. Wu, S. Han, J. Huang and L. Guan, J. Mater. Chem. A, 2023, 11, 4977–4986 RSC .
8. X. Li, M. Jiang, Y. Du, X. Ding, C. Xiao, Y. Wang, Y. Yang, Y. Zhuo, K. Zheng, X. Liu, L. Chen, Y. Gong, X. Tian and X. Zhang, Mater. Horiz., 2023, 10, 2945–2957 RSC .
9. K. Xue, C. Shao, J. Yu, H. Zhang, B. Wang, W. Ren, Y. Cheng, Z. Jin, F. Zhang, Z. Wang and R. Sun, Adv. Funct. Mater., 2023, 33, 2305879 CrossRef .
10. Q. Cao and W. Yuan, Chem. Eng. J., 2024, 494, 153254 CrossRef .
11. Z. Ma, Y. Zhang, R. Jiang, L. Shao, J. Cao, H. Guo and G. Zhang, Compos. Sci. Technol., 2024, 248, 110460 CrossRef .
12. L. Chen, X. Chang, H. Wang, J. Chen and Y. Zhu, Nano Energy, 2022, 96, 107077 CrossRef CAS .
13. T. Chen, Q. Wei, Y. Ma, Y. Tang, L. Ma, S. Deng and B. Xu, Nano Energy, 2024, 127, 109752 CrossRef CAS .
14. J. Cao, C. Lu, J. Zhuang, M. Liu, X. Zhang, Y. Yu and Q. Tao, Angew. Chem., Int. Ed., 2017, 56, 8795–8800 CrossRef CAS .
15. J. Wu, F. Zeng, Z. Fan, S. Xuan, Z. Hua and G. Liu, Adv. Funct. Mater., 2024, 34, 2410518 CrossRef CAS .
16. L. Gao, W. Jiang, X. Zhang, Y. Sun, K. Chen, W. Li, H. Xie and J. Liu, Chem. Eng. J., 2024, 479, 147822 CrossRef CAS .
17. W. Zhao, Y. Zheng, A. Huang, M. Jiang, L. Wang, Q. Zhang and W. Jiang, Adv. Mater., 2024, 36, 2402386 CrossRef CAS .
18. B. Qin, S. Zhang, P. Sun, B. Tang, Z. Yin, X. Cao, Q. Chen, J.-F. Xu and X. Zhang, Adv. Mater., 2020, 32, 2000096 CrossRef CAS PubMed .
19. J. Lai, M. Xie, Q. Zhao, C. Zhang, Z. Wang and H. Xia, Adv. Funct. Mater., 2025, 35, 2415732 CrossRef CAS .
20. Y. Zhang, S. Zhang, M. Zhai, B. Wei, B. Lyu and L. Liu, ACS Appl. Polym. Mater., 2024, 6, 8399–8408 CrossRef CAS .
21. D. H. Ho, Y. M. Kim, U. J. Kim, K. S. Yu, J. H. Kwon, H. C. Moon and J. H. Cho, Adv. Energy Mater., 2023, 13, 2301133 CrossRef CAS .
22. Y. Tian, Y. Wei, M. Wang, J. Wang, X. Li, X. Qin and L. Zhang, Nano Energy, 2025, 139, 110908 CrossRef CAS .
23. J. Chen, Y. Gao, L. Shi, W. Yu, Z. Sun, Y. Zhou, S. Liu, H. Mao, D. Zhang, T. Lu, Q. Chen, D. Yu and S. Ding, Nat. Commun., 2022, 13, 4868 CrossRef CAS .
24. S. Peng, N. Thirunavukkarasu, J. Chen, X. Zheng, C. Long, X. Huang, Z. Weng, L. Zheng, H. Wang, X. Peng and L. Wu, Chem. Eng. J., 2023, 463, 142312 CrossRef CAS .
25. R. Zhou, Y. Jin, W. Zeng, H. Jin, L. Shi, L. Bai and X. Shang, Adv. Funct. Mater., 2023, 33, 2301921 CrossRef CAS .
26. T. Ye, J. Tan, T. Wu, F. Zhang, S. Chen and C. Wang, Sci. China:Chem., 2025, 68, 1998–2009 CrossRef CAS .
27. X. Xun, Z. Zhang, X. Zhao, B. Zhao, F. Gao, Z. Kang, Q. Liao and Y. Zhang, ACS Nano, 2020, 14, 9066–9072 CrossRef CAS .
28. Y. Zhu, Y. He, W. Lu, H. Tian, F. Fei, P. Zhou and J. Wang, J. Mater. Chem. A, 2024, 12, 28716–28730 RSC .
29. F. Xu, H. Li and Y. Li, Adv. Mater., 2024, 36, 2412317 CrossRef CAS PubMed .
30. J. Yuan, Y. Zhang, C. Wei and R. Zhu, Adv. Sci., 2023, 10, 2303114 CrossRef .
31. P. Wang, J. Liu, W. Yu, G. Li, C. Meng and S. Guo, Nano Energy, 2022, 103, 107768 CrossRef CAS .
32. L. Liu, J. Li, Z. Tian, X. Hu, H. Wu, X. Chen, L. Zhang and W. Ou-Yang, Nano Energy, 2024, 128, 109817 CrossRef CAS .
33. P. Parashar, M. K. Sharma, B. K. Nahak, A. Khan, W.-Z. Hsu, Y.-H. Tseng, J. R. Chowdhury, Y.-H. Huang, J.-C. Liao, F.-C. Kao and Z.-H. Lin, J. Mater. Chem. A, 2025, 13, 13750–13762 RSC .
34. S. P. Sreenilayam, I. U. Ahad, V. Nicolosi, V. Acinas Garzon and D. Brabazon, Mater. Today, 2020, 32, 147–177 CrossRef .
35. K. Meng, X. Xiao, W. Wei, G. Chen, A. Nashalian, S. Shen, X. Xiao and J. Chen, Adv. Mater., 2022, 34, 2109357 CrossRef CAS PubMed .
36. Y. Zhang, X. Zuo, S. Zhang, Z. Ma and G. Zhang, Compos. Sci. Technol., 2024, 256, 110756 CrossRef .
37. M. Lin, Z. Zheng, L. Yang, M. Luo, L. Fu, B. Lin and C. Xu, Adv. Mater., 2022, 34, 2107309 CrossRef .
38. Y. Lu, H. Zhang, Y. Zhao, H. Liu, Z. Nie, F. Xu, J. Zhu and W. Huang, Adv. Mater., 2024, 36, 2310613 CrossRef .
39. X. Zhu, W. Zhang, G. Lu, H. Zhao and L. Wang, ACS Nano, 2022, 16, 16724–16735 CrossRef PubMed .
40. N. Yu, B. Cheng, Y. Liu, W. Wu, R. K. Y. Li, Z. Liang, F. Cheng and H. Zhao, Small, 2024, 20, 2405700 CrossRef PubMed .
41. B. Li, F. Xu, T. Guan, Y. Li and J. Sun, Adv. Mater., 2023, 35, 2211456 CrossRef .
42. X. Zhu, Y. Hao, L.-F. Huang, H. Zhao and L. Wang, J. Mater. Chem. A, 2024, 12, 26158–26169 RSC .
43. C. Luo, Y. Chen, Z. Huang, M. Fu, W. Ou, T. Huang and K. Yue, Adv. Funct. Mater., 2023, 33, 2304486 CrossRef .
44. Y. Guo, H. Zhang, L. Fang, Z. Wang, W. He, S. Shi, R. Zhang, J. Cheng and P. Wang, Nano Energy, 2024, 123, 109427 CrossRef CAS .

This journal is © The Royal Society of Chemistry 2025