Bioinspired poly(acrylic acid)-regulated crosslinked self-healing, quasi-solid polymer electrolytes for flexible supercapacitor applications

Abstract

The increasing demand for efficient power solutions in portable and wearable electronics has highlighted flexible supercapacitors as a critical energy storage technology, offering high-power density, exceptional cycle life, and enhanced safety. Inspired by biomimetic design strategies, which have increasingly been employed to enhance the mechanical integrity of materials, this study adopts a bio-inspired approach to develop a structurally stable architecture through synthetic modification. We introduce a self-healing quasi-solid polymer electrolyte (QSPE) that addresses the limitations of conventional gel electrolytes by leveraging poly(acrylic acid) (PAA) for enhanced electrolyte retention and self-repair capabilities. Incorporating PAA facilitates the strong hydrogen bonding networks and optimizes the crosslinking density, achieving compatibility between mechanical robustness and ion transport. In this research, synaptic-inspired coPUAA₃₃₀ has been identified as the most ideal composition. When integrated into supercapacitors soaked with tetraethylammonium tetrafluoroborate (TEABF₄) in propylene carbonate (PC) as the electrolyte, the cell with coPUAA₃₃₀ retains nearly 96% of its initial capacitance, even in relatively high water-content organic electrolytes. Furthermore, after undergoing cutting and self-repair, it retains 91% of its capacitance, demonstrating the strong self-healing ability. Additionally, the cell preserves 70% of its capacitance after 3000 charge–discharge cycles, highlighting its excellent cycling stability. Notably, flexible devices utilizing coPUAA₃₃₀ exhibit stable capacitance retention up to a bending angle of 135° and retained 78% of their capacitance upon returning to their original state. These findings establish coPUAA₃₃₀ as a high-performance, self-healing QSPE with exceptional mechanical flexibility and electrochemical reliability, reinforcing its potential for next-generation flexible supercapacitors.

---

1. Introduction

The rapid development of flexible electronics and wearable technology has driven significant interest in supercapacitors with high power density, rapid charge/discharge capability, and superior cycle life.¹ However, conventional liquid electrolytes (LEs) present challenges such as corrosiveness, low reliability, and costly packaging, limiting their use in flexible devices. Solid-state electrolytes offer a potential solution by preventing electrolyte leakage, but they suffer from poor electrode contact and low ionic conductivity (10⁻⁸ to 10⁻⁵ S cm⁻¹), hindering their practical implementation.²

To overcome these limitations, quasi-solid polymer electrolytes (QSPEs) have emerged as promising alternatives, effectively integrating the mechanical stability of solid electrolytes with the high ionic transport efficiency of liquid electrolytes. Among various QSPE materials, polyethylene glycol (PEG) is widely used due to its high dielectric constant and excellent chain flexibility.³ However, single-component QSPEs often exhibit relatively low ionic conductivity, poor thermal stability, and insufficient mechanical strength due to high crystallinity, limiting their practical applications.⁴ Alternatively, inorganic nanoparticles have been introduced to modify polymer electrolytes, effectively disrupting crystallinity and enhancing ionic transport pathways.^5,6 Nevertheless, their limited compatibility with polymer matrices can lead to phase separation and mechanical instability.⁷ To address these issues, block copolymers with multi-phase architectures^8,9 and tailored crosslinking densities have been extensively explored, offering a synergistic strategy to regulate the polymer crystallinity while improving ionic mobility, mechanical integrity, and electrolyte stability.

To develop reliable, flexible, and wearable devices, incorporating QSPEs with self-healing capability is not just beneficial but essential.^10,11 Unlike conventional polymer electrolytes, recent self-healing QSPEs typically rely on dynamic covalent chemistries (e.g., imine,¹² disulfide,^13,14 and boronic/boronate ester exchange¹⁵) or non-covalent interactions (e.g., hydrogen bonds¹⁶ and metal–ligand coordination networks^17,18). These dynamic interactions serve as sacrificial yet reversible bonds, enabling the restoration of structural integrity and functionality either spontaneously or upon mild external stimuli. Such self-healing behavior not only enhances the mechanical resilience of QSPEs but also significantly improves device durability, extends operational lifespan, and ensures stable and reliable electrochemical performance under mechanical deformation.¹⁹

Inspired by biomimetic design strategies that enhance material integrity,²⁰ we adopted a bio-inspired approach to construct a structurally stable, self-healing QSPE through synthetic modification.²¹ Drawing from the synaptic structure of neurons, which facilitates the rapid signal transmission, we hypothesized that introducing a branched architecture could similarly enhance the ionic conductivity.²² To verify this idea, poly(acrylic acid) (PAA) was introduced as a crosslinker into the coPU block copolymer backbone, effectively modulating the polymer chain dynamics.²³ The high ionic-storage capacity of PAA mimics the synaptic enhancements, improves the ion-transport efficiency, and simultaneously reinforces the mechanical stability.^24,25 Furthermore, the abundant hydrogen bonding arising from the carboxyl groups in PAA, together with the urethane-derived hydrogen bonds, endows the coPUAA materials with intrinsic and efficient self-healing capability, significantly enhancing their structural durability and resistance to mechanical fatigue (Fig. 1(a)). By preserving structural integrity and functionality even after repeated deformation, the self-healing QSPEs are believed to play a pivotal role in advancing the next-generation energy storage applications.

Fig. 1 (a) Conceptual illustration of PAA as a synaptic branching structure introduced into the coPU polymer matrix to enhance its performance as a QSPE in the energy storage application. (b) FTIR spectra tracking the polymerization reaction of coPUAA₃₃₀. (c) Summary of T_g values as a function of the PAA content. (d) Temperature-programming FTIR spectra of coPUAA₃₃₀. (e) Stress–strain curves of coPUAA_x. (f) Scratch testing results for coPUAA₃₃₀ conducted under ambient conditions. (g) Stress–strain curves of coPUAA₃₃₀ before and after healing.

2. Results and discussion

2.1. Materials characterization

The synthesis routes of coPU and coPUAA_x (where x represents various PAA contents) are depicted in Fig. S1, demonstrating that the incorporation of PAA facilitates the formation of diverse hydrogen bonds and synergistic interactions. Fig. 1(b) presents the FTIR spectra, confirming the successful synthesis of coPU, which exhibits a decrease in the intensity of the isocyanate peak at 2275 cm⁻¹.^26,27 Following a 2-h reaction period, the remaining isocyanate groups further react with the carboxylic acid groups of PAA, leading to the formation of amide linkages and the successful preparation of coPUAA_x. Fig. S2 presents the gel permeation chromatography (GPC) analysis of coPUAA_x, confirming successful polymerization through the measurement of the number-average molecular weight (M_n). The impact of varying PAA ratios on the material properties was evaluated using dynamic mechanical analysis (DMA) to measure the glass transition temperature (T_g), as shown in Fig. 1(c) and Table 1. The results indicate that at PAA contents below 330 mg, coPUAA_x (x = 100, 200) slightly increased the crosslinking density but still exhibited properties similar to pristine coPU (Fig. S3(a) and (b)). However, at 330 mg, microphase separation first emerges, accompanied by a significant T_g shift, suggesting that the PAA crosslinking density begins to influence the polymer chain mobility (Fig. S3(c)). When the PAA content exceeded 440 mg, a highly crosslinked network dominated by PAA was formed, and the pronounced microphase separation was observed, indicating that excessive PAA incorporation disrupts the local continuity of the polymer network and reduces the effective crosslinking efficiency, resulting in a high mechanical modulus but restricted chain mobility (Fig. S3(d) and (f)).

Table 1 Glass transition temperature (T_g), stress, strain, and toughness of coPU and coPUAA_x (x = 110, 220, 330, 440, 550, 660)

Sample	T g (°C)	Stress (MPa)	Strain (%)	Toughness (MJ m^−3)
coPU	−37.13347	1.69	749.85	10.92
coPUAA110	−32.99912	0.87	1343.22	6.91
coPUAA220	−28.89518	0.83	3325.36	17.75
coPUAA330	−13.26305	1.44	5057.61	36.22
coPUAA440	4.96255	1.63	3037.32	29.16
coPUAA550	11.69653	1.82	2932.79	42.85
coPUAA660	16.05044	1.75	2133.23	26.89

---

Temperature-programming FTIR experiments were performed to validate the presence of hydrogen bonding within the polymer network. As shown in Fig. 1(d), the FTIR spectra exhibit a clear peak shift of hydrogen bond-related functional groups of coPUAA₃₃₀ as the temperature increases from 25 °C to 110 °C, providing evidence of the reversible nature of hydrogen bonds in coPUAA₃₃₀. Further insights are obtained from the two-dimensional correlated FTIR spectroscopic (2D-FTIR) analysis. As shown in the synchronous 2D-FTIR spectrum (Fig. S4(a)), the cross-peaks are observed at 1545–1639 cm⁻¹, indicating a strong correlation between amide C=O stretching and NH bending vibrations, suggesting that these hydrogen bonds form and break synchronously as temperature increases. Additionally, the asynchronous 2D-FTIR spectra (Fig. S4(b)) reveal a negative cross-peak at 1639–1695 cm⁻¹, suggesting a sequential structural rearrangement between carbonyl hydrogen bonding and free carbonyl groups. This result implies that as temperature increases, the hydrogen-bonded carbonyls gradually dissociate before free carbonyls are fully disrupted, highlighting the hierarchical nature of hydrogen bond breakage in coPUAA_x.

Uniaxial tensile tests at room temperature (displacement rate: 100 mm min⁻¹) further quantify their mechanical properties (Fig. 1(e) and Table 1). The pristine coPU exhibits the high tensile strength (1.7 ± 0.2 MPa) but limited strain (800 ± 40%). Incorporating PAA into the polymer matrix, as demonstrated in coPUAA₁₁₀, effectively reduces the initial tensile strength, as observed in coPUAA₁₁₀ (0.8 ± 0.1 MPa), but markedly increases the strain capability (1400 ± 50%). This behavior suggests that PAA disrupts the tightly packed PU structure, resulting in enhanced flexibility. At an optimal PAA content of 330 mg, coPUAA₃₃₀ achieves the remarkable tensile strain (5000 ± 100%) with moderate tensile strength (1.4 ± 0.2 MPa). Beyond this ratio, coPUAA_x (x = 440, 550, 660) shows increased tensile strength but reduced strain due to a rigid PAA-dominated crosslinked network.

The presence of multiple hydrogen bonds in coPUAA_x imparts the self-healing capability to the material. To confirm this behavior, optical microscopy was employed to observe deliberately notched coPUAA_x films at room temperature (20 °C). The results indicate rapid self-healing in coPUAA₃₃₀, achieving complete healing within 1 h (Fig. 1(f) and S5). Further confirmation is provided by the stress–strain curves of healed films (Fig. 1(g)), which quantitatively demonstrate the self-healing efficiency of coPUAA₃₃₀. The results show that coPUAA₃₃₀ restores its 96% tensile stress of the original value, highlighting the critical role of PAA in forming the intermolecular hydrogen bonds and dynamic networks that drive the self-healing process. In addition, the contact angle measurements using 1 M TEABF₄/PC electrolytes (Fig. S6) reveal that coPUAA₃₃₀ exhibits superior electrolyte affinity due to its abundant carboxyl groups, which promote ionic conduction. The complementary water contact angle tests (Fig. S7) further confirm that the optimized polymer structure in coPUAA₃₃₀ facilitates the efficient ion transport, further reinforcing its promising potential for QSPE applications.

2.2. Electrochemical analysis

Ionic conductivity is a critical performance metric for polymer electrolytes in energy storage devices. However, variations in film thickness can significantly impact the conductivity measurements. To address this, a controlled hot-pressing method was employed to ensure consistent processing conditions, leveraging the free-standing nature of coPUAA to fabricate thin films (detailed in Fig. S8). Under identical conditions, coPUAA₃₃₀ exhibits the thinnest film thickness (70 µm), achieving a balance between optimal flexibility and mechanical stability (shown in Table S1).

The ionic conductivity measurements conducted in coin cells (Fig. S9 and Table S1) reveal that the unmodified coPU exhibits a notably low ionic conductivity (0.0513 mS cm⁻¹), underscoring its limited suitability for electrochemical applications. In contrast, the introduction of PAA as a branching structure within the coPU matrix significantly enhances ionic transport, verifying our design idea. Specifically, coPUAA₃₃₀ achieves a maximum ionic conductivity of 0.485 mS cm⁻¹, demonstrating that PAA incorporation not only facilitates ion transport by providing abundant –COO⁻ groups but also improves material flexibility.²⁸ However, a further increase in the PAA content, as in coPUAA_x (x = 440, 550, 660), leads to a marked decrease in the ionic conductivity, attributable to the elevated crystallinity that impedes ion migration.²⁹ This trend is further supported by the contact angle measurements using the 1 M TEABF₄/PC electrolyte (Fig. S6), where coPUAA₃₃₀ exhibits the lowest contact angle of 34.56°, indicating superior electrolyte affinity and a well-regulated polymer structure.^30,31

The electrochemical properties of the synthesized coPU and coPUAA_x polymer films were evaluated using a symmetric supercapacitor with two identical activated carbon (AC) electrodes to assess their suitability for high-power applications under various operating conditions. Cyclic voltammograms (CVs) of supercapacitors using the QSPE are shown in Fig. 2(a); both the original coPU and the modified coPUAA exhibit ideally capacitive behavior without faradaic reactions within a voltage window of 0–2.5 V.³² Notably, the introduction of PAA improves the reversibility of the capacitive response, with coPUAA₃₃₀ demonstrating the most rectangular CV profile, attributable to its optimized ionic conductivity and moderate crystallinity. In contrast, with the further increase in the crystallization of coPUAA₄₄₀, the PAA-dominated crosslinked network restricts the ion transport, resulting in high iR responses and the emergence of distorted, leaf-like CVs.^33,34 Galvanostatic charge–discharge (GCD) measurements (Fig. 2(b)) further confirm the enhanced ionic transport in the coPUAA_x system. Specifically, the cell using coPUAA₃₃₀ exhibits the smallest iR-drop and achieves a remarkable specific capacitance (27.7 F g⁻¹), nearly 10 times that of the cell using pristine coPU (2.9 F g⁻¹), as shown in Table S2. This enhancement can be attributed to the branched polymer architecture and abundant carboxyl groups introduced by PAA, which facilitate the ion transport within the QSPE. Conversely, the excessive crosslinking in coPUAA₆₆₀ leads to high internal resistance, as reflected by an iR-drop exceeding 2 V, rendering it unsuitable for QSPE applications. In contrast, coPUAA₃₃₀ effectively reduces the internal resistance via maximizing the ionic transport, meeting the essential criteria for the application of high-performance QSPE-based supercapacitors.

Fig. 2 (a) CVs measured at 5 mV s⁻¹, (b) GCD curves obtained at 0.5 A g⁻¹, and (c) EIS analysis for the cells using coPU and coPUAA_x with various PAA ratios. (d) CVs measured at 5–25 mV s⁻¹ and (e) GCD curves obtained at 0.5–1 A g⁻¹ for the coPUAA₃₃₀ cell. (f) The specific capacitance vs. current density comparison for AC-based cells using coPUAA₃₃₀ and coPU. (g) Cycle performance of the coPUAA₃₃₀ cell at 0.5 A g⁻¹. (h) The Ragone plot comparison for supercapacitors with various types of polymer electrolytes.^38–49 All values are normalized to the total mass of active materials.

Building on these observations, electrochemical impedance spectroscopy (EIS) was employed to further investigate the impact of the PAA content on the impedance characteristics of the cells using coPUAA_x-based QSPEs in Fig. 2(c). To establish a reliable interpretation, pristine coPU and coPUAA₆₆₀ were chosen as typical samples to define the equivalent circuit, given their high impedance and distinct semicircle features. Based on this framework, the EIS responses can be divided into five characteristic parts. The real part resistance at the high-frequency end indicates the intrinsic resistance of the QSPE (i.e., R_S), which shows similar information to that in Fig. S9. The two small semicircles in the high- and middle-frequency regions are attributable to the contact impedance at the AC/current collector interface (R_if and CPE_if in parallel) and interparticle impedance among AC particles (R_P and CPE_P in parallel). The third semicircle in the relatively low-frequency region could be linked to the interfacial resistance when ions transport through the AC/electrolyte interface (R_PD and CPE_PD in parallel). The ion-diffusion impedance within porous particles of supercapacitors using liquid electrolytes was commonly fitted by the transmission line model (TLM) proposed by Levie, denoted as M_a. Finally, at frequencies close to the low-frequency end, a capacitor-like behavior is modeled by a constant phase element, as shown in Fig. S10.³⁵ As summarized in Table S3, the pristine coPU and coPUAA₆₆₀ exhibit the highest R_S, attributable to two reasons in increasing the intrinsic resistance of QSPEs. For the pristine coPU, the absence of crosslinking within the coPU limits the formation of ion channels, reducing the ionic conductivity. For the coPUAA₆₆₀, on the other hand, the excessive crosslinking within the polymer reduces the electrolyte affinity, decreasing the moving rate of ions.³⁰ Moreover, both materials display prominent semicircles in the low-frequency region, indicating elevated R_PD values that are unfavorable for supercapacitor applications.

In contrast, the incorporation of an optimized PAA content in coPUAA₃₃₀ significantly reduces R_S, R_P and R_PD, revealing the impact of a QSPE polymer network design, promoting not only the intimate interfacial contact with the AC matrix but also the ion movement. With the optimized branching structure, coPUAA₃₃₀ minimizes spatial hindrance, enhances material flexibility, and facilitates efficient ionic transport channels, effectively achieving the lowest equivalent series resistance (ESR) among all coPUAA samples.

CVs measured at various scan rates were performed for the coPUAA₃₃₀ and coPU cells (Fig. 2(d) and S11). The cell with the pristine coPU displays distorted responses, indicating the suboptimal ionic transport. In contrast, the supercapacitor using coPUAA₃₃₀ exhibits highly symmetric CV profiles even at high scan rates. These findings highlight the significant role of the PAA-induced synaptic architecture in facilitating the ionic transport pathways and capacitive properties. Besides, the GCD tests at various current densities further emphasize the stark contrast. As shown in Fig. 2(e), the cell using coPUAA₃₃₀ shows nearly ideal charge–discharge curves with a minimal iR-drop, even at 1 A g⁻¹, whereas the one with coPU exhibits the pronounced asymmetric charge–discharge behavior and significant voltage losses (Fig. S12). These phenomena confirm that the PAA incorporation enables coPUAA₃₃₀ to enhance its ionic conductivity, rendering the cell to achieve an ideal electric double-layer charge storage mechanism suitable for high-power supercapacitor applications. The specific capacitance comparison (Fig. 2(f)) demonstrates that the specific capacitance of ACs from the coPUAA₃₃₀ cell is 2.5 times that of the coPU-based cell at low current densities, while this difference increases to nearly 10-fold at 1 A g⁻¹, highlighting its superior rate capability under high current conditions. The long-term cycling test shown in Fig. 2(g) reinforces the durability of coPUAA₃₃₀, retaining approximately 70% of its initial capacitance even after 3000 cycles. The gradual decline likely arises from the limited compatibility between the PVDF binder and the coPUAA₃₃₀ electrolyte, which probably disrupts the ionic pathways and increases the interfacial resistance during the cycling test.

To further assess the performance of coPUAA₃₃₀, we compared the specific energy and power of the coPUAA₃₃₀ cell with other high-efficiency supercapacitor systems using a Ragone plot, as shown in Fig. 2(h). The energy densities at different current densities were obtained using the following formula:^36,37

where C is the specific cell capacitance based on the total mass of active materials coated on both electrode and V is the voltage window (2.5 V). The corresponding power density was calculated according to the formula:where E is from eqn (1) and Δt is the galvanostatic discharge time. From eqn (1) and (2), the supercapacitor employing the coPUAA₃₃₀ QSPE achieves a specific energy of approximately 52 Wh kg⁻¹ and a specific power of 1250 W kg⁻¹. On the same calculation basis, these values significantly surpass other QSPE cells reported in the literature. This superior performance is attributable to the optimized crosslinked network, which enhances charge mobility and improves electrolyte affinity. These findings highlight the immense potential of coPUAA₃₃₀ in revolutionizing the high-performance flexible supercapacitors and solid-state energy storage devices.

Material characterization and electrochemical testing revealed several limitations of pristine coPU as a QSPE for supercapacitors. Despite its abundance of ether groups, the insufficient flexibility and low ionic conductivity hinder its practical applicability. The incorporation of PAA as a crosslinker fundamentally alters the coPU structure. As a structural skeleton, the long-chain architecture of PAA enhanced the flexibility, and its abundant carboxyl groups serve as synaptic-like ionic transport sites, significantly improving the ionic conductivity. Additionally, PAA introduces dynamic hydrogen-bonding interactions, conferring effective self-healing capabilities.

Furthermore, our findings demonstrate that the mechanical properties and electrochemical performance of coPUAA_x as a QSPE are closely regulated by the degree of PAA incorporation. At low PAA ratios, the limited number of synaptic sites results in a loosely crosslinked network, which moderately enhances the ionic transport through carboxyl groups but compromises mechanical robustness and overall electrochemical performance. Conversely, an excessive PAA content leads to an over-crosslinked network dominated by PAA, where the carboxyl groups tend to form intramolecular hydrogen bonds, increasing the crystallinity of QSPEs and impeding the ion mobility. As shown in Fig. 1(a), the optimal PAA content, precisely tuned to align with the T_g point of coPUAA_x, promotes a well-balanced synaptic structure that optimizes the chain dynamics, facilitates the ionic conduction, and enables the self-healing behavior. This synergistic effect, driven by the PAA-based synaptic architecture, establishes coPUAA_x as a promising QSPE for flexible supercapacitors.

2.3. Self-healing and flexible energy storage applications

A self-healing QSPE is indispensable for improving the safety and environmental tolerance of energy storage devices. Benefiting from the PAA synaptic branching structure, coPUAA₃₃₀ exhibits remarkable self-healing capabilities, fully restoring mechanical integrity after being severed and rejoined. To investigate its self-healing property in energy storage systems, two complementary evaluation strategies were employed within a supercapacitor system, as shown in Fig. 3(a). First, the physical cutting test: after self-healing for 1 h, a slight increase in the impedance of coPUAA₃₃₀ is visible (Fig. 3(b)). This phenomenon can be attributed to the localized defects and inhomogeneous network reconstruction during the healing process, which may introduce additional ionic transport barriers.⁵⁰ Despite this minor increase in impedance, the ion mobility is largely preserved, indicating stable electrochemical interfacial properties of coPUAA₃₃₀ following self-healing. Consistent results are observed in the CVs shown in Fig. S13, which maintain ideal capacitive behavior with 97.9% capacitance retention, showcasing how the self-healing mechanism retains the rapid ionic conduction capabilities of coPUAA₃₃₀ without compromise. Similarly, the GCD curves in Fig. 3(c) display only a slight increase in the iR-drop following the self-healing process, yet the material consistently retains 91.3% of its initial capacitance after healing, demonstrating the strong physical-healing capability of coPUAA₃₃₀ in maintaining reversible electrochemical energy storage performance even after mechanical damage.

Fig. 3 (a) Schematic illustration of the self-healing mechanism of coPUAA₃₃₀, (b) negative Nyquist plots, (c) GCD curves obtained at 0.5 A g⁻¹ for the cells using coPUAA330 before/after self-healing (cut-and-healed sample), and (d) GCD curves obtained at 0.1 A g⁻¹, (e) iR drop on the charge curves (initial voltage) and specific capacitance across charge–discharge cycles at 0.1 A g⁻¹ for the cell assembled by adhering two separate coPUAA₃₃₀ films.

Second, an alternative evaluation method was employed to investigate the reconstruction of ionic conduction pathways. In this ion channel reconstruction test, two identical 70 µm coPUAA₃₃₀ membranes were aligned and assembled into a full EDLC configuration. As shown in Fig. 3(d), the bilayer structure successfully exhibited ideal capacitive behavior at 0.1 A g⁻¹, confirming the feasibility of ionic transport across the healed interface. Notably, the GCD curve in the first cycle displayed a significantly higher initial voltage compared to subsequent cycles, as further illustrated in Fig. 3(e). This elevated onset voltage is indicative of a transient increase in internal resistance, likely due to incomplete ionic and interfacial recovery at the newly formed junction. The progressive decrease and stabilization of the initial voltage in later cycles suggest the gradual reformation of continuous ion transport pathways and restoration of interfacial integrity, thereby highlighting the effective ionic pathway reconstruction capability of coPUAA₃₃₀. Although the first-cycle GCD profile exhibited a notably high iR drop (from 0 to 0.86 V) on the charge curve, the discharge cell capacitance was comparable to that of subsequent cycles. In addition, Fig. S14 shows the direct current ESR values derived from the slope of the iR-drop against the applied current during the GCD discharge process. The bilayer coPUAA₃₃₀ system showed only 43.5% higher resistance than twice that of the monolayer configuration, indicating that the reconstructed interface imposed minimal additional transport resistance. This is reasonably attributed to the highly interconnected ion conduction pathways formed by the dynamic hydrogen bonding network in coPUAA₃₃₀, which facilitates efficient ionic mobility across the self-healed junction. This observation suggests that, despite temporary resistance at the interface, the reconstructed ion transport network is sufficiently effective to support stable energy storage performance from the onset of operation.

In addition to mechanical resilience, coPUAA₃₃₀ addresses challenges posed with water in organic electrolyte systems, where moisture can trigger side reactions above 1 V,⁵¹ increasing the risk of explosions,^52,53 and causing performance degradation in low-temperature environments.⁵⁴ Given the unavoidable presence of trace water, materials capable of stabilizing water molecules are crucial for retaining electrochemical stability in high water-content environments. According to the literature, PAA, due to its abundant hydrogen bonding sites, is a highly hydrophilic polymer.^55,56 Leveraging the hydrophilic nature of PAA, coPUAA₃₃₀ effectively immobilizes water through extensive hydrogen bonding, mitigating moisture-induced performance losses (as shown in Fig. 4(a)). To evaluate this capability of coPUAA₃₃₀, experiments were conducted using TEABF₄/PC electrolytes stored in a glove box under ambient conditions for a long period, during which the water content reached 150 ppm (measured by Karl Fischer titration). Fig. 4(b) illustrates that coPUAA₃₃₀, characterized by its high ionic conductivity and abundant hydrogen bonding sites, effectively sustains the internal resistance in the relatively high water-content electrolytes. Similar results are observed in the GCD curves in Fig. 4(d), where the negligible variation in iR-drop between systems with 8 ppm and 150 ppm water, preserving the electrochemical performance. The CV & GCD measurements (Fig. 4(c) and (d)) further confirm that the cell using coPUAA₃₃₀-based QSPE maintains reversible capacitive behavior even when soaked in high water-content electrolytes. The GCD curves under high water content retain 96.8% of their original capacitance, as shown in Table S4. The phenomenal self-healing ability of coPUAA₃₃₀ can be attributed to its PAA synaptic branching, which promotes the formation of abundant hydrogen bonds. Importantly, the contact angle measurements (refer to Fig. S6) further validate the strong hydrophilicity of coPUAA₃₃₀, with a measured contact angle of approximately 40.44°, indicative of strong water affinity. This suggests that water molecules are more readily attracted to and retained by the polymer chains, confirming the strong interactions between the polymer and water. These interactions not only stabilize the electrolyte composition but also enhance the water molecule immobilization, contributing to the durability and environmental tolerance of the QSPE in electrochemical energy storage devices.

Fig. 4 (a) Schematic illustration of water molecule immobilization by coPUAA₃₃₀, (b) negative Nyquist plots, (c) CVs measured at 5 mV s⁻¹, and (d) GCD curves obtained at 0.5 A g⁻¹ for the cells using coPUAA₃₃₀ soaked with 1 M TEABF₄/PC containing 8 ppm or 150 ppm water content.

To assess the feasibility of coPUAA₃₃₀ in flexible supercapacitor applications, a symmetric supercapacitor was assembled with coPUAA₃₃₀ as the QSPE and two identical AC-coated copper foil electrodes, as illustrated in Fig. 5(a). The annealed Cu foil was employed as the substrate owing to its favorable mechanical flexibility and packaging compatibility, without compromising the electrochemical performance. The CVs in Fig. 5(b) display a near-rectangular shape, indicating typical capacitive behavior, while the GCD curves exhibit consistent triangular shape across current densities from 0.5 A g⁻¹ to 1 A g⁻¹. Although a higher iR-drop is observed compared to the coin-cell system, likely due to the absence of internal cell pressure, coPUAA₃₃₀ maintains efficient ion transport and electrode contact (Fig. 5(c)). Given the demands of flexible devices, the mechanical resilience of coPUAA330 was assessed under various bending conditions. As shown in Fig. S15, the charge–discharge profiles remain stable at a current density of 0.5 A g⁻¹ across bending angles from 0° to 180°. The corresponding specific capacitance values, shown in Fig. 5(d), remain stable up to 135°, with 78% capacitance retention when the device returns to 0°. This mechanical durability is attributed to the dynamic hydrogen-bonding network within coPUAA₃₃₀, which supports repeated deformation while preserving the structural integrity. Moreover, its inherent self-healing capability mitigates the microstructural damage that may accumulate during flexing, ensuring sustained electrochemical performance over the extended use. These results demonstrate that coPUAA₃₃₀ combines the mechanical adaptability, self-repairability, and electrochemical stability, which confirm this polymer as a promising QSPE for the next-generation flexible supercapacitors capable of withstanding mechanical fatigue without compromising energy storage performance.

Fig. 5 (a) Schematic representation of the symmetric flexible supercapacitor configuration, (b) CVs measured at various scan rates ranging from 5 mV s⁻¹ to 50 mV s⁻¹, (c) GCD curves at current densities from 0.5 A g⁻¹ to 1 A g⁻¹ of a flexible supercapacitor with coPUAA₃₃₀, and (d) the specific cell capacitance of a flexible supercapacitor with coPUAA₃₃₀ at various fold angles.

3. Conclusions

In this study, a bioinspired QSPE was developed by introducing PAA as a synaptic-like crosslinker within a coPU matrix. The incorporation of PAA plays a critical role in regulating the crosslinked network architecture, directly influencing the balance between mechanical flexibility, ionic conductivity, and self-healing capability. Systematic evaluation revealed that the PAA content critically governs the formation of ion-conductive pathways and dynamic hydrogen-bonding networks. Among the tested compositions, coPUAA₃₃₀ exhibits the optimal PAA ratio, enabling a well-regulated crosslinked structure that maximizes free volume for ion transport while maintaining sufficient mechanical integrity. The coPUAA₃₃₀ formulation exhibited good ionic conductivity (0.485 mS cm⁻¹), high tensile strain (4700%), and robust self-healing properties through an optimized network of carboxyl-rich transport sites and reversible hydrogen bonds.

These synergistic properties enable coPUAA₃₃₀ to maintain outstanding electrochemical reliability under mechanical stress, with 78% capacitance retention after repeated bending, 91.3% retention post self-healing, and 70% retention after 3000 charge–discharge cycles. Furthermore, coPUAA₃₃₀ demonstrates exceptional environmental tolerance, preserving performance even when soaked in relatively high water-content electrolytes. These results highlight the pivotal role of PAA-regulated crosslinking in tuning QSPE properties and establish coPUAA₃₃₀ as a highly promising candidate for the next-generation flexible supercapacitors, offering a synergistic combination of mechanical resilience, self-healing capability, and electrochemical reliability essential for wearable and portable energy storage applications.

Author contributions

Yun Ku: conceptualization, methodology, validation, visualization, formal analysis, investigation, data curation, writing – original draft preparation, writing – review & editing. Yi-An Chen: conceptualization, methodology, validation, visualization, formal analysis, investigation, data curation, writing – original draft preparation, writing – review & editing. Hung-Yi Huang: conceptualization, methodology, validation, visualization, formal analysis, investigation, data curation, writing – original draft preparation, writing – review & editing. Rou-Han Lai: visualization, investigation, data curation. Yi-Heng Tu: visualization, writing – review & editing. Ho-Hsiu Chou: conceptualization, methodology, validation, visualization, formal analysis, writing – review & editing, supervision, project administration. Chi-Chang Hu: conceptualization, methodology, validation, visualization, formal analysis, writing – review & editing, supervision, project administration.

Conflicts of interest

There are no conflicts of interest to declare.

Data availability

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

Supplementary information (SI): a schematic illustration of the coPUAA_x synthesis process, GPC spectra, dynamic mechanical analysis (DMA), two-dimensional FTIR (2D-FTIR) analysis, optical microscopy (OM) images, water contact angle measurements, electrochemical impedance spectroscopy (EIS) fitting results, a schematic representation of the coin-cell configuration, calculations of specific capacitance and iR-drop, and a comparative evaluation of electrochemical performance among coPUAA_x. See DOI: https://doi.org/10.1039/d5ta06583f.

Acknowledgements

The financial support of this work by the Ministry of Science and Technology (MOST) under contract no. MOST 109-2221-E-007-038-MY3 and National Science and Technology Council (NSTC) of Taiwan under contract no. NSTC 113-2923-E-007-007-MY3, NSTC 113-2823-8-007-003, NSTC 112-2622-E-007-032, NSTC 112-2223-E-007-006-MY3, NSTC 112-2221-E-007-081, and NSTC 112-2221-E-007-021-MY3 is gratefully acknowledged.

References

1. W. Raza, F. Ali, N. Raza, Y. Luo, K.-H. Kim, J. Yang, S. Kumar, A. Mehmood and E. E. Kwon, Nano Energy, 2018, 52, 441–473.
2. R. Chen, Q. Li, X. Yu, L. Chen and H. Li, Chem. Rev., 2019, 120, 6820–6877.
3. M. Sadiq, M. A. Khan, M. M. Hasan Raza, S. M. Aalam, M. Zulfequar and J. Ali, ACS Omega, 2022, 7, 40116–40131.
4. X. Lu, Y. Wang, X. Xu, B. Yan, T. Wu and L. Lu, Adv. Energy Mater., 2023, 13, 2301746.
5. F. Croce, G. B. Appetecchi, L. Persi and B. Scrosati, Nature, 1998, 394, 456–458.
6. T. J. Singh and S. V. Bhat, J. Power Sources, 2004, 129, 280–287.
7. X.-L. Wang, A. Mei, M. Li, Y.-H. Lin and C.-W. Nan, J. Appl. Phys., 2007, 102, 054907.
8. T. Feng, X. Li, G. Li, X. Luo, X. Liu, K. Zhang, X. Jia, Y. Wang, W. Sun and K. Wang, J. Energy Storage, 2024, 102, 114131.
9. A. J. Butzelaar, P. Röring, M. Hoffmann, J. Atik, E. Paillard, M. Wilhelm, M. Winter, G. Brunklaus and P. Theato, Macromolecules, 2021, 54, 11101–11112.
10. R.-H. Lai, Y.-A. Chen, C.-Y. Chou, H.-Y. Huang, W. Mongkonkan, C.-A. Chiu, Y.-H. Chen, M.-H. Yu, C.-C. Hu and S. Jungsuttiwong, J. Mater. Chem. A, 2025, 13, 14588–14600.
11. Y.-A. Chen, R.-H. Lai, W.-C. Lin, H.-Y. Huang, S.-J. Chen, C.-M. Yeh, H.-L. Huang, M. M. Elsenety, C.-C. Hu and C.-H. Yu, ACS Appl. Polym. Mater., 2024, 6, 6976–6987.
12. K. Deng, S. Zhou, Z. Xu, M. Xiao and Y. Meng, Chem. Eng. J., 2022, 428, 131224.
13. H. Liu, S. Luo, Y. Yang, X. Zhao, G. Huang, X. Jin, T. Zhong, M. Guan, J. Liu and Y. Li, Energy Storage Mater., 2025, 78, 104250.
14. C.-M. Yeh, C.-H. Lin, T.-Y. Han, Y.-T. Xiao, Y.-A. Chen and H.-H. Chou, J. Mater. Chem. A, 2021, 9, 6109–6116.
15. J. Zheng, X. Y. Oh, E. Ye, W. H. Chooi, Q. Zhu, X. J. Loh and Z. Li, Mater. Chem. Front., 2023, 7, 381–404.
16. Y. Zhou, Y. Li, X. Liu, J. Lv, Y. Su and L. Weng, Energy Storage Mater., 2025, 104265.
17. J. Li, X. Zhu, Z. Xu, H. Ren, S. Muhammad, H. Yang, B. Luo, M. U. Javed, C. Yu and Y. Zhang, Adv. Funct. Mater., 2025, e10177.
18. T.-Y. Han, C.-H. Lin, Y.-S. Lin, C.-M. Yeh, Y.-A. Chen, H.-Y. Li, Y.-T. Xiao, J.-W. Chang, A.-C. Su and U.-S. Jeng, Chem. Eng. J., 2022, 438, 135592.
19. L. Zhang, P. Zhang, C. Chang, W. Guo, Z. H. Guo and X. Pu, ACS Appl. Mater. Interfaces, 2021, 13, 46794–46802.
20. H. Dou, M. Xu, Y. Zheng, Z. Li, G. Wen, Z. Zhang, L. Yang, Q. Ma, A. Yu and D. Luo, Adv. Mater., 2022, 34, 2110585.
21. A. Li, X. Liao, H. Zhang, L. Shi, P. Wang, Q. Cheng, J. Borovilas, Z. Li, W. Huang and Z. Fu, Adv. Mater., 2020, 32, 1905517.
22. H. Tan, Q. Tao, I. Pande, S. Majumdar, F. Liu, Y. Zhou, P. O. Persson, J. Rosen and S. van Dijken, Nat. Commun., 2020, 11, 1369.
23. J.-A. Wang, Y.-T. Lu, S.-C. Lin, Y.-S. Wang, C.-C. M. Ma and C.-C. Hu, ACS Appl. Mater. Interfaces, 2018, 10, 17871–17882.
24. W. Li and Y. Cui, Sol. Energy Mater. Sol. Cells, 2023, 250, 112071.
25. Z. Song, X. Liu, J. Ding, J. Liu, X. Han, Y. Deng, C. Zhong and W. Hu, ACS Appl. Mater. Interfaces, 2022, 14, 49801–49810.
26. D. P. Queiroz, M. N. de Pinho and C. Dias, Macromolecules, 2003, 36, 4195–4200.
27. Y.-A. Chen, S.-J. Chen, L.-Y. Lee, R.-H. Lai, C.-M. Yeh, C.-A. Chiu, J.-Y. Lai, Y.-C. Lai and H.-H. Chou, Nano Energy, 2023, 117, 108882.
28. J. Kim, J. Choi, K. Park, S. Kim, K. W. Nam, K. Char and J. W. Choi, Adv. Energy Mater., 2022, 12, 2103718.
29. W. Zhu, S. Wang, X. Zhang, A. Wang, C. Wu and F. Hao, Small, 2022, 18, 2105783.
30. H.-Y. Huang, Y.-H. Tu, Y.-H. Yang, Y.-A. Chen, W.-L. Lee, M.-F. Wu, H.-H. Chou and C.-C. Hu, J. Mater. Chem. A, 2024, 12, 4312–4324.
31. H.-Y. Huang, Y.-H. Tu, Y.-H. Yang, Y.-T. Lu and C.-C. Hu, Chem. Eng. J., 2023, 457, 141373.
32. Y. Ku, H.-Y. Ku, A.-L. Huang, H.-Y. Huang, W.-Y. Chang, J.-C. Liang and C.-C. Hu, J. Energy Storage, 2025, 115, 115863.
33. C.-M. Chen, Q. Zhang, M.-G. Yang, C.-H. Huang, Y.-G. Yang and M.-Z. Wang, Carbon, 2012, 50, 3572–3584.
34. M.-H. Tsai, S.-J. Chao, D. C.-I. Wang, I.-H. Lin, L.-C. Chung, P.-I. Liu, L.-C. Hua, H.-Y. Huang and C.-C. Hu, Chem. Eng. J., 2025, 521, 166833.
35. Z. Yang, H. Jiang, X. Li, X. Liang, J. Wei, Z. Xie, B. Tang, Y. Zhang and Z. Zhou, Adv. Funct. Mater., 2024, 34, 2407713.
36. H. Yang, S. Kannappan, A. S. Pandian, J.-H. Jang, Y. S. Lee and W. Lu, Nanotechnology, 2017, 28, 445401.
37. Y. Zhu, S. Murali, M. D. Stoller, K. J. Ganesh, W. Cai, P. J. Ferreira, A. Pirkle, R. M. Wallace, K. A. Cychosz and M. Thommes, science, 2011, 332, 1537–1541.
38. T. Mitravinda, K. Nanaji, S. Anandan, A. Jyothirmayi, V. S. K. Chakravadhanula, C. S. Sharma and T. N. Rao, J. Electrochem. Soc., 2018, 165, A3369.
39. S. Senthilkumar and R. K. Selvan, ChemElectroChem, 2015, 2, 1111–1116.
40. M. U. Rani, K. Nanaji, T. N. Rao and A. S. Deshpande, J. Power Sources, 2020, 471, 228387.
41. M. Y. Bhat, N. Yadav and S. Hashmi, Electrochim. Acta, 2019, 304, 94–108.
42. N. Yadav, M. K. Singh, N. Yadav and S. Hashmi, J. Power Sources, 2018, 402, 133–146.
43. P. Wang, Y. Zhang, H. Jiang, X. Dong and C. Meng, Chem. Eng. J., 2022, 427, 131548.
44. Z. Pan, M. Liu, J. Yang, Y. Qiu, W. Li, Y. Xu, X. Zhang and Y. Zhang, Adv. Funct. Mater., 2017, 27, 1701122.
45. Y. Wang, X. Wang, X. Li, X. Li, Y. Liu, Y. Bai, H. Xiao and G. Yuan, Adv. Funct. Mater., 2021, 31, 2008185.
46. A. K. Mohammed, A. Pandikassala, P. P. Sánchez, S. A. Gaber, S. Canossa, M. Kurian, G. Xavier, Y. He, F. Gándara and S. Kurungot, Chem. Eng. J., 2024, 496, 153589.
47. Y. Wang, N. Chen, Y. Liu, X. Zhou, B. Pu, Y. Qing, M. Zhang, X. Jiang, J. Huang and Q. Tang, Chem. Eng. J., 2022, 450, 138398.
48. C. Xu, Z. Li, C. Yang, P. Zou, B. Xie, Z. Lin, Z. Zhang, B. Li, F. Kang and C.-P. Wong, Adv. Mater., 2016, 28, 4105–4110.
49. Y. Zeng, Y. Han, Y. Zhao, Y. Zeng, M. Yu, Y. Liu, H. Tang, Y. Tong and X. Lu, Adv. Energy Mater., 2015, 5, 1402176.
50. Y. Zhou, L. Li, Z. Han, Q. Li, J. He and Q. Wang, Chem. Rev., 2022, 123, 558–612.
51. H. Kurig, M. Vestli, A. Jänes and E. Lust, Electrochem. Solid-State Lett., 2011, 14, A120.
52. Q. Xu, Y. Zhu, T. Xie, C. Shi and N. Zhang, ChemCatChem, 2021, 13, 592–602.
53. S. Chen, J. Liao, Z. Zhou, S. Yang, Q. Gao, X. Cai, F. Peng, Y. Fang and S. Zhang, Appl. Catal., B, 2021, 291, 120139.
54. A. Tatlisu, Z. Huang and R. Chen, ChemSusChem, 2018, 11, 3899–3904.
55. H. Liao, F. Zhou, Z. Zhang and J. Yang, Chem. Eng. J., 2019, 357, 428–434.
56. J. H. Park, S. Jung, Y. H. Song, N. R. Aluru, T. Kim, S. B. Lee, U. H. Choi and J. Lee, ACS Appl. Energy Mater., 2020, 3, 10119–10130.

---

Footnote

† These authors contributed equally to this work.

---