Development of highly robust polyurethane elastomers possessing self-healing capabilities for flexible sensors

Abstract

Traditional flexible electronic sensing materials have fallen short in meeting the diverse application needs and environments of modern times. Hence, we require a multi-functional elastomer material to improve the overall performance and expand the functionality of flexible electronic sensors. In this study, we fabricated a multi-block polyurethane (PU) elastomer based on semi-crystalline polycaprolactone (PCL) chain segments and highly flexible polydimethylsiloxane (PDMS) chain segments, which showcases outstanding mechanical properties, self-healing capabilities, and recyclability. By adjusting the ratio parameters of the chain segments, we were able to modulate the thermodynamic behavior, hydrophobicity, mechanical behavior, and self-healing properties of the designed PU elastomers. The optimized ratios exhibited good tensile strength (16.26 MPa), high elongation at break (3300.84%), good toughness (278.82 MJ m⁻³, fracture energy ≈ 234.96 KJ m⁻²), high self-repairing (≈100%, at room temperature for 12 h), efficient recyclability, and puncture resistance. Self-healing is accomplished through the interactions between dynamic disulfide bonds, dynamic boron–oxygen bonds, and hydrogen bonds. The conductive ink (PEDOT:PSS) was encapsulated within this elastomer to construct a flexible electronic sensor, attaining excellent sensing performance (stable output for 1000 cycles). This multi-functional polyurethane elastomer acts as an ideal matrix material for flexible electronic sensors, offering novel concepts and perspectives for the next generation of green electronic flexible materials, electronic flexible robots, and other stimulus-responsive materials.

---

New concepts

In order to advance the development of a sustainable society, it is imperative that next-generation elastomers exhibit exceptional healing and recycling capabilities, in addition to damage tolerance, so as to prolong service life and reduce raw material consumption and environmental pollution. In this work, we have engineered dynamic crosslinked network structures within polydimethylsiloxane (PDMS)/polycaprolactone (PCL) multiblock polymers, showcasing outstanding mechanical self-healing properties, recyclability, flexibility, stretchability, and toughness. These dynamic crosslinked structures are characterized by covalent and hydrogen bonds that are dynamically locked within the crystalline PCL chain segments. This design not only enhances the elastomer's rigidity but also enables efficient energy dissipation through deformation and dissociation processes, thereby significantly improving damage tolerance. The reversibility of these bonds under heating facilitates effective healing and recycling processes that restore the elastomer's original mechanical strength and integrity. Furthermore, the material has been fabricated into a sandwich-structured sensor with exceptional sensing performance. This multifunctional polyurethane elastomer serves as an ideal matrix for flexible electronic sensors, presenting innovative concepts with promising applications for next-generation green electronic flexible materials, electronic flexible robots, and other stimulus-responsive materials.

1. Introduction

In recent years, flexible electronic sensors have seen significant development in various fields such as biomedicine, motion detection, information transmission, and automotive industries.^1–8 Traditional strain sensors that only require stable electrical conductivity and tensile elasticity can no longer meet the evolving needs of these industries. Consequently, the preparation of multifunctional flexible sensing materials has become a focal point of current research. Thermoplastic polyurethane (TPU), as a linear polymer, holds an important position in the scientific community due to its easy processability and excellent mechanical properties.⁹ TPU is typically produced by reacting isocyanate-based end-capped prepolymers with small-molecule chain extenders, which allows for the extension of chain segments. This process enables TPU to achieve different effects by altering the prepolymers and chain extenders, making it widely applicable in industry, the medical field, and flexible electronic materials.^10–12 However, TPU also has limitations that restrict its range of applications, such as low heat resistance, short service life, and low elongation at break.

Polydimethylsiloxane (PDMS) has many special properties such as good flexibility, low surface tension, good hydrophobicity, good thermal stability, and a very low glass transition temperature (T_g, −120 °C) due to the alternating presence of organic and inorganic components in its internal structure and the unique structure of its silicone main chain.^13,14 However, its special main chain structure also gives it poor mechanical strength. In PU, the mechanical properties can be adjusted by reconfiguring the polymer chain segments, for example, by adding crystalline polymer chain segments to enhance the strength and stiffness. Polycaprolactone (PCL), as a well-known semi-crystalline biodegradable biopolymer, is widely used in biomedical materials because of its good biocompatibility, good compatibility with organic polymers, low melting point (about 60 °C), low glass transition temperature (about −60 °C), good biodegradability, and good shape memory temperature-control properties. Moreover, PCL and PDMS have a large interaction parameter (χ), which leads to a strong phase separation behavior between them.^15–18 The high strength and low melting point of PCL and the high flexibility, high melting point, and low strength of PDMS complement each other. As demonstrated in previous studies, the combination of PDMS and PCL has been extensively utilized in self-healing polyurethanes, offering valuable insights that have significantly guided our subsequent research.^19–21 Therefore, it is necessary to discuss in depth the relationship between different ratios of PDMS and PCL on the mechanical properties and thermodynamics of polyurethane elastomers to provide ideas for future high-performance thermoplastic polyurethane elastomers.

However, the wear and tear that PU materials are subjected to during daily use greatly reduce their service life and contribute significantly to environmental pollution. As a result, developing materials with self-healing abilities has become a mainstream approach to extending service life and improving material stability.^22–24 So far, two types of self-healing materials have been reported: (i) exogenous self-healing materials with pre-embedded healing agents (e.g., microcapsules²⁵ and micro vessels²⁶) within the material, and (ii) materials that incorporate reversible chemical bonds (e.g., imine bonds,²⁷ disulfides,²⁸ boron–oxygen bonds,²⁹ Diels–Alder reactions,³⁰ free radical dimerization reactions,³¹ and cycloaddition reactions³²) or dynamic interactions in the polymer chain segments (such as hydrogen bonding,³³ π–π interactions,³⁴ host–guest interactions,³⁵ and metal–ligand coordination³⁶). However, exogenous self-healing materials lose their self-repairing ability after the healing agent is depleted as the number of self-repairs increases. Conversely, it is easier to introduce self-healing groups between chain segments in PUs. Therefore, the introduction of reversible chemical bonds is our primary solution to endow PUs with self-healing capability.

Among these reversible dynamic bonds, aromatic disulfide bonds provide the material with the ability to self-heal at low temperatures because these bonds have a high dynamic exchange rate. Self-healing is achieved by the formation of a cross-linking network between boroxine and boronic acid in the boron–oxygen bond in response to changes in external conditions. This cross-linking network also strengthens the mechanical properties of the material. The key factor in self-repairing is to promote the movement of the chain segments. When the polymer chain segments can move along the damaged interface, the polymer can successfully heal. However, if the polymer chain segment mobility is too strong, it can lead to low crosslinking between the polymer chain segments, thus reducing the mechanical properties. This creates a contradiction between the material's mechanical properties and its self-repairing performance.^37,38 Therefore, it is critical to balance the mechanical and self-repairing properties of the material. A rational design is needed so that an elastomer with both high mechanical and self-healing properties can be synthesized.

In this work, we prepared a multifunctional polyurethane elastomer consisting of PCL and PDMS as soft segments, isophorone diisocyanate (IPDI) as hard segments, and 4-(hydroxymethyl) phenylboronic acid (HMBA) and 2,2′-diaminodiphenyl disulphides (DPS) as chain extenders. The effects of different PCL and PDMS contents on the mechanical properties, thermal behavior, and hydrophobicity of the elastomers were systematically investigated. The relationship between different HMBA and DPS contents was explored to achieve a balance between mechanical and self-healing properties. Due to the crystallinity of PCL, the material exhibits shape memory functionality. The rational design of the elastomer chain segments ensures recyclability. Finally, its potential in flexible electronic sensors was demonstrated by applying conductive ink (PEDOT:PSS) uniformly on the surface of the material, followed by encapsulation.

2. Experimental section

2.1 Materials

Polycaprolactone diol (HO-PCL-OH, M_n = 2000 Da) was purchased from Shanghai Dibo Chemicals Technology Co., Ltd (Shanghai, China). Bis(hydroxyalkyl)-terminated poly(dimethylsiloxane) (HO-PDMS-OH, M_n = 4200 Da) was purchased from Sigma Aldrich (Shanghai, China). Di-n-butyltin dilaurate (DBTDL) was purchased from Beijing InnoChem Science & Technology Co., Ltd (Beijing, China). Tetrahydrofuran (THF), isophorone diisocyanate (IPDI), 4-(hydroxymethyl)phenylboronicacid (HMBA) and 2,2′-diaminodiphenyl (DPS) disulphide were purchased from Shanghai Titan Scientific Co., Ltd (Shanghai, China).

2.2 Preparation of the elastomers

Synthesis of prepolymers. HO-PCL-OH (2 g, 1 mmol) and HO-PDMS-OH (4.2 g, 1 mmol) were dissolved uniformly in THF. Subsequently, they were gradually heated in N₂ to 40 °C. IPDI (0.89 g, 4 mmol) and DBTDL (0.10 g, 0.16 mmol) were added to the solution. Employing an identical preparation approach, by regulating the ratio of PCL and PDMS for the pair's preparation, the prepolymers of PU-1, PU-2, and PU-3 were synthesized. The specific feedings were presented as follows: HO-PCL-OH (2 g, 1 mmol) and HO-PDMS-OH (0.46 g, 0.111 mmol) for PU-1; HO-PCL-OH (2 g, 1 mmol) and HO-PDMS-OH (1.05 g, 0.25 mmol) for PU-2; HO-PCL-OH (2 g, 1 mmol) and HO-PDMS-OH (1.8 g, 0.429 mmol) for PU-3.

Synthesis of PUBS-3. For the synthesis of PUBS-3, DPS (0.248 g, 1 mmol) was dissolved in THF, the uniformly dissolved solution was added to the prepolymer, and was stirred at 40 °C in a N₂ atmosphere for 24 h. Subsequently, HMBA (0.152 g, 1 mmol) was dissolved in methanol, the uniformly dissolved solution was added to the mixture, and was stirred at 40 °C in an N₂ atmosphere for 24 h. The reacted solution was carefully transferred into a PTFE mold, whereupon the solvents volatilized at room temperature for 24 h and further dried in vacuum for 12 h at 60 °C, resulting in the formation of elastomer PU-B_0.5S_0.5. Following the same procedure, PUBS-1, PUBS-2, and PUBS-4 were synthesized by changing the feed mass of DPS and HMBA as follows: DPS (0.207 g, 0.8325 mmol) and HMBA (0.042 g, 0.2775 mmol) for PUBS-2; DPS (0.276 g, 1.11 mmol) and HMBA (0.00 g, 0.00 mmol) for PUBS-1; DPS (0.069 g, 0.2775 mmol) and HMBA (0.127 g, 0.8325 mmol) for PUBS-4.

2.3 Characterization

The thermogravimetric analyzer (TA Instruments Model Q2000) was utilized to analyze the thermal stability of the samples within a nitrogen atmosphere. Each sample (3–8 mg) was heated from 30 °C to 800 °C at a rate of 20 °C min⁻¹. The differential scanning calorimetry (DSC) test was conducted using TA Instrument – Waters LLC, where the sample was heated to 100 °C, cooled to −80 °C, and then reheated to 200 °C under nitrogen, with a rise/fall rate of 10 °C min⁻¹ during the test. The contact angle was measured through a droplet shape analyzer (Kruss DSA30). The infrared thermal imaging instrument (LT7-P) was employed to record the samples. The photos were captured with a Canon EOS80D digital camera. The sensing performance was tested with a Keithley 2450 digital multimeter.

2.4 Mechanical properties

The high-precision tensile and pressure testing machine (HY-025CS, Shanghai Hengyu Instrument Co., Ltd) was adopted to test the fatigue resistance of standard dumbbell-type samples by means of tensile cycling. The same deformation amount (600%) was controlled for 8 cycles and different deformation amounts (ranging from 100% to 1500%) were controlled for 7 cycles. The electronic universal testing machine (SHIMADZU, AGS-X-10kN) was utilized to conduct the tensile test on the dumbbell sample at a rate of 100 mm min⁻¹. The fracture strength, fracture strain, fracture toughness, and elastic modulus are calculated from the stress–strain curve obtained through the test. Fracture toughness (W) is the energy dissipated during the tensile process of the sample before fracture, obtained by the following formula for the area under the stress–strain curve:where σ is the stress and ε is the strain. The fracture energy (G_c) is calculated by the method reported in the literature, and the formula is as follows:³⁹

In this formula, λ_c is the fracture strain of the notched specimen, W is the energy obtained by integrating the stress–strain curve of the intact specimen from 0 to λ_c, and C is the notch length in the notched specimen.

2.5 Self-healing experiments

After the specimen was pulled apart by the electronic universal testing machine, it was re-contacted at room temperature and 80 °C for diverse durations. The healed sample was stretched using the same method as in Section 2.4 to test the recovery of mechanical properties. Additionally, in order to observe the self-healing situation more directly, the film was cut and restored. The film was severed with a blade and then heated, and the image of the material self-healing process was observed with a metallographic microscope (BMM-580V).

3. Results and discussion

3.1 Synthesis and structural characterization of PUBS elastomers

As depicted in Fig. 1a, a thermoplastic polyurethane elastomer PUBS was designed. Polycaprolactone diol (PCL) and bis(hydroxyalkyl)-terminated poly(dimethylsiloxane) (PDMS) were employed as soft segments, while isophorone diisocyanate (IPDI) was utilized as the hard segment. Subsequently, the chain was elongated with 4-(hydroxymethyl) phenylboronic acid (HMBA) and 2,2′-diaminodiphenyl disulphide amines (DPS) serving as chain extenders. The polyurethane (PU) integrates the high flexibility of the PDMS chain segment with the semi-crystalline characteristics of PCL. The micro curvature of the crystalline PCL, acting as the physical crosslinking point within the chain, enhances the mechanical strength of PU, while the high flexibility of PDMS imparts excellent tensile properties to PU. As chain extenders, HMBA and DPS not only introduce hydrogen bonds between carbamate bonds and urea bonds within the chain segment but also incorporate disulfide bonds, boron–oxygen bonds, and π–π interactions, enabling PU materials to form a dynamic cross-linked network and possess the capability of self-healing (Fig. 1b).

Fig. 1 (a) Synthesis of PUBS. (b) Schematic diagram of the segments and dynamic bonds in PUBS.

The preparation of PUs was verified by ATR-FT-IR (Fig. 2a), ¹H NMR (Fig. 2b), and EDS (Fig. 2c). In the infrared spectrum of PUBS-3, it can be initially observed that the characteristic peak of the N=C=O group on IPDI at 2242 cm⁻¹ vanished on PU, and the characteristic peak detected at 3360 cm⁻¹ corresponded to the stretching vibration of the N–H bond in the carbamic ester bond. The characteristic peaks at 1538 cm⁻¹ pertained to the bending vibration of N–H. Additionally, the characteristic peaks at 2950 and 1727 cm⁻¹ respectively corresponded to the stretching vibration of the C–H bond and the stretching vibration of the C=O bond in the carbamate bond, indicating that the hydroxyl groups on PDMS and PCL as soft segments reacted completely with the N=C=O groups on IPDI. The success of the prepolymer synthesis for the PU elastomer was confirmed by the formation of the –C(=O)–NH– group. The C–C main chain vibration peaks of the benzene ring at 1636 cm⁻¹ and 1560 cm⁻¹ confirmed the presence of DPS in the PU main chains. The emergence of new weak bands at 734 and 681 cm⁻¹, a phenomenon distinctive to boroxine formation, substantiated that HMBA was already present in the main PU-4 chain (Fig. S1, ESI†).²⁹ The H atomic peak in PCL and PDMS and the typical chemical shift peak of C–H in the benzene rings of HMBA and DPS can be discerned in ¹H NMR. EDS analysis revealed that carbon (C), nitrogen (N), oxygen (O), silicon (Si), and sulfur (S) were uniformly distributed, thereby proving the successful synthesis of PUBS-3.

Fig. 2 (a) FT-IR spectra of PUBS-3. (b) ¹H NMR spectra of PUBS-3. (c) Distribution of elements in PUBS-3.

3.2 Thermal behavior and hydrophobicity of PU-x elastomers

We investigated the influence of PCL and PDMS on the elastomers at varying ratios. These materials, designated as PU-x, were prepared with specific compositional variations as detailed in Table S1 (ESI†). The crystallization temperature (T_g) and melting temperature (T_m) of PUs can be seen from the DSC curves in Fig. 3a and b. The values of enthalpy of melting, enthalpy of crystallization, and degree of crystallinity of the elastomers were obtained through calculations (Table S3, ESI†). The degree of crystallinity was calculated in relation to the thermodynamic enthalpy of melting of perfect crystals of PCL and PDMS, with corresponding values of 139 J g⁻¹ and 63.8 J g⁻¹ for PCL and PDMS, respectively. We found that as the proportion of PDMS in PUs gradually increased, the melting temperature of the PUs also gradually increased, while the degree of crystallinity decreased. Conversely, as the proportion of PCL increased, the enthalpy of crystallization of the PUs increased, and the crystallization temperature also rose. For example, in PU-1 and PU-4, the molar ratio of PCL increased from 50% to 90%, and the crystallinity increased from 21.66% to 45.06%.

Fig. 3 (a) DSC crystallization temperature curves of PU-x. (b) DSC melting temperature curves of PU-x. (c) TGA curves of PU-x. (d) Infrared thermography of PU-1, PU-3, and PU-4, respectively. (e) Plot of thermal infrared imaging temperature for PU-1, PU-3 and PU-4 as a function of time during the heating process. (f) Dynamic contact angle of PU-1. (g) Image of PU-4 contact angle magnitude as a function of time. (h) AFM images of PU-x: (i) PU-1. (ii) PU-2. (iii) PU-3. (iv) PU-4. (I) Storage modulus, loss modulus and tan δ of PU-4 measured by DMA.

The thermal stability of PU elastomers was examined using TGA, and the results were shown in Fig. 3c. The weight loss before 200 °C was attributed to the volatilization of solvents and the small molecules incompletely reacted, which was not very significant. In the TGA curves, comparing the changes among three samples PU-1, PU-3 and PU-4, we can see that these samples started to degrade at 238 °C, 256 °C, and 276 °C, respectively. This stage mainly corresponded to the decomposition of the hard segments of the polyurethane, as the bond energy of C–N was weaker than that of C–O and C–C, making it easier to decompose as the temperature rises. As the temperature increased, the degradation temperature became closely related to the content of PDMS. On one hand, the bond energy of the Si–O bond is higher than that of the C–C chain, enhancing heat resistance in the PU elastomers. On the other hand, when PU elastomers form a crosslinked network structure, and the PDMS is uniformly dispersed between the chain segments. This homogeneous chain segment structure acts like a protective layer of siloxane, shielding the chain segments from decomposition. Therefore, as the PDMS content gradually increased, the heat resistance was improved from 471.71 °C to 567.17 °C, an increase of 20.24%, compared to PU-1 and PU-3. We used an infrared thermography camera to record the infrared thermography images and temperature changes of three PU elastomer samples, PU-1, PU-3 and PU-4 (Fig. 3d and e). The target temperature of the heating table was set to 90 °C, and the PU elastomers were placed on the heating stage for 30 s. It was observed that the heating rate of the elastomer gradually decreased as the PDMS content increased. This is because the thermal conductivity of PDMS is only 0.134 W (m K)⁻¹, which is lower than that of PCL, 0.2 W (m K)⁻¹. The maximum temperatures of the three groups of samples were 84 °C, 76 °C, and 68 °C, respectively, after heating for 30 s. This makes the PU elastomers more susceptible to temperature change, giving them great potential for use at high temperatures.

The significant changes observed in the DSC and TGA curves, as well as the thermal imaging images, indicated that as the proportion of PDMS in PUs gradually increased, both heat resistance and thermal stability were markedly enhanced. PU elastomers with this improved thermal stability can meet the requirements of most application environments.

Using water droplets to test the contact angle on the surface of a material can characterize the material's hydrophilicity, and the data derived from the contact angle can visualize changes in the material's surface and structure.⁴⁰ In this regard, we conducted dynamic and static contact angle tests on PU elastomers. A 50 μl microsyringe was used to place a water droplet on the sample surface, and the dynamic contact angles (Fig. S3–S6, ESI†) represent the samples PU-1 (Fig. S3 and S4, ESI†), PU-3 (Fig. S5 and S6, ESI†), and PU-4 (Fig. 3f and g), respectively. The static contact angle was tested after the stabilization of the water droplets, as shown in Fig. S2 (ESI†). The contact angle of the water droplets was used to calculate the dispersive surface energy of the samples (γ^d_S), and according to Fowkes’ study, the surface tension can be calculated by the following equations:

γ_SL = γ_SV − γ_LV cos θ (4)

where γ_LV represents the surface tension at the liquid–gas interface, γ_SL represents the surface tension at the solid–liquid interface, and γ_SV represents the surface tension at the solid–gas interface. Simplifying eqn (3) and (4), we can derive the dispersive surface energy equation.

γ^d_S = [γ_LV(1 + cos θ)]²/4γ^d_L (5)

Because we used water droplets as the test fluid, γ_LV = 72.7 mN m⁻¹ and γ^d_L = 23.9 mN m⁻¹, which resulted in the final data (Fig. S2, ESI†).⁴¹

From the dynamic contact angle, we observed that with the extension of time, the contact angle of PU elastomers gradually decreased and tended to stabilize. When the ratio of PDMS increased, the rate of change slowed down. For example, in PU-1, the contact angle decreased from 107.4° at 2 s to 95.3° at 60 s, with a loss rate of 11.27%. In contrast, for PU-4, the contact angle decreased from 110.2° at 2 s to 104.5° at 60 s, with a loss rate of 5.17%. In the static contact angle measurements, comparing PU-1 and PU-4, the interfacial energy of PU-1 was 42.13 mN m⁻¹, while that of PU-4 was 16.58 mN m⁻¹. This indicated a reduction of 60.64% in interfacial energy for PU-4 compared to PU-1, which corresponded to a higher contact angle. The results suggested that a lower surface energy corresponded to a larger contact angle. From the dynamic and static contact angle results, it was evident that as the PDMS content increased, the structure and composition of the PU elastomer surface changed, leading to variations in the sample properties. This was due to the nonpolar nature of PDMS causing microphase separation within the internal structure. Additionally, the significant thermodynamic difference between PDMS and PCL resulted in PDMS chain segments near the surface layer tending to migrate to the surface of the PU elastomers. The unique inorganic backbone Si–O–Si and organic side chain –CH₃ in PDMS cause the inorganic backbone to protrude to the surface when PDMS partially migrated, forming a low-surface-energy and hydrophobic surface. Meanwhile, the organic side chain tended to interact with the internal structure of the material. Consequently, this phenomenon occurred more prominently as the PDMS content increased, endowing PU elastomers with unique surface properties.

The phase morphology of the PU-x elastomers was investigated via atomic force microscopy (AFM). Fig. 3h presented the distribution of soft and hard segments along with the phase-separated structure at diverse PCL and PDMS ratios. When the PCL content was high (Fig. 3h(i)), the PU surface was predominantly covered by crystalline PCL domains. The darker areas in the images signify PDMS-rich regions, which arose due to the aggregation of siloxane chains at room temperature. This aggregation was driven by the high mobility and compressibility of siloxanes. Furthermore, as the PDMS content increased, the darker regions became more pronounced, indicating a greater surface coverage by PDMS. This was most likely attributed to the significantly lower surface energy and extreme backbone flexibility of PDMS, enabling better spreading on the surface compared to the crystalline PCL segments.

We undertook to carry out a more elaborate analysis of the microphase separation and crystalline behaviors of PU-x elastomers through the employment of XRD. Fig. S7 (ESI†) showcased the XRD patterns of PCL and PDMS at diverse ratios. As illustrated in Fig. S7 (ESI†), the crystallinity of PCL was characterized by the existence of two small but distinct XRD peaks at 2θ ≈ 21.5° and 23.7°. It was obvious that the intensity of these peaks augments with the copolymer content, attaining a maximum at the PCL threshold. This finding indicated the enhanced crystallinity of the PCL phase. Additionally, a broad amorphous background peak was observed, with a maximum at 2θ ≈ 12°. These XRD outcomes disclosed the coexistence of amorphous PDMS and crystalline PCL chain segments, a clearly defined microphase separation in the PU-x elastomers.

The viscoelastic behavior of PU-4 was characterized via DMA. The tan δ curves (Fig. 3f) disclosed a significant loss peak for PU-4 at −23.7 °C and a minor one near −65 °C, respectively corresponding to the glass transition temperatures (T_g) of the PCL and PDMS phases. These T_g values were higher compared to those of the pure soft segments (−60 °C for PCL and −120 °C for PDMS). This can be ascribed to the augmented molecular weight and the integration of urethane hard segments, both of which constrained chain mobility. The storage modulus curves of PU-4 (Fig. 3f) manifested two distinct descents, corresponding to the glass transition and melting of the PCL chain segments. Remarkably, the storage modulus persisted above 1 MPa even when the temperature exceeded 100 °C, and the material sustained its elastomeric state above 115 °C.

3.3 Mechanical performance of PUBS-x elastomers

Flexible electronic materials are often repeated many times in use. To compare the mechanical properties between PU elastomers, we controlled the molar ratios of the chain extenders HMBA and DPS to be equal on one hand to explore the relationship between PCL and PDMS with different molar ratios (Fig. 4a and Table S4, ESI†), and on the other hand, we controlled the molar ratios of the soft segments PCL and PDMS to be equal on one hand to explore the effects of different molar ratios of the chain extenders HMBA and DPS (Fig. 4b and Table S5, ESI†). First, we controlled the thickness of all test specimens to be ∼1 mm, used the same tensile rate of 100 mm min⁻¹, and performed three repetitions of each sample to select its average value. With different PCL and PDMS molar ratios, we found that the prepared PUs exhibited classical elastomeric behavior. They showed no obvious yielding phenomenon in the stress–strain curves and displayed strain hardening after elastic deformation until fracture. Interestingly, compared to PU-1, the mechanical curves of PU-2, PU-3, and PU-4 reflected better resilience. The difference observed in the PU-1 samples can be attributed to the fact that PDMS comprised only 10% of the soft segments. From a microscopic perspective, PCL, with its semi-crystalline nature, dominated the chain segments. This made the mechanical behavior of PU-1 significantly different from the other samples. As the PDMS content in the soft segments increased in PU-2, PU-3 and PU-4, the fracture strength, elongation at break, and toughness also increased. This was because a higher PDMS content leads to a more uniform distribution of PDMS and PCL within the chain segments. The Si–O bonds in PDMS, characterized by their unique flexibility, were better distributed between the chain segments, enhancing the elongation at break and elastic modulus of the PU elastomers. Additionally, the crystallinity of PCL was better utilized, improving the mechanical properties of PDMS. Comparing PU-1 with PU-4, the tensile strength increased from 9.36 MPa to 16.26 MPa, the toughness from 60.28 MJ m⁻³ to 278.82 MJ m⁻³, and the elongation at break from 1932.52% to 3300.84%. This represented a 67.1% increase in tensile strength, a 362.5% increase in toughness, and a 70.8% increase in elongation at break. Therefore, the mechanical properties were optimized when the molar ratio of PCL to PDMS was 1 : 1.

Fig. 4 (a) Stress–strain curves of PU-x. (b) Stress–strain curves of PUBS-x. (c) Schematic illustration of the single-edge notched elastomer used for determining the fracture energy. (d) Tensile stress–strain curve of notched PU-4 compared with that of the original one. (e) Continuous cyclic tensile curves of PU-4 at 600% strain. (f) Comparison of PU-4 with other reported elastic materials regarding fracture energy, strain, and toughness.

Among the different molar ratios of chain extenders HMBA and DPS, we selected the best mechanical properties at a PCL to PDMS molar ratio of 1 : 1. Observations from the images showed that, unlike the varying molar ratios of PCL and PDMS, there was no significant difference in the images under different molar ratios of chain extenders, all displaying classical elastomer behavior. However, as the HMBA ratio increased, the mechanical strength showed a gradual upward trend, while the elongation at break decreased. This was due to the formation of a robust cross-linked 3D network structure between boron and oxygen during the chain extension process. This cross-linked network enhanced the mechanical strength and toughness of the material, but also restricted chain segment extension, resulting in a decrease in elongation at break. For example, comparing PUBS-2 and PUBS-4, the tensile strength increased from 14.19 MPa to 16.90 MPa, a 19.1% increase, and the toughness increased from 263.71 MJ m⁻³ to 295.63 MJ m⁻³, a 12.1% increase, when the HMBA content increased from 25% to 75%. The elongation at break decreased from 3497.24% to 2779.06%, a reduction of 20.5%. When DPS was used as the sole chain extender, the strong cross-linked 3D network structure between boron and oxygen was absent. The chain segments were only cross-linked through hydrogen bonding and π–π interactions, leading to a decrease in mechanical strength and an increase in elongation at break. Therefore, at an HMBA to DPS molar ratio of 1 : 1, the mechanical strength, toughness, and elongation at break are optimized.

Based on previous results of mechanical strength and elongation at break for different ratios, we selected PU-4 to further investigate the toughness of PU elastomers. The original samples were quartered to prepare notched specimens (Fig. 4d). Tensile testing of these notched specimens revealed that the cracks were deflected and became obtuse, suggesting that the polymer network internally prevented crack propagation and extension. The maximum tensile strength of the notched specimens was 10.61 MPa, the fracture toughness was 163.33 MJ m⁻³, and the calculated fracture energy was 234.96 KJ m⁻². In comparison, we compared the elongation at break, toughness, and fracture energy data of representative self-healing PU materials.^42–52 The performance of PU-4 in this study surpassed that of the reported materials (Fig. 4f).

The strengthening and toughening mechanisms of PU-4 elastomers were further investigated through cyclic tensile testing (Fig. S8, ESI†). The hysteresis area for each loading–unloading cycle was plotted as a function of strain in Fig. S9 (ESI†). At low strains, the hysteresis area showed little change, suggesting that the material undergoes primarily time-dependent elastic deformation. As the strain increased, the hysteresis area grew significantly, demonstrating efficient energy dissipation, which enhanced the toughness and damage tolerance of the elastomer. After eight cycles at 600% strain, the material retained 94.91% of its initial stress, and the cyclic hysteresis area remained at 91% of the first cycle (Fig. 4h). These results indicated that the elastomer exhibited notable fatigue resistance.

3.4 Self-healing properties of PUBS-x elastomers

Next, we examine the influence of disulfide and boron–oxygen bonds on the elastomer at varying ratios. These materials, designated as PUBS-x, were prepared with specific compositional variations as detailed in Table S2 of the ESI.† As a PU elastomer material, it is inevitable to encounter various degrees of damage in our daily life. For this reason, we have incorporated self-healing capabilities into the material to extend its service life and reduce waste generation, thereby achieving environmental protection. As shown in Fig. 5a, PUBS-x elastomers, through dynamic boron–oxygen bonding, dynamic disulfide bonding, and the formation of abundant multiple hydrogen bonds between carbonyl, ureido, and carbamate groups within the chain segments, enhanced the material's self-healing ability. Additionally, the spatially asymmetric configuration of the hard-segmented IPDI and cis-chair-type chain extenders reduced the crystalline orientation of the polyurethanes, resulting in an irregular chain arrangement within the molecules. This ensured sufficient chain mobility within the molecular network, which facilitated the rearrangement of chain segments to achieve self-repairing. This rational design between chain segments balanced the mechanical properties and self-healing performance of PU elastomers, endowing them with excellent mechanical properties while also achieving outstanding self-healing performance.

Fig. 5 (a) Self-healing mechanism diagram of PU elastomers. (b) Selected spinning and stretching diagrams self-healed for different colors of PUBS-3. (c) Stress–strain curves of PUBS-3 after self-healing at room temperature and 80 °C. (d) Stress–strain curves of PUBS-3 under multiple self-healing cycles. (e) SEM images of scratched PUBS-3 during the healing process. (f) Metallographic microscopy image of the self-healing process of PUBS-3.

Through self-healing experimental tests, we verified the self-repairing performance of PU elastomers. We chose PUBS-3 elastomer as the test sample. First, the elastomer was cut into two sections, and one half was stained with Congo red. After contacting the fractured sections and allowing self-healing at room temperature for 3 h, the self-repaired samples were subjected to simple tensile and torsion experiments and then tested on a universal testing machine for tensile testing (Fig. 5b). The results indicated that the elastomer possessed basic self-healing capabilities. Self-repairing occurs primarily through dynamic boron–oxygen bonding, dynamic disulfide bonding, and hydrogen bonding within the chain segments. To verify the self-repairing performance among PU elastomers, we tested different ratios of boron–oxygen and disulfide bonds (Fig. 5c and Fig. S10–S12, ESI†). Firstly, we explored the effect of self-repairing duration on self-repairing efficiency by comparing PUBS-1 and PUBS-2 (Fig. S10 and S11, ESI†). The specific data are presented in Tables S6 and S7 (ESI†).

The former contains only dynamic disulfide and hydrogen bonds, while the latter also includes dynamic boron–oxygen bonds. At room temperature, after 3 h of self-repairing, the efficiency of PUBS-1 reached 63.12%, while that of PUBS-2 was 59.59%. However, with increased time, the self-repairing efficiency of PUBS-2 gradually surpassed that of PUBS-1. After 12 h, the self-repairing efficiency of PUBS-2 reached 98.64%, compared to 94.78% for PUBS-1. This is because in PUBS-1, under the self-repairing environment of dynamic disulfide bonding, supplemented by hydrogen bonding, the chain segments quickly form cross-links. However, as time progressed, the cross-linking density reached its upper limit, gradually slowing down the self-repairing efficiency. In PUBS-2, the addition of dynamic boron–oxygen bonds resulted in two dynamic bonds cross-linking simultaneously within the polymer chain segments in a short period of time, delaying the flow of the chain segments. Over time, the chain segments had sufficient time for cross-linking, and with the aid of the additional dynamic bond, the cross-linking of the chain segments improved, thus enhancing self-repairing efficiency over time. We hypothesize that a single dynamic bond can enhance self-repairing efficiency in the short term. However, over an extended period, the formation of multiple self-repairing networks is more favorable for improving the material's self-repairing efficiency. To further investigate the effects of dynamic disulfide bonding and dynamic boron–oxygen bonding on self-repairing efficiency, we tested PUBS-2, PUBS-3, and PUBS-4 under room temperature conditions after 3 h of self-repairing. The specific data are shown in Tables S7–S9 (ESI†). We found that the self-repairing efficiency of the material gradually decreased with the increase in the ratio of boron–oxygen bonds. This indicated that the cross-linking density of dynamic boron–oxygen bonds formed between the polymer chain segments is higher than that of dynamic disulfide bonds, leading to decreased self-repairing efficiency. The PUBS-4 sample required 4 h to achieve the same 3-hour self-repairing efficiency as the PUBS-3 sample. With the extension of self-repairing time to 12 h, the self-repairing efficiency of the PUBS-3 sample reached 100%, indicating full restoration to the pre-damaged state. In contrast, the PUBS-4 sample required 16 h to achieve 100% self-repairing efficiency. This demonstrated that dynamic boron–oxygen bonding provides strong cross-linking between the polymer chain segments, enhancing the material's mechanical properties while limiting its self-repairing efficiency. Comparing the four groups of samples, PUBS-1, PUBS-2, PUBS-3, and PUBS-4, we found that PUBS-3 exhibited the most superior self-repairing properties. Finally, we compared the self-repairing efficiency at different temperatures. At 80 °C, the material achieved the self-repairing effect obtained at room temperature in a significantly shorter time. This is because elevated temperatures enhance the mobility of the polymer chain segments, overcoming the limitations imposed by cross-linking and accelerating the dynamic bond exchange rate. To evaluate the long-term stability of the self-repairing performance, we performed five consecutive self-repair cycle tests on PUBS-3 samples (Fig. 5d). The results demonstrated that PUBS-3 maintained a self-repairing efficiency of over 98% even after five cycles, indicating excellent stability. For PU elastomers, mechanical properties and self-healing properties are often contradictory; an increase in mechanical properties results in tighter cross-linking within the material, which hinders self-healing. However, from our experiments on mechanical and self-repairing properties, we concluded that PUBS-3 samples perfectly balanced the contradiction between mechanical and self-repairing properties, greatly extending the material's service life and offering significant application potential. Comparing various self-repairing elastomers, the self-healing capability of the PUBS-3 sample surpasses that of previously reported self-repairing elastomers (Table S10, ESI†). Additionally, we used metallographic microscopy and SEM to observe the self-repairing recovery of damaged PUBS-3 samples (Fig. 5e and f). The observation revealed a clear crack in the damaged samples, which disappeared after 3 h of self-repairing at room temperature, indicating successful healing.

3.5 Recycling, shape memory and puncture resistance of PUBS-3 elastomer

With the rapid depletion of global resources and the environmental impact of waste materials, there is an increasing emphasis on material sustainability. Consequently, the development of green polymers has become essential. In PU elastomers, the presence of a stable hydrogen bonding network within the polymer chain segments, combined with the synergistic effects of dynamic boron–oxygen and dynamic disulfide bonds, contributes to their unique dynamic structure. This structure facilitates the rearrangement of the polymer network, ensuring excellent recyclability and reprocess ability.

As demonstrated in Fig. 6a, we cut PUBS-3 films into small pieces and dissolved them in a vial containing tetrahydrofuran (THF). After stirring for 10 min at room temperature, the sample completely dissolved in THF. The resulting solution was poured onto a PTFE mold, and after complete evaporation of the solvent, a new PUBS-3 film was obtained. We evaluated the tensile properties of the recycled PUBS-3 films over multiple cycles. The mechanical properties of the films after three recycling cycles were comparable to those of the original films (Fig. 6b). The third recycled sample could support a dumbbell weight 12 500 times its own weight (Fig. 6c). Additionally, puncture resistance tests showed that the recycled samples retained their resistance to punctures from water-based pens even after three recycling cycles (Fig. 6e). The tensile strength, elongation at break, and toughness of the recycled samples were all above 95% of the original values (Fig. 6d), demonstrating their excellent recyclability and reprocessing capability. Due to the high crystallinity of PCL, it serves as a physical cross-linking agent and switchable chain segment, resulting in unidirectional shape memory properties. The shape memory effect arises from entropy changes and structural transformations. In this system, both PDMS and PCL chains within the polymer matrix can align in the direction of the applied stress, thus reducing entropy. Meanwhile, the crystalline PCL microregions and the urethane hard segments stabilize the deformation. Upon heating, the polymer chains revert to a more entropy-favorable, convoluted structure, leading to the restoration of the original shape. In our experiments, we fixed a dumbbell-shaped PU-3 sample into a helical shape by heating it to its melting temperature (T_m) and then setting the shape by cooling while applying force. Upon reheating the sample in an oven to a temperature greater than T_m, the helical dumbbell-shaped sample reverted to its original shape (Fig. 6f).

Fig. 6 (a) Recyclability and reprocess ability of PUBS-3 films. (b) Stress–strain curves of PUBS-3 after different recycling times. (c) PUBS-3 with stands 12 500 times its weight after the third recovery. (d) Comparison of mechanical properties of PUBS-3 films after multiple recycling. (e) Puncture resistance test of the PUBS-3 film. (f) PUBS-3 film was programmed to a temporary spiral shape and then unraveled back to the initial shape.

3.6 Strain sensing properties of PUBS-3 elastomer

Through the rational design of polymer chain segments, PUBS-3 demonstrated excellent mechanical and self-healing properties, making it highly suitable for various applications, particularly as a substrate for flexible sensing materials. To explore its potential, we first pre-stretched the PUBS-3 film and then fabricated a thin-film sensor by uniformly applying a highly transparent and conductive PEDOT:PSS ink to the material surface using a spin-coater, followed by encapsulation into a sandwiched structure (Fig. 7a). The self-healing properties of the PUBS-3 film were tested by connecting it to a small light bulb via an electric wire. When the wire was activated, the bulb lit up, indicating a closed current path. Cutting the film disrupted the current path, causing the bulb to go out. The film was then left to self-heal at room temperature for 30 s, during which the bulb emitted a faint glow. This partial restoration of the current path resulted in increased resistance and thus a dim light. Subsequently, heating the film in an 80 °C oven for 20 min allowed the electrical connections to fully reform, restoring the bulb to its original brightness (Fig. 7b). This demonstrates that the self-healing properties significantly extend the lifespan of the sensing material. As flexibility is crucial for sensors, we performed a cyclic stretching test on the film, stretching it 1000 times from 0% to 100% strain (frequency is once a second). The sensor showed stable signal output (Fig. 7c). Additionally, we evaluated the response time of the film sensor to strain changes, finding that it exhibited a fast response time of 86 ms during stretching and 98 ms during recovery, reflecting high signal transmission speed (Fig. 7d). Further tests assessed the film's performance under varying strains and frequencies. Stretching the film from 25% to 150% strain revealed significant changes in performance (Fig. 7e). Additionally, testing different stretching rates (from 1 Hz to 0.125 Hz) at a fixed strain of 100% showed that the signal peak varied with the stretching rate (Fig. 7f). The strain sensing mechanism revealed that the sensor has a high gauge factor (GF = R/(R₀·ε), where ε represents different strains across various strain ranges. The highest GF value of 29.175 with R² > 93% was observed in the 125–225% strain range, while GF values were 2.9548 with R² > 96% in the low strain range (0–125%) and 15.176 with R² > 99% in the high strain range (225–300%). These results indicate that the sensor has high sensitivity across different strain ranges. Finally, we tested the sensor's ability to detect changes in body movements, such as finger bending and leg bending (Fig. 7g and i). The sensor demonstrates good stability and responsiveness in monitoring different body parts during daily activities.

Fig. 7 (a) Conductive PUBS-3 film preparation process. (b) The small bulb brightness was used to indicate the magnitude of the resistance of the PUBS-3 composite flexible sensors in the initial state, in the state with cut marks, and in the state of healed at room temperature for 30 seconds and at 80 °C for 20 min. (c) Response stability of the strain sensor at a constant strain of 100% for 1000 cycles. (d) Response time and recovery time at 100% strain. (e) Response to change in relative resistance for different stretching conditions. (f) Relative resistance change response at different frequencies at 100% strain. (g) Change in relative resistance of the sensor at different strains. (h) Relative resistance changes in knee flexion. (i) Changes in relative resistance at different angles of finger flexion.

4. Conclusions

In summary, we have successfully developed a self-healing, highly ductile, and recyclable multiblock polymer elastomer by utilizing PCL and PDMS as soft segments and incorporating dynamic boron–oxygen and dynamic disulfide bonding into the polymer chain segments. This material combines the crystallinity of PCL and the flexibility of PDMS, with synergistic multilevel dynamic interactions and phase separation between the soft segments, which contribute to its remarkable mechanical properties. The elastomer demonstrated high tensile strength (16.26 MPa), exceptional elongation at break (3300.84%), and significant toughness (278.82 MJ m⁻³, fracture energy ≈ 234.96 KJ m⁻²), as well as self-healing capabilities. The inclusion of PDMS enhanced the hydrophobicity and heat resistance of the elastomer, while PCL's crystallinity endowed it with shape memory properties. Additionally, the linear chain segment structure allowed for recyclability. The combination of these features makes the material suitable for applications requiring enhanced reliability and extended service life. Finally, we prepared a sandwich-structured flexible sensor using transparent conductive ink encapsulated in PUBS-3 and evaluated its sensing performance. This study offers a novel approach for developing next-generation functional flexible sensors.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of interest

The authors have no competing interests to declare.

Acknowledgements

This work was supported by the Natural Science Foundation of Shanghai (24ZR1427000) and Class III Peak Discipline of Shanghai-Materials Science and Engineering (High-Energy Beam Intelligent Processing and Green Manufacturing).

References

1. N. Ashammakhi, A. L. Hernandez, B. D. Unluturk, S. A. Quintero, N. R. De Barros, E. Hoque Apu, A. Bin Shams, S. Ostrovidov, J. Li, C. Contag, A. S. Gomes and M. Holgado, Biodegradable Implantable Sensors: Materials Design, Fabrication, and Applications, Adv. Funct. Mater., 2021, 31, 2104149,  DOI:10.1002/adfm.202104149.
2. T. Falcucci, K. F. Presley, J. Choi, V. Fizpatrick, J. Barry, J. Kishore Sahoo, J. T. Ly, T. A. Grusenmeyer, M. J. Dalton and D. L. Kaplan, Degradable Silk-Based Subcutaneous Oxygen Sensors, Adv. Funct. Mater., 2022, 32, 2202020,  DOI:10.1002/adfm.202202020.
3. C. Xu, J. Chen, Z. Zhu, M. Liu, R. Lan, X. Chen, W. Tang, Y. Zhang and H. Li, Flexible Pressure Sensors in Human–Machine Interface Applications, Small, 2024, 20, 2306655,  DOI:10.1002/smll.202306655.
4. J. Yi and Y. Xianyu, Gold Nanomaterials-Implemented Wearable Sensors for Healthcare Applications, Adv. Funct. Mater., 2022, 32, 2113012,  DOI:10.1002/adfm.202113012.
5. Z. Shi, L. Meng, X. Shi, H. Li, J. Zhang, Q. Sun, X. Liu, J. Chen and S. Liu, Morphological Engineering of Sensing Materials for Flexible Pressure Sensors and Artificial Intelligence Applications, Nano-Micro Lett., 2022, 14, 141,  DOI:10.1007/s40820-022-00874-w.
6. G. Maduraiveeran and W. Jin, Nanomaterials based electrochemical sensor and biosensor platforms for environmental applications, Trends Environ. Anal. Chem., 2017, 13, 10–23,  DOI:10.1016/j.teac.2017.02.001.
7. S. Pyo, J. Lee, K. Bae, S. Sim and J. Kim, Recent Progress in Flexible Tactile Sensors for Human-Interactive Systems: From Sensors to Advanced Applications, Adv. Mater., 2021, 33, 2005902,  DOI:10.1002/adma.202005902.
8. Y. Zhao and J. Li, Sensor-Based Technologies in Effective Solid Waste Sorting: Successful Applications, Sensor Combination, and Future Directions, Environ. Sci. Technol., 2022, 56, 17531–17544,  DOI:10.1021/acs.est.2c05874.
9. W. Xu, R. Zhang, W. Liu, J. Zhu, X. Dong, H. Guo and G.-H. Hu, A Multiscale Investigation on the Mechanism of Shape Recovery for IPDI to PPDI Hard Segment Substitution in Polyurethane, Macromolecules, 2016, 49, 5931–5944,  DOI:10.1021/acs.macromol.6b01172.
10. S. Zhu, N. Lempesis, P. J. In ‘T Veld and G. C. Rutledge, Molecular Simulation of Thermoplastic Polyurethanes under Large Tensile Deformation, Macromolecules, 2018, 51, 1850–1864,  DOI:10.1021/acs.macromol.7b02367.
11. L. Wan, C. Deng, H. Chen, Z.-Y. Zhao, S.-C. Huang, W.-C. Wei, A.-H. Yang, H.-B. Zhao and Y.-Z. Wang, Flame-retarded thermoplastic polyurethane elastomer: From organic materials to nanocomposites and new prospects, Chem. Eng. J., 2021, 417, 129314,  DOI:10.1016/j.cej.2021.129314.
12. C. Xu and Y. Hong, Rational design of biodegradable thermoplastic polyurethanes for tissue repair, Bioact. Mater., 2022, 15, 250–271,  DOI:10.1016/j.bioactmat.2021.11.029.
13. D. Qi, K. Zhang, G. Tian, B. Jiang and Y. Huang, Stretchable Electronics Based on PDMS Substrates, Adv. Mater., 2021, 33, 2003155,  DOI:10.1002/adma.202003155.
14. J. Fan, J. Huang, Z. Gong, L. Cao and Y. Chen, Toward Robust, Tough, Self-Healable Supramolecular Elastomers for Potential Application in Flexible Substrates, ACS Appl. Mater. Interfaces, 2021, 13, 1135–1144,  DOI:10.1021/acsami.0c15552.
15. M. Labet and W. Thielemans, Synthesis of polycaprolactone: a review, Chem. Soc. Rev., 2009, 38, 3484,  10.1039/b820162p.
16. G. Rivero, L.-T. T. Nguyen, X. K. D. Hillewaere and F. E. Du Prez, One-Pot Thermo-Remendable Shape Memory Polyurethanes, Macromolecules, 2014, 47, 2010–2018,  DOI:10.1021/ma402471c.
17. A. Lendlein and O. E. C. Gould, Reprogrammable recovery and actuation behaviour of shape-memory polymers, Nat. Rev. Mater., 2019, 4, 116–133,  DOI:10.1038/s41578-018-0078-8.
18. M. Dziadek, K. Dziadek, K. Checinska, B. Zagrajczuk, M. Golda-Cepa, M. Brzychczy-Wloch, E. Menaszek, A. Kopec and K. Cholewa-Kowalska, PCL and PCL/bioactive glass biomaterials as carriers for biologically active polyphenolic compounds: Comprehensive physicochemical and biological evaluation, Bioact. Mater., 2021, 6, 1811–1826,  DOI:10.1016/j.bioactmat.2020.11.025.
19. E. Yilgör, M. Isik, C. K. Söz and I. Yilgör, Synthesis and structure-property behavior of polycaprolactone-polydimethylsiloxane-polycaprolactone triblock copolymers, Polymer, 2016, 83, 138–153,  DOI:10.1016/j.polymer.2015.12.024.
20. X. Wang, S. Zhan, Z. Lu, J. Li, X. Yang, Y. Qiao, Y. Men and J. Sun, Healable, Recyclable, and Mechanically Tough Polyurethane Elastomers with Exceptional Damage Tolerance, Adv. Mater., 2020, 32, 2005759,  DOI:10.1002/adma.202005759.
21. Z.-X. Fei, C. Yin, J.-R. Sun, L. Yuan and L.-Y. Shi, Healable and recyclable multiblock polyurethanes with mechanical performance Tailorability based on hierarchical phase separation and dynamic bond interaction, Polymer, 2023, 289, 126467,  DOI:10.1016/j.polymer.2023.126467.
22. K. Cerdan, M. Thys, A. Costa Cornellà, F. Demir, S. Norvez, R. Vendamme, N. Van Den Brande, P. Van Puyvelde and J. Brancart, Sustainability of self-healing polymers: A holistic perspective towards circularity in polymer networks, Prog. Polym. Sci., 2024, 152, 101816,  DOI:10.1016/j.progpolymsci.2024.101816.
23. J. Zhao, J. Zhu, J. Zhang, Z. Huang and D. Qi, Review of research on thermoplastic self-healing polyurethanes, React. Funct. Polym., 2024, 199, 105886,  DOI:10.1016/j.reactfunctpolym.2024.105886.
24. S. An, S. S. Yoon and M. W. Lee, Self-Healing Structural Materials, Polymers, 2021, 13, 2297,  DOI:10.3390/polym13142297.
25. H. C. Erythropel, J. B. Zimmerman, T. M. De Winter, L. Petitjean, F. Melnikov, C. H. Lam, A. W. Lounsbury, K. E. Mellor, N. Z. Janković, Q. Tu, L. N. Pincus, M. M. Falinski, W. Shi, P. Coish, D. L. Plata and P. T. Anastas, The Green ChemisTREE: 20 years after taking root with the 12 principles, Green Chem., 2018, 20, 1929–1961,  10.1039/C8GC00482J.
26. Q. Sun, X.-Y. Wang, S. Wang, R.-Y. Shao and J.-F. Su, Investigation of Asphalt Self-Healing Capability Using Microvasculars Containing Rejuvenator: Effects of Microvascular Content, Self-Healing Time and Temperature, Materials, 2023, 16, 4746,  DOI:10.3390/ma16134746.
27. L. Bai, X. Yan, B. Feng and J. Zheng, Mechanically strong, healable, and reprocessable conductive carbon black/silicone elastomer nanocomposites based on dynamic imine bonds and sacrificial coordination bonds, Composites, Part B, 2021, 223, 109123,  DOI:10.1016/j.compositesb.2021.109123.
28. S. Zheng, H. Xue, J. Yao, Y. Chen, M. A. Brook, M. E. Noman and Z. Cao, Exploring Lipoic Acid-Mediated Dynamic Bottlebrush Elastomers as a New Platform for the Design of High-Performance Thermally Conductive Materials, ACS Appl. Mater. Interfaces, 2023, 15, 41043–41054,  DOI:10.1021/acsami.3c09826.
29. J.-C. Lai, J.-F. Mei, X.-Y. Jia, C.-H. Li, X.-Z. You and Z. Bao, A Stiff and Healable Polymer Based on Dynamic-Covalent Boroxine Bonds, Adv. Mater., 2016, 28, 8277–8282,  DOI:10.1002/adma.201602332.
30. X. Chen, M. A. Dam, K. Ono, A. Mal, H. Shen, S. R. Nutt, K. Sheran and F. Wudl, A Thermally Re-mendable Cross-Linked Polymeric Material, Science, 2002, 295, 1698–1702,  DOI:10.1126/science.1065879.
31. B. Ghosh and M. W. Urban, Self-Repairing Oxetane-Substituted Chitosan Polyurethane Networks, Science, 2009, 323, 1458–1460,  DOI:10.1126/science.1167391.
32. C.-M. Chung, Y.-S. Roh, S.-Y. Cho and J.-G. Kim, Crack Healing in Polymeric Materials via Photochemical [2+2] Cycloaddition, Chem. Mater., 2004, 16, 3982–3984,  DOI:10.1021/cm049394+.
33. H. Jiang, T. Yan, W. Pang, M. Cheng, Z. Zhao, T. He, Z. Wang, C. Li, S. Sun and S. Hu, Incomplete ionic interactions and hydrogen bonds constructing elastomers with water accelerated Self-Healing and self-healing strengthening capacities, Chem. Eng. J., 2024, 489, 151074,  DOI:10.1016/j.cej.2024.151074.
34. S. Burattini, B. W. Greenland, W. Hayes, M. E. Mackay, S. J. Rowan and H. M. Colquhoun, A Supramolecular Polymer Based on Tweezer-Type π–π Stacking Interactions: Molecular Design for Healability and Enhanced Toughness, Chem. Mater., 2011, 23, 6–8,  DOI:10.1021/cm102963k.
35. Advanced Materials – 2013 – Kakuta – Preorganized Hydrogel Self-Healing Properties of Supramolecular Hydrogels Formed by.pdf, (n.d.).
36. S. Bode, L. Zedler, F. H. Schacher, B. Dietzek, M. Schmitt, J. Popp, M. D. Hager and U. S. Schubert, Self-Healing Polymer Coatings Based on Crosslinked Metallosupramolecular Copolymers, Adv. Mater., 2013, 25, 1634–1638,  DOI:10.1002/adma.201203865.
37. L. He, J. Shi, B. Tian, H. Zhu and W. Wu, Self-healing materials for flexible and stretchable electronics, Mater. Today Phys., 2024, 44, 101448,  DOI:10.1016/j.mtphys.2024.101448.
38. H. Xu, J. Tu, H. Li, J. Ji, L. Liang, J. Tian and X. Guo, Room-temperature self-healing, high ductility, recyclable polyurethane elastomer fabricated via asymmetric dynamic hard segments strategy combined with self-cleaning function application, Chem. Eng. J., 2023, 454, 140101,  DOI:10.1016/j.cej.2022.140101.
39. H. W. Greensmith, Rupture of rubber. X. The change in stored energy on making a small cut in a test piece held in simple extension, J. Appl. Polym. Sci., 1963, 7, 993–1002,  DOI:10.1002/app.1963.070070316.
40. A. Mathew, S. Kurmvanshi, S. Mohanty and S. K. Nayak, Influence of diisocyanate, glycidol and polyol molar ratios on the mechanical and thermal properties of glycidyl-terminated biobased polyurethanes: Mechanical and thermal properties of biobased polyurethanes, Polym. Int., 2017, 66, 1546–1554,  DOI:10.1002/pi.5412.
41. C. Xu, Z. Qu, Z. Tan, B. Nan, H. Meng, K. Wu, J. Shi, M. Lu and L. Liang, High-temperature resistance and hydrophobic polysiloxane-based polyurethane films with cross-linked structure prepared by the sol-gel process, Polym. Test., 2020, 86, 106485,  DOI:10.1016/j.polymertesting.2020.106485.
42. Y. Xu, S. Zhou, Z. Wu, X. Yang, N. Li, Z. Qin and T. Jiao, Room-temperature self-healing and recyclable polyurethane elastomers with high strength and superior robustness based on dynamic double-crosslinked structure, Chem. Eng. J., 2023, 466, 143179,  DOI:10.1016/j.cej.2023.143179.
43. L. Wang, X. Wang, T. Liu, F. Sun, S. Li, Y. Geng, B. Yao, J. Xu and J. Fu, Bio-inspired self-healing and anti-corrosion waterborne polyurethane coatings based on highly oriented graphene oxide, npj Mater. Degrad., 2023, 7, 96,  DOI:10.1038/s41529-023-00415-9.
44. F. Dong, X. Yang, L. Guo, Y. Qian, P. Sun, Z. Huang, X. Xu and H. Liu, A tough, healable, and recyclable conductive polyurethane/carbon nanotube composite, J. Colloid Interface Sci., 2023, 631, 239–248,  DOI:10.1016/j.jcis.2022.11.045.
45. D. Wu, L. Liu, Q. Ma, Q. Dong, Y. Han, L. Liu, S. Zhao, R. Zhang and M. Wang, Biomimetic supramolecular polyurethane with sliding polyrotaxane and disulfide bonds for strain sensors with wide sensing range and self-healing capability, J. Colloid Interface Sci., 2023, 630, 909–920,  DOI:10.1016/j.jcis.2022.10.058.
46. C. Li, X. Yang, Y. Wang, J. Liu and X. Zhang, Core–Shell Nanostructured Assemblies Enable Ultrarobust, Notch-Resistant and Self-Healing Materials, Adv. Funct. Mater., 2024, 2410659,  DOI:10.1002/adfm.202410659.
47. X. Duan, W. Cao, X. He, M. Wang, R. Cong, Z. Zhang, C. Ning, C. Wang, S. Zhao, Z. Li and W. Gao, Realization of dual crosslinked network robust, high toughness self-healing polyurethane elastomers for electronics applications, Chem. Eng. J., 2023, 476, 146536,  DOI:10.1016/j.cej.2023.146536.
48. H. Yang, S. Song, X. Yang, F. Fei, C. Zhang, Z. Jiang and Y. Zhang, Dynamic Adaptive Cross-linked Elastomers with Highly Robust, Recyclable, and Conductive Abilities toward Strain Sensors and Self-Sensing Actuators, ACS Sustainable Chem. Eng., 2024, 12, 3111–3120,  DOI:10.1021/acssuschemeng.3c06981.
49. H. Feng, F. Yu, Y. Guo, W. Wang, L. Xiao and Y. Liu, An exceptional strength, self-healing and recyclability polyurethane elastomers via multiple hydrogen bonds optimization strategy, Appl. Surf. Sci., 2024, 655, 159560,  DOI:10.1016/j.apsusc.2024.159560.
50. L. Xia, H. Tu, W. Zeng, X. Yang, M. Zhou, L. Li and X. Guo, A room-temperature self-healing elastomer with ultra-high strength and toughness fabricated via optimized hierarchical hydrogen-bonding interactions, J. Mater. Chem. A, 2022, 10, 4344–4354,  10.1039/D1TA08748G.
51. K. Song, W. Ye, X. Gao, H. Fang, Y. Zhang, Q. Zhang, X. Li, S. Yang, H. Wei and Y. Ding, Synergy between dynamic covalent boronic ester and boron–nitrogen coordination: strategy for self-healing polyurethane elastomers at room temperature with unprecedented mechanical properties, Mater. Horiz., 2021, 8, 216–223,  10.1039/D0MH01142H.
52. S. Li, X. Lin and S. Gong, High strength, high stiffness and toughness, defect-tolerant waterborne polyurethane with healing and antibacterial abilities, J. Polym. Sci., 2024, 62, 388–400,  DOI:10.1002/pol.20230423.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5mh00022j

---