Engineering metal–organic framework towards suppressed leakage current in polymer nanocomposites

Abstract

The development of high-performance power electronics is constrained by the lack of polymer dielectrics capable of operating under extreme conditions. Conventional dielectrics face significant challenges at high temperatures and electric fields due to increased leakage currents and dielectric loss. In this work, we introduce zeolitic imidazolate framework (ZIF-11) as a filler for polyetherimide (PEI), showcasing its dual functionality in suppressing both ionic and electronic conduction. The strong electrostatic interactions between ZIF-11 and the PEI matrix reduce free volume, inhibiting ionic species' mobility, while the high electron affinity of ZIF-11 introduces multiple traps for electronic charge carriers, significantly lowering electronic conduction. Leveraging these synergistic effects, the composite achieves an order-of-magnitude reduction in leakage current with ≤1 wt% ZIF-11. It maintains >80% efficiency at 625 MV m⁻¹ and 150 °C, achieving an energy density of 5.44 J cm⁻³—a 1.38-fold improvement over unfilled PEI and surpassing current dielectric polymers and polymer nanocomposites. This scalable strategy demonstrates the transformative potential of engineering MOF-based fillers in designing advanced polymer dielectrics for next-generation power electronics.

---

1. Introduction

The worldwide goal of peak carbon emissions places increasing pressure on renewable energy, recyclable materials, and highly efficient energy use as conventional energy sources are gradually depleted.^1–5 Polymer film capacitors are extensively utilized in wearable electronics, hybrid electric vehicles, and pulse power systems because of their large power density, high breakdown strength (E_b), and fast charge–discharge responses with mechanical flexibility.^6–9 Dielectric capacitors are always exposed to harsh environments and operate at temperatures higher than 140 °C in practical applications.^10–12 Unfortunately, one of the most advanced commercial dielectrics for film capacitors, biaxially oriented polypropylene (BOPP) film, typically operates only at a maximum temperature of less than 100 °C when exposed to an external electric field.¹³ This is significantly lower than what is required for the high temperatures of the aforementioned electric fields. Consequently, the development of dielectric polymers with exceptional stability and dependability at high temperatures is crucial.

Polymers with high glass transition temperatures (T_g) (like polyimide [PI], polyetherimide [PEI], fluorene polyester [FPE], divinyltetramethyldisiloxane-bis(benzocyclobutene) [BCB], polycarbonate [PC], etc.) are prospective dielectric materials for use in high-temperature settings.^14,15 Among them, PEI is a member of the PI class, whose T_g can reach up to 260 °C and outstanding thermal stability.¹⁶ The aromatic imide unit in PEI's molecular chain increases the polymer's mechanical strength and heat resistance, while the ether unit provides mechanical flexibility and fluidity. Moreover, PEI has high E_b, high T_g, and exceptional mechanical properties due to the polar groups phthalimide's high dipole moments. PEI is widely regarded as an ideal linear dielectric for developing high-temperature dielectric capacitors. However, despite its excellent thermal resistance, stable dielectric constant (ε_r), and low dielectric loss (tan δ) at high temperatures and low electric fields, PEI experiences a significant increase in conductivity loss under high temperatures and high electric fields. Ultimately, the energy storage performance of pristine PEI at high temperatures significantly fails to meet expectations.

In recent years, researchers have been working on creating composites that exhibit high energy storage density (U_e) and can operate effectively at elevated temperatures for a while now. To enhance the energy storage capabilities of polymer-based composites at elevated temperatures, a number of ingenious techniques were widely used, including the design of chemical crosslinking structural all-organic polymers, the construction of polymer-based composites with multilayers or smart filler–matrix interfacial regions and the introduction of wide bandgap inorganic nanofillers.^17–21 For example, Dong et al.²² prepared the PI films that were filled with silicon dioxide (SiO₂)-coated barium titanate (BaTiO₃) nanofibers and encased in alumina (Al₂O₃) layers. The composites' E_b is enhanced by the wide band gap SiO₂ and Al₂O₃, while BaTiO₃ contributes to their higher polarization. This allows the composites to achieve a high U_e and charge–discharge efficiency (η) of 1.75 J cm⁻³ and above 90% at 200 °C. By leveraging electrostatic interactions among their oppositely charged phenyl groups, Yang et al.²³ incorporate three-dimensional (3D) rigid aromatic molecules into aromatic polyimides to create physical crosslinking networks. The dense physical crosslinking networks enhance the polyimides, increasing the E_b, while the aromatic molecules reduce loss by trapping charge carriers. In extreme environments, the prepared composites demonstrate consistent performance over an extended period of 10⁵ charge–discharge cycles. All of these efforts have produced exciting results, raising the high-temperature U_e and η of polymer-based dielectrics to unprecedented heights. However, due to the contradictory and mutually exclusive relationship between E_b and polarization in polymer-based composites at high temperatures, enhancing the energy storage performance of polymer dielectrics at high temperatures continues to be a significant challenge when compared to their room-temperature performance.^24,25

Metal–organic frameworks (MOFs) are network-like materials that resemble networks that are created by joining organic linkers with metal nodes. MOFs are renowned for their adaptable size, porosity, specific surface area, and crystallinity tunability.^26–28 Consequently, a wide range of applications including electrochemistry, sensing, catalysis, separation, gas storage, and energy storage have made extensive use of MOFs with porosity, structural diversity, and functionally adaptive properties.^29–32 In comparison, studies investigating the use of MOFs' 3D topological structure in electronic devices, particularly flexible ones, remain at a nascent stage. Small amounts of MOFs can be added to the polymer to reduce conductivity loss and improve the U_e and η values of polymer dielectrics under high-temperature conditions, as demonstrated in a few prior works.^16,33–36 Nevertheless, limited research has been conducted on utilizing diverse MOFs as fillers with highly tunable structures to enhance the energy storage capabilities of polymers in high-temperature and strong external electric fields environments.

In this work, a novel type of classical MOF (ZIF-11) has been prepared and integrated into PEI for high-temperature capacitive energy storage applications, where the manufactured nanocomposites are referred to as ZIF-11/PEI. It was found that the electrostatic interactions between oppositely charged phenyl groups in ZIF-11 and PEI can result in the creation of physical crosslinking networks, reducing free volume and ionic conduction, which serve to enhance the mechanical strength of PEI and thus increase its E_b. Additionally, ZIF-11 is capable of trapping charge carriers to reduce the electronic conduction, thus further improving the energy storage performance of ZIF-11/PEI nanocomposites. Consequently, the ZIF-11/PEI nanocomposites demonstrate notably enhanced high-temperature energy storage performance by the addition of an ultralow amount of ZIF-11 into the PEI matrix. For example, 1 wt% ZIF-11/PEI nanocomposites can attain a U_e of 5.44 J cm⁻³ with an η of 80.2%, representing a 138% increase compared to pure PEI (2.29 J cm⁻³) at 150 °C. This work points to a new avenue for investigating the application of MOFs with framework structures in high temperature energy storage materials.

2. Experimental section

2.1 Materials

Zinc nitrate hexahydrate (Zn(NO₃)₂·6H₂O, A.R.) and benzimidazole (BIm, 98%) were obtained from Shanghai Aladdin Biochemical Technology Co., Ltd. Toluene (A.R.) and ammonium hydroxide (NH₃, 28–30% aqueous solution) were purchased from Tianjin Fuyu Fine Chemical Co., Ltd. 1-Methyl-2-pyrrolidone (NMP, A.R.) and methanol (MeOH, A.R.) were supplied by Shanghai Macklin Biochemical Co., Ltd. PEI was bought from SABIC. Every commercial reagent was utilized without additional purification unless otherwise noted.

2.2 Synthesis of ZIF-11 particles

Following a standard process, 0.47 g of BIm (4 mmol) was dissolved in 19.22 g of methanol (600 mmol), after which 18.43 g of toluene (200 mmol) and 0.24 g of ammonia hydroxide (4 mmol NH₃) were added under continuous stirring at room temperature. After that 0.59 g of zinc nitrate hexahydrate (2 mmol) was introduced and the mixture was stirred for 1 h at ambient temperature (∼25 °C). Ultimately, the resulting white precipitated solution was centrifuged at 9000 rpm for 5 min using a centrifuge, subjected to three washes with methanol, and subsequently vacuum-dried at 80 °C for 12 h to yield ZIF-11 particles. The synthesis procedure for ZIF-11 is depicted in Fig. S1a.†

2.3 Synthesis of ZIF-11/PEI nanocomposites

As seen in Fig. S1b,† the ZIF-11/PEI nanocomposite films were fabricated through a typical solution casting technique. To prepare ZIF-11/PEI nanocomposite films with varying filler contents, the necessary mass fraction of ZIF-11 (0.25 wt%, 0.5 wt%, 1 wt%, 2 wt%, and 3 wt%) was first distributed ultrasonically into NMP for a 1 h at room temperature in order to ensure thorough mixing. After that, the mixed solution received 0.6 g of PEI, which was agitated for 24 h at ambient temperature. Subsequently, the solution was uniformly applied onto a glass plate using the casting method. It was subsequently oven-dried at 80 °C for 6 h, and for 6 h at 200 °C in a vacuum in order to totally eliminate the solvent. The same process was used to prepare pure PEI film without ZIF-11 fillers for comparative purposes.

2.4 Characterization

The chemical structure of the fillers and PEI-based nanocomposites was examined using Fourier transform infrared spectroscopy (FT-IR, Bruker EQUINOX-55). X-ray diffraction (XRD, D8 Advance, German) was used to analyze the structures of MOFs and PEI-based nanocomposites. Cu Kα radiation was used, and the scan speed was 5° min⁻¹. The thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) analysis of ZIF-11 was conducted using a TGA/DSC1 (Swiss) synchronous thermal analysis system. A DSC analyzer (DSC214, German) was used to measure the PEI films' thermal behavior at a heating rate of 10 °C min⁻¹ in a nitrogen atmosphere. Using a field emission scanning electron microscope (FESEM, FEI Quanta450) fitted with a IE250X-Max50 energy-dispersive spectrometer (EDS) to show the distribution of elements on the sample surface, the morphological structures of MOFs and their nanocomposites were examined. Isothermal surface potential decay (ISPD) curves were measured using a Trek P0865 surface potentiometer. Prior to the test, the surface of the samples was charged using the corona discharge method, and the surface potential decay curve of the sample after charging was measured with a Trek P0865 surface potentiometer. Before the films were put through subsequent dielectric and ferroelectric performance tests, gold electrodes with a diameter of 4 mm were sputtered on both surfaces. The high temperature dielectric temperature spectrum measurement system (TZDM-RT-1000, China) and a precise impedance analyzer (Agilent 4294 A, USA) were used to measure the temperature dependence (from 25 °C to 200 °C) and frequency dependence (from 100 Hz to 10⁷ Hz) of the dielectric properties of the nanocomposites, respectively. A TF Analyzer 2000E ferroelectric polarization system operating at 100 Hz was used to measure the E_b and the electric displacement–electric field curves (D–E loop) of all polymer films. Based on the acquired D–E loops, the U_e and η of the nanocomposites were computed.

2.5 Density functional theory (DFT) computation

The DMol3 module of Materials Studio (MS) software was utilized to conduct DFT calculations for the geometry optimization of PEI and ZIF-11, as well as their electrostatic potential distribution. The calculations employed a hybrid approach incorporating the large basis set (DNP 4.4) and a three-parameter Becke-style hybrid functional (B3LYP) exchange–correlation functional.

2.6 Molecular dynamics (MD) simulation

Different models with varying proportions were developed using the AC module of the MS software in the MD simulations. To achieve stable configurations, the constructed models underwent geometric structure optimization. Iterative optimization was employed to determine the optimal force field parameters, using Compass as the selected force field, combined with the force field assigned charge distribution method and the smart optimization calculation method. The microcanonical ensemble (NVT) was chosen for the MD calculations, utilizing the Nose thermostat for temperature control and the Velocity Verlet algorithm applied to resolve Newton's equations of motion. The van der Waals interactions were addressed using an atom-based method with a cutoff radius set at 12.5 Å, and the simulation was conducted for 500 ps with a time step of 1 fs. The resulting data were imported into the MS software to determine the appropriate polymer configurations for calculating the free volume, occupied volume, and binding energy.

2.7 Finite element method simulations

To visualize the temperature distribution in the capacitor made of dielectric films under high temperature conditions, finite element analysis was conducted by COMSOL Multiphysics. The heat transfer model used for the mathematical models of the PEI and ZIF-11/PEI films was as follows:Here, ρ, C_p, t, T, K and Q represent density, specific heat capacity, time, temperature, thermal conductivity and thermal power density, separately. The value of Q was determined using the energy loss derived from the D–E loop, calculated with the equation: Q = f × U_loss. U_loss represents the energy loss in the D–E loop of the materials at 300 MV m⁻¹, while f denotes the frequency. The simulation employed a cylindrical sample measuring 5 mm in radius and 10 mm in height with the ambient temperature was fixed at 150 °C.

3. Results and discussion

3.1 Morphology and structure of ZIF-11

The depiction of the correlated structure and characteristics of ZIF-11 particles is illustrated in Fig. 1. Fig. 1a depicts the FT-IR spectrum of the synthesized ZIF-11 particles. The FT-IR spectra reveal that the peaks at 1609 and 1467 cm⁻¹ are associated with the stretching of the C–C bond in ZIF-11, whereas the peak at 424 cm⁻¹ corresponds to the vibration of the Zn–N bond in ZIF-11.³⁷ In Fig. 1b, the XRD pattern of ZIF-11 is presented. The XRD pattern indicates that the peaks at 2θ = 6.09°, 7.51°, 15.70°, 16.86°, and 18.51° correspond to the (002), (112), (015), (044), and (006) planes, separately. These peaks are considered to be the typical characteristic features of ZIF-11, aligning with those documented in existing literature.³⁸ Meanwhile, it aligns with the simulated ZIF-11 patterns, indicating that the products were characterized by a pure phase and high crystallinity. The TGA of the synthesized ZIF-11 reveals a progressive weight decrease of 4.01% between 30 and 350 °C, attributed to the evaporation of guest molecules (e.g. H₂O) and residual solvents within ZIF-11. A subsequent plateau observed from approximately 580 °C aligns with the thermal decomposition of the ZIF-11 framework, as demonstrated in Fig. 1c. Additionally, a distinct exothermic peak emerges in the DSC curve of the synthesized ZIF-11 around 572 °C, indicating the onset of decomposition and transformation into ZnO at this temperature stage. It is noteworthy that the maximum temperature for the thermal treatment process of the ZIF-11/PEI nanocomposites is 200 °C. Accordingly, the thermal treatment of the PEI-based composites does not lead to the decomposition of the ZIF-11. Furthermore, the removal of guest molecules and residual solvents may impart high crystallinity to the ZIF-11, potentially enhancing the overall performance of the ZIF-11/PEI nanocomposites. Utilizing FESEM, the microstructural analysis of ZIF-11 was conducted, as depicted in Fig. 1d. The ZIF-11 particles exhibit a typical rhombic dodecahedron morphology, with sizes mainly falling between 400 and 600 nm. Moreover, based on the elemental mapping, both Zn, N and C elements displayed uniform distribution throughout the entirety of ZIF-11 (Fig. 1e). In conclusion, the aforementioned characterizations confirm the successful synthesis of ZIF-11.

Fig. 1 (a) FT-IR spectrum, (b) XRD pattern, (c) TGA&DSC analysis, (d) FESEM image and (e) element mapping images of the ZIF-11.

3.2 Quantum chemical computation

To explore the influence of ZIF-11 on the molecular chain structure of PEI, we initially explored the electrostatic potential distribution in ZIF-11 and PEI using DFT calculations. Fig. 2a illustrates that some of the phenyl groups within ZIF-11 carry negative charged (highlighted in red), while certain phenyl groups in PEI display positive charged (highlighted in blue), indicating potential electrostatic interaction between the ZIF-11 and host PEI. To gain further insight into the interaction between ZIF-11 and PEI, we conducted MD simulations (Fig. 2b and S3†). The comparisons of the computed occupied volume, free volume, free volume proportion (FFV, free volume/sum of free and occupied volume) and the binding energy between PEI and ZIF-11/PEI are depicted in Fig. 2c. In the molecular structure model, the blue color represents free volume, while the other colors represent different molecular chains. PEI initially exhibits a free volume of 4709.2 ų and a free volume fraction of 12.64%. Upon incorporating ZIF-11, these values decrease to 3932.4 ų and 10.87%, respectively. The decrease in free volume within the ZIF-11/PEI nanocomposite is a result of ZIF-11's electrostatic crosslinking of polymer chains, which encourages a denser arrangement of chains. This is further supported by the higher binding energy observed between the ZIF-11/PEI chains. Meanwhile, electrostatic interactions improve the interfacial compatibility between ZIF-11 and PEI, which will facilitate the dispersion of ZIF-11 within the PEI matrix. In summary, the simulation results suggest that incorporating ZIF-11 effectively decreases the free volume in PEI, thereby impacting its dielectric and energy storage properties.

Fig. 2 (a) Schematic diagrams of structures and electrostatic potential distributions of ZIF-11 and PEI, with oppositely charged phenyl groups highlighted (red circles indicate negatively charged phenyl groups in ZIF-11, and blue circles indicate positively charged phenyl groups in PEI). (b) Polymer configuration of PEI and ZIF-11/PEI. (c) The occupied volume, free volume, free volume proportion, and binding energy for both PEI and ZIF-11/PEI nanocomposite.

3.3 Characterizations of the nanocomposites

Fig. 3a and b depict the cross-sectional SEM micrographs of PEI and the 1 wt% ZIF-11/PEI nanocomposite. The films' thickness measures around 10 μm, and their cross-section appears smooth without visible defects. As illustrated by the EDS mapping in Fig. 3b, the Zn element is uniformly distributed across the film area, demonstrating that ZIF-11 particles are evenly distributed within the PEI matrix, benefiting from their porous structure and functionalized surface. Fig. 3c displays the XRD patterns of both pure PEI and ZIF-11/PEI nanocomposites, where all polymer shows a broad diffuse scattering peak at ∼15°, characteristic of amorphous PEI. The spacing of all polymer chains was calculated using XRD data, as depicted in Fig. 3d. The interchain spacing reduces from 6.28 Å in pristine PEI to 5.68 Å in the 1 wt% ZIF-11/PEI nanocomposite, demonstrating tighter chain packing due to electrostatic interactions within the blended film. When the filler content is higher (>1 wt%), the interchain spacing of PEI nanocomposites increases instead, possibly due to excessive filler introduction, which adds additional defects and, in turn increases the free volume within the material. To examine the functional group present in ZIF-11/PEI, the FT-IR spectra of the nanocomposites were obtained, as shown in Fig. 3e. The characteristic peaks in the nanocomposites predominantly correspond to those observed in PEI. Specifically, PEI exhibits characteristic absorption peaks at 1777, 1716, 1351, and 741 cm⁻¹, representing symmetric stretching of C=O, asymmetric stretching of C=O, stretching of C–N bonds, and bending of C–N bonds of imide carbonyls, separately.³⁹ Based on the DSC results depicted in Fig. 3f, the T_g and the specific heat capacity change (ΔC_p) were determined for both PEI and ZIF-11/PEI nanocomposites, as illustrated in Fig. 3g. The T_g of 1 wt% ZIF-11/PEI nanocomposites increases slightly from 210.2 °C in PEI to 215.5 °C in 1 wt% ZIF-11/PEI, suggesting enhanced structural stability of the ZIF-11/PEI nanocomposites at elevated temperatures. Based on the DSC results, the blended films based on PEI demonstrate high temperature resistance. Significantly, ΔC_p drops from 0.116 J g⁻¹ K⁻¹ for PEI to 0.098 J g⁻¹ K⁻¹ in the 1 wt% ZIF-11/PEI, indicating that the electrostatic attraction between ZIF-11 and PEI matrix straightens the PEI chains compared to their naturally entangled state, which typically exhibits more voids and defects in pure polymers. The solubility test was also conducted on both pure PEI and the ZIF-11/PEI nanocomposites. According to Fig. 3h, pure PEI was completely dissolved in NMP within 1 h. However, the ZIF-11/PEI nanocomposite was retained after 1 h and exhibited a high gel content. These aforementioned results suggest that the molecular structure of 1 wt% ZIF-11/PEI is the densest, aligning with the outcomes of MD simulations. This further verifies that the ZIF-11 reduces the free volume and molecular spacing between PEI molecular chains, which increasing the polymer density. In brief, the creation of physical crosslinking networks significantly enhances the density of polymer chain packing and strengthens the polymers, suggesting enhanced dielectric and energy storage properties.

Fig. 3 (a) Cross-sectional FESEM image of raw PEI. (b) Cross-sectional FESEM diagram of the 1 wt% ZIF-11/PEI nanocomposite, accompanied by corresponding elemental mapping image for Zn, N, O, and C elements. (c and d) XRD patterns for ZIF-11, raw PEI, and ZIF-11/PEI nanocomposites, as well as the interchain spacing calculated from XRD data. (e) FT-IR spectra of the raw PEI and the ZIF-11/PEI nanocomposites. (f) DSC curves for raw PEI and 1 wt% ZIF-11/PEI. (g) T_g and ΔC_p as determined from the DSC analysis of PEI and 1 wt% ZIF-11/PEI. (h) The macroscopic photograph of raw PEI and 1 wt% ZIF-11/PEI nanocomposite before and after undergoing a solubility test in NMP.

3.4 Dielectric performances of polymer nanocomposites

Dielectric properties, such as the ε_r and tan δ, are highly important in enabling film capacitors to achieve superior energy storage capacity. To examine how ZIF-11 impacts the dielectric characteristics of PEI, we analyzed the ε_r and tan δ across varying frequencies and temperatures (Fig. 4). Fig. 4a illustrates the variation of the ε_r and tan δ of all polymers at ambient temperature across frequencies ranging from 100 Hz to 10⁷ Hz. The ε_r of the nanocomposite exhibits a gradual decrease, while the tan δ shows a corresponding gradual increase as the frequency rises. This behavior conforms to the characteristic dielectric relaxation phenomenon.^40–42 As shown in Fig. 4c, at identical frequencies, the ε_r of the nanocomposite first reduces and then raises as the ZIF-11 loading is increased. Typically, the ε_r of materials displays a negative correlation with E_b,^43,44 as empirically described by the relationship: E_b ∼ ε_r^−0.65. Therefore, a lower ε_r contributes to an increase in the E_b of the material. In contrast, with the further incorporation of ZIF-11, the nanocomposites experience a slight increase in ε_r due to the expanded interface area. The tan δ values for all nanocomposites remain below 0.015, attributed to the robust backbone structure of the PEI matrix and excellent interface compatibility. It's worth mentioning that the tan δ of most nanocomposites decreases compared to that of PEI. For example, the nanocomposite 1 wt% ZIF-11/PEI demonstrates the smallest tan δ value of 0.0024 at 10² Hz, notably lower than PEI's tan δ of 0.0035. The reduction in ε_r observed in nanocomposites with low levels of ZIF-11 fillers can be attributed to ZIF-11's superior insulation and low dielectric properties. Additionally, the porous nature of ZIF-11 fillers, combined with the interaction between ZIF-11 and PEI, inhibits dipole polarization in polymers, leading to a reduction in relaxation loss. As a result, the tan δ stays at a low level, contributing to the enhancement of η. We also plotted the impedance (Z) versus frequency for the pristine PEI and ZIF-11/PEI nanocomposites, as shown in Fig. 4b. The impedance can be formulated as:

Z = R + j(X_L + X_C) (2)

where R denotes resistance, j represents imaginary units, X_L signifies inductive reactance and X_C denotes capacitive reactance. According to experimental results, the impedance of all nanocomposites studied displays a similar frequency dependence, decreasing as the frequency increases. At identical frequencies, the Z of the 1 wt% ZIF-11/PEI nanocomposite notably exceeds that of pure PEI, suggesting a decrease in its conductivity (σ) (corresponding to the results in Fig. S4†), which will be beneficial for increasing its E_b.

Fig. 4 (a) ε_r, tan δ and (b) Z dependent on frequency for ZIF-11/PEI nanocomposites with varying filler contents. (c) The comparison of ε_r and tan δ for PEI and ZIF-11/PEI nanocomposites at 10² Hz. (d) The temperature dependence of ε_r and tan δ at 10³ Hz for PEI and 1 wt% ZIF-11/PEI.

To investigate how the addition of ZIF-11 fillers impacts the temperature stability of the ε_r and tan δ in nanocomposites, we conducted tests on the dielectric performance of polymer films across a temperature range spanning from ambient temperature up to 200 °C. Fig. 4d depicts the changes in ε_r and tan δ of the nanocomposites at 10³ Hz across various temperatures. The ε_r and tan δ of the ZIF-11/PEI exhibits more excellent stability as temperature rises. Additionally, thanks to the suppression of both ionic and electronic conduction by ZIF-11, the σ of the ZIF-11/PEI nanocomposite at high temperatures was significantly reduced, as shown in Fig. S5.† For example, the σ specific of 1.0 wt% ZIF-11/PEI nanocomposites is 2.31 × 10⁻¹¹ S cm⁻¹ at 150 °C, which is significantly lower compared to pure PEI (3.36 × 10⁻¹¹ S cm⁻¹ at 150 °C). This confirms the significant thermal stability of the nanocomposites mentioned earlier.

To investigate the influence of ZIF-11 on the E_b of the samples, the characteristic E_b was assessed using the two-parameter Weibull distribution, as described by the following formula:

P = 1 − exp(−(E/E_b))^β (3)

P, E, and β denote the cumulative failure probability, the experimental breakdown strength value, and the shape parameter indicating the dispersion degree of sample data, separately. E_b represents the characteristic breakdown strength, signifying the electric field strength at which the sample's breakdown probability reaches 63.2%. Fig. S6† and 5a illustrate the Weibull distribution of pristine PEI and the ZIF-11/PEI nanocomposites with varying filler loadings at room temperature (25 °C) and 150 °C, respectively. At 150 °C, all E_b values are slightly lower than those at 25 °C, primarily because the leakage current within the nanocomposites increases as the temperature rises with an applied electric field. Moreover, as the ZIF-11 filler content increases, the E_b of the nanocomposite films initially rises before subsequently declining. According to classical dielectric breakdown theory,⁴⁵ the inclusion of free volumes (e.g., pores) typically lowers the E_b of nanocomposites. Surprisingly, the incorporation of ZIF-11, known for its porous structure, not only does not decrease the E_b but also enhances it in PEI at lower concentrations. This surprising enhancement demonstrates that the ZIF-11 has the ability to produce a highly effective 3D interface interaction with PEI and successfully construct a physical crosslinking network within the PEI matrix. With a ZIF-11 loading of 1 wt%, the nanocomposite exhibits a significantly elevated E_b of 645.5 MV m⁻¹ at 150 °C (43% higher than that of pure PEI). This enhancement in E_b stems from the effective establishment of physical crosslinking networks. The compact structure of the ZIF-11/PEI nanocomposites indicates the minimal physical defects, which restricts the relaxation of polymer chains at high temperature and inhibits the mobility of ionic species, thereby reducing the likelihood of breakdown in the nanocomposite (as depicted in Fig. 6). However, additional filler content will result in filler agglomeration, facilitating the formation of a conductive pathway that diminishes the material's E_b.

Fig. 5 (a) Weibull distribution analysis of the breakdown strength for PEI and ZIF-11/PEI nanocomposites at 150 °C. (b) The dependence of ln σ on the inverse temperature (1000/T) for both PEI and the 1 wt% ZIF-11/PEI. (c) The schematic representation of energy distribution across various regions. (d) Leakage current density versus electric field for ZIF-11/PEI nanocomposites at 150 °C. (e) A comparison of the injection barrier heights and hopping distances in PEI and ZIF-11/PEI nanocomposites. (f) Trapped density distribution of PEI and 1 wt% ZIF-11/PEI.

Fig. 6 The schematic diagram illustrates the improvement in E_b of PEI after the introduction of ZIF-11.

3.5 Conductivity mechanism of polymer nanocomposites

To further investigate the breakdown mechanism of polymer nanocomposites, we also calculated the activation energy (E_a) for charge carrier migration in the polymers based on the Arrhenius equation. The σ exhibits a strong dependence on temperature (T), and this relationship can be expressed by the following equation:where σ₀ is the pre-exponential factor, representing the conductivity at the high-temperature limit, while T and k_B denote the absolute temperature and the Boltzmann constant, separately. The E_a can be derived from the plots of ln σ versus 1000/T (Fig. 5b), which represents the energy barrier that must be overcome for carrier transport. The improved E_a suggests that the movement of carriers is restricted under the application of an electric field, contributing positively to the enhancement of insulation strength. The enhanced E_b of ZIF-11/PEI nanocomposites is further supported by the reduced leakage current observed at high temperatures in Fig. 5d. As observed, the leakage current density (J) decreases from 9.82 × 10⁻⁷ A cm⁻² in PEI to 1.57 × 10⁻⁷ A cm⁻² in 1 wt% ZIF-11/PEI at 300 MV m⁻¹ and 150 °C, indicating a significant reduction in conduction loss and improved η. It should be noted the substantial reduction in leakage current is likely contributed by the suppression of both ionic and electronic charge transport. The zinc ions in the ZIF-11 exhibit intrinsically higher electron affinity than the PEI matrix, leading to the formation of deep traps that effectively capture and immobilize electronic carriers. Simultaneously, the imidazolate ligands of ZIF-11 interact strongly with PEI (via strong electrostatic interactions between oppositely charged phenyl groups, Fig. 2a and 6), constraining polymer chain mobility and therefore reducing ionic conduction.

It is important to emphasize that these two charge suppression mechanisms operate via distinct physical principles, which are often conflated in the literature.^14,22,33 Electronic charge transport is primarily influenced by band structure, where high-electron-affinity species immobilize charge carriers and create deep traps. In contrast, ionic conduction is governed by segmental polymer mobility, free volume, and electrostatic interactions, rather than energy band modifications.

In this study, we leverage ZIF-11's dual structural characteristics to simultaneously suppress both charge transports, demonstrating a rational design strategy for achieving multifunctional charge suppression in polymer dielectrics.

By fitting the field-dependent current curves with various conduction equations, further examination of the charge transport behavior reveals that Schottky emission is the dominant mechanism in the lower electric field region, whereas hopping conduction becomes predominant in the higher electric field region (Fig. 5d). The Schottky emission model is described by the following expression:

where A represents the Richardson constant, T denotes the absolute temperature, e is the electronic charge, eφ stands for the Schottky barrier height, ε₀ is the vacuum dielectric constant, ε_i is the optical dielectric constant and k_B is the Boltzmann constant. The calculated height of the injection barrier is shown in Fig. 5e, and we can clearly see that the injection barriers of all nanocomposites are significantly enhanced compared to pure PEI. The reason for this is that the ZIF-11 introduces additional traps, which effectively restrict the de-trapping and migration of injected charges. These trapped electrons accumulate near the electrode/dielectric interface, forming a charge-accumulated region. This creates a built-in electric field oriented opposite to the applied field, thereby inhibiting the injection of interface charges into the polymer film. Under high-temperature and high-electric-field conditions, the hopping conduction model can be expressed as:Here, n represents carrier concentration, V is the attempt-to-escape frequency, λ denotes the hopping distance, and E_a is the activation energy. This expression can be reduced to:

J(E) = A × sinh(BE) (7)

Here, A and B are two lumped parameters. The hopping distance significantly is reduced from 0.92 nm in virgin PEI to 0.36 nm in the 1.0 wt% ZIF-11/PEI sample, as shown in Fig. 5e. A shorter hopping distance suggests a greater number of traps, which are more effective at capturing and dispersing charge carriers, consequently suppressing leakage current. The variations in leakage current and hopping distance correspond to the trend observed in E_b, which also confirms that optimizing the ZIF-11 doping level results in the most effective exciton trapping in ZIF-11/PEI nanocomposite films, thus inhibiting leakage conduction. Moreover, the significant interaction between ZIF-11 and PEI results in a high energy potential barrier at the bonding interface (as depicted in Fig. 5c). This barrier likewise substantially decreases the leakage current density and increases the E_b. As depicted in Fig. 5c, E_V, E_F, and E_C denote the valence band, Fermi level and conduction band, separately. The parameters φ_d(h) and φ_d(e) observed in the bonded area suggest the presence of deep traps with elevated energy barriers, while φ_s(h) and φ_s(e) in the transition zone signify lower the energy barriers. As carriers transition from the transition region to the bonded region under the influence of electric fields, their energy diminishes as the barrier energy increases.

To further verify the introduction of traps in PEI, the isothermal surface potential decay (ISPD) method was used to analyze the energy distribution of charge traps (Fig. 5f). Fig. 5f showcases that the density of traps in the 1 wt% ZIF-11/PEI is significantly greater than that in neat PEI. This indicates that the ZIF-11 particles help to enhance the trap density and introduce more localized deep charge traps on the PEI surface. The additional traps can limit the de-trapping and movement of injected charges, preventing interface charges from entering the PEI matrix. In a word, the 3D structure of ZIF-11 offers additional trapping sites, which efficiently lower the carrier density in the nanocomposites. This reduction in carrier energy and density also significantly increases the E_b.

3.6 Energy storage performances of polymer nanocomposites

We evaluate the energy storage capabilities of both raw PEI and the ZIF-11/PEI nanocomposites under various temperature conditions (25 °C and 150 °C) using D–E loops (as illustrated in Fig. S7† and 7). It is evident that the maximum displacement (D_max) and remnant displacement (D_r) steadily rise as the electric field increases. At 25 °C, the narrow D–E loops indicate the low energy loss and sustained high η. To assess the electric displacements in the ZIF-11/PEI nanocomposite films at elevated temperatures, the values of D_max and D_r are graphically represented in Fig. 8a. Clearly, both the maximal E and D_max of the nanocomposite thin films first rise and then fall as the ZIF-11 content increases, yet they overall exceed those of raw PEI. For instance, the 1 wt% ZIF-11/PEI nanocomposite achieves a maximum E of 625 MV m⁻¹, which is 1.39 times higher than that of pure PEI (450 MV m⁻¹), while its D_max at 625 MV m⁻¹ (2.06 μC cm⁻²) is 1.44 times greater than that of pure PEI (1.43 μC cm⁻²). Furthermore, the D–E loops of the polymer films containing ZIF-11 filler remain notably narrow, suggesting that the incorporation of ZIF-11 does not lead to a decrease in η. Therefore, by incorporating a small quantity of ZIF-11 into the PEI matrix, energy loss is effectively minimized at high temperatures, leading to improved η of the PEI nanocomposites.

Fig. 7 D–E loops of (a) raw PEI. (b) 0.25 wt% ZIF-11/PEI. (c) 0.5 wt% ZIF-11/PEI. (d) 1 wt% ZIF-11/PEI. (e) 2 wt% ZIF-11/PEI and (f) 3 wt% ZIF-11/PEI at different electric field at 150 °C.

Fig. 8 (a) Comparison of the maximum polarization and remanent polarization in pure PEI versus ZIF-11/PEI nanocomposites at 150 °C. (b) Energy storage properties of raw PEI and ZIF-11/PEI nanocomposites at 150 °C. (c) Variation chart of U_e and η as ZIF-11 content increases at 150 °C. (d) Simulated rise in temperature of capacitors constructed from pristine PEI and 1 wt% ZIF-11/PEI nanocomposite under sustained operation at 300 MV m⁻¹ and 150 °C. (e) Cyclic durability of raw PEI and 1 wt% ZIF-11/PEI nanocomposite at 300 MV m⁻¹ and 150 °C. (f) Comparison of U_e at η ≈ 80% between this work and other works at 150 °C.

The values of U_e and η are determined by integrating the D–E loops using the energy storage formula:

U_e = ∫EdD (8)

η = U_e/U_c (9)

In the given formula, E denotes the applied electric field, D represents induced electrical displacement, and U_c signifies charge energy density. Fig. S8† displays the U_e and η of ZIF-11/PEI nanocomposites at room temperature (25 °C). Thanks to the incorporation of ZIF-11, the E_b of the nanocomposites is significantly enhanced, and the U_e of ZIF-11/PEI nanocomposites with low filler content is notably higher than that of raw PEI. At 25 °C, the 1 wt% ZIF-11/PEI nanocomposite exhibits a U_e of 7.97 J cm⁻³ (700 MV m⁻¹), representing a notable improvement over the 3.85 J cm⁻³ of raw PEI, while maintaining an η of 88.52%.

Fig. 8b depicts the U_e and η at 150 °C for all nanocomposite materials. As the temperature rises, the benefits of ZIF-11 fillers become more pronounced. In contrast, pure PEI experiences a significant reduction in E_b and η at high temperatures due to an increase in leakage current. The ZIF-11 fillers substantially enhance the thermal stability of PEI at elevated temperatures, allowing ZIF-11/PEI nanocomposites to retain outstanding energy storage performance even at elevated temperature (150 °C). The nanocomposite with 1 wt% ZIF-11 exhibits the highest U_e of 5.44 J cm⁻³ at 625 MV m⁻¹ and 150 °C, representing a 1.38-fold increase compared to raw PEI (2.29 J cm⁻³ at 450 MV m⁻¹). Meanwhile, the η is enhanced too, with 1 wt% ZIF-11/PEI nanocomposites demonstrating elevated η values and maintaining a high level of over 80% within 625 MV m⁻¹ at 150 °C. A visual comparison of the maximum U_e and η values of pure PEI with those of nanocomposites containing varying loadings of ZIF-11 particles is shown in Fig. 8c. We can clearly see that the energy storage properties of all nanocomposites are significantly enhanced compared to those of pure PEI. These findings suggest that adding ZIF-11 effectively improves the energy storage properties of PEI. This improvement is ascribed to the minimized leakage current and lower conduction loss, resulting from the development of physical crosslinking networks within host PEI matrix.

In the actual application scenario, dielectric materials must operate under specified electric stress. High-field leakage conduction can cause dielectric materials to rapidly heat up, and if this accumulated heat is not effectively dissipated, thermal runaway or heat-die may occur.^46–48 Accordingly, we also simulated the temperature rise in capacitors at 150 °C and 300 MV m⁻¹ using pristine PEI and 1 wt% ZIF-11/PEI nanocomposites as dielectric films, with the simulations performed in COMSOL Multiphysics. As illustrated in Fig. 8d, the capacitor using raw PEI presents a more pronounced temperature increase compared to the capacitor employing 1 wt% ZIF-11/PEI nanocomposites as the dielectric film. The reduced temperature rise is primarily due to the suppression of conduction and energy losses resulting from the incorporation of ZIF-11 particles into the PEI matrix. To evaluate dielectric stability under challenging conditions, the cyclic stability of the 1 wt% ZIF-11/PEI and pure PEI was tested at 300 MV m⁻¹ and 150 °C, as depicted in Fig. 8e. Throughout 10⁵ consecutive charge–discharge cycles, the U_e and η values remain nearly constant, demonstrating that the 1 wt% ZIF-11/PEI nanocomposite maintains exceptional long-term stability at high temperatures. Additionally, we measured and compared the dielectric and energy storage properties of the 1 wt% ZIF-11/PEI nanocomposite before and after multiple cycles, as shown in Fig. S9.† It can be observed that the dielectric and energy storage performances before and after cycling are nearly unchanged. This further demonstrates that the prepared ZIF-11/PEI nanocomposite exhibits excellent long-term operational reliability. Fig. 8f illustrates a comparison of U_e and η between the ZIF-11/PEI nanocomposite and the composites documented in the literature.^{10,14–16,21,22,33,41,49–51} The results indicate that the 1 wt% ZIF-11/PEI nanocomposite manufactured in this work exhibits superior energy storage performance. Moreover, the max. U_e obtained in this study exceeds four times that of BOPP (1.24 J cm⁻³ at 120 °C), indicating that the nanocomposites produced here are promising for high-temperature dielectric energy storage applications.

4. Conclusion

This work presents a novel approach for enhancing dielectric performance of polymeric materials by engineering MOF and harnessing its multifunctional capabilities. Through the incorporation of ZIF-11 into a polyetherimide (PEI) matrix, we achieved remarkable improvement in the dielectric properties of the polymer nanocomposites, particularly under demanding conditions. The ZIF-11 effectively inhibits ion transport through its strong electrostatic interaction with PEI, and meanwhile suppresses electronic conduction by creating multiple trapping sites for electrons/holes, overcoming the limitations of traditional materials. With an ultra-low filler loading (≤1 wt%), the composite achieves a dramatic reduction in leakage currents, enabling a 1.38-fold increase in energy density (5.44 J cm⁻³ at 150 °C and 625 MV m⁻¹) while maintaining high charging–discharging efficiency (>80%), setting a new benchmark for polymer dielectrics.

Looking forward, this work highlights a promising direction for future research: engineering MOFs by selecting organic ligands with tailored interactions to further enhance compatibility with polymer matrices and suppress ionic conduction, while leveraging their intrinsic high electron affinity to efficiently trap electronic carriers. This dual-benefit strategy provides a unique opportunity to address the limitations of conventional fillers, which often lack the ability to simultaneously manage ionic and electronic conduction. Exploring MOF structures with optimized ligand–metal configurations could further expand the scope of high-performance polymer dielectrics, opening new pathways for their application in next-generation power electronics.

Data availability

Data for this article is available on request from the corresponding authors.

Author contributions

Fanrong Kong (first author): investigation, conceptualization, data curation, formal analysis, writing – original draft. Wenying Zhou (corresponding author): resources, visualization, conceptualization, funding acquisition, writing – review & editing. Fan Zhang: resources, investigation. Weiwei Li: data curation, software. Haomiao Li: resources. Yuanwei Zhu: resources. Bin Zhou: resources. Tian Yao: visualization. Bo Li (corresponding author): conceptualization, writing – review & editing.

Conflicts of interest

The authors have no conflicts of interest to declare that are relevant to the content of this article. All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.

Acknowledgements

This work was supported by funding from the National Natural Science Foundation of China (No. 52277028), and acknowledge the Analytic Instrumentation Center of XUST. The authors would like to express their gratitude to Prof. Xiaoxu Liu for his guidance on this work.

References

1. B. Huang, J. Yu, J. Dong, Y. Zhou, L. Zhai, L. Dou, C. Wu, X. Liang, C. Zhang, K. (Ken) Ostrikov and T. Shao, Adv. Mater., 2024, 36(52), 2311713.
2. L. Ghadbeigi, Z. Liu, T. D. Sparks and A. V. Virkar, J. Electrochem. Soc., 2016, 163, A1560–A1565.
3. W. Zhou, Y. Wang, F. Kong, W. Peng, Y. Wang, M. Yuan, X. Han, X. Liu and B. Li, Energy Environ. Mater., 2024, 7(4), e12698.
4. D. Zhao, S. Jiang, S. Yu, J. Ren, Z. Zhang, S. Liu, X. Liu, Z. Wang, Y. Wu and Y. Zhang, Carbon, 2023, 201, 864–870.
5. A. Kamali, Y. Zhang, Z. Liu, E. Schulman, M. Almafrachi, S. Zhang, T. Wang, L. Gonzalez-Lopez, X. Zhao, W. Zhang, Q. Zhou, M. Al-Sheikhly, L. Hu and D. Liu, Mol. Catal., 2023, 551, 113589.
6. Z. Wang, G. Yang, X. Cao, M. Li, T. Zhang, C. Liu, Y. Zhou and Y. Cheng, High Voltage, 2024, 9, 1021–1032.
7. S. Wang, X. Bao, B. Gao and M. Li, Dalton Trans., 2019, 48, 8288–8296.
8. Z. Wang, M. Li, B. Liu, G. Yang, M. Luo, T. Zhang, L. Li, Y. Cheng, Z. Jia and G. Wu, J. Mater. Sci. Technol., 2024, 183, 12–22.
9. M. Luo, Y. Zhou, R. Wang, X. Cao and Z. Wang, Chem. Eng. J., 2024, 501, 157623.
10. H. Li, L. Ren, D. Ai, Z. Han, Y. Liu, B. Yao and Q. Wang, InfoMat, 2020, 2, 389–400.
11. L. Yang, M. Hao, K. Yang, D. Lan, X. Zhang, X. Tian and Z. Wang, IEEE Trans. Dielectr. Electr. Insul., 2024, 32(1), 117–126.
12. M. Yuan, G. Zhang, B. Li, T. C. M. Chung, R. Rajagopalan and M. T. Lanagan, ACS Appl. Mater. Interfaces, 2020, 12, 14154–14164.
13. L. Ren, H. Li, Z. Xie, D. Ai, Y. Zhou, Y. Liu, S. Zhang, L. Yang, X. Zhao, Z. Peng, R. Liao and Q. Wang, Adv. Energy Mater., 2021, 11(28), 2101297.
14. H. Li, Y. Li, X. Yang, D. Liu and X. Liu, Appl. Phys. Lett., 2023, 122(20), 202903.
15. C. Yuan, Y. Zhou, Y. Zhu, J. Liang, S. Wang, S. Peng, Y. Li, S. Cheng, M. Yang, J. Hu, B. Zhang, R. Zeng, J. He and Q. Li, Nat. Commun., 2020, 11, 3919.
16. J. Li, X. Liu, B. Huang, D. Chen, Z. Chen, Y. Li, Y. Feng, J. Yin, H. Yi and T. Li, Mater. Horiz., 2023, 10, 3651–3659.
17. B. Li, C. A. Randall and E. Manias, J. Phys. Chem. C, 2022, 126, 7596–7604.
18. Z. Wang, T. Zhang, M. Hao, M. Li, Y. Zhou, W. Sun, J. Wang and Y. Cheng, Composites, Part A, 2023, 169, 107495.
19. X. Duan, J. Wang, Q. Chai, P. Ma, H. Du, L. Jin, F. Lai, Z. Peng, X. Chao and J. Lu, Ceram. Int., 2024, 50, 18797–18805.
20. K. Gong, Y. Peng, A. Liu, S. Qi and H. Qiu, Composites, Part A, 2024, 176, 107857.
21. M. Fan, P. Hu, Z. Dan, J. Jiang, B. Sun and Y. Shen, J. Mater. Chem. A, 2020, 8, 24536–24542.
22. J. Dong, R. Hu, Y. Niu, L. Sun, L. Li, S. Li, D. Pan, X. Xu, R. Gong, J. Cheng, Z. Pan, Q. Wang and H. Wang, Nano Energy, 2022, 99, 107314.
23. M. Yang, L. Zhou, X. Li, W. Ren and Y. Shen, Adv. Mater., 2023, 35(35), 2302392.
24. B. Li, M. Sarkarat, A. Baker, C. A. Randall and E. Manias, MRS Adv., 2021, 6, 247–251.
25. M. Luo, X. Cao, Y. Zhou, R. Wang and Z. Wang, IEEE Trans. Dielectr. Electr. Insul., 2024, 31, 2539–2547.
26. M. Sun, X. Wang, F. Gao, M. Xu, W. Fan, B. Xu and D. Sun, Microstructures, 2023, 3(4), 2023032.
27. Y. Pan, A. Bhowmick, W. Wu, Y. Zhang, Y. Diao, A. Zheng, C. Zhang, R. Xie, Z. Liu, J. Meng and D. Liu, ACS Catal., 2021, 11, 9970–9985.
28. P. Chen, J. Hou and L. Wang, Microstructures, 2022, 2, 14.
29. F. Gao, X.-A. Yue, X.-Y. Xu, P. Xu, F. Zhang, H.-S. Fan, Z.-L. Wang, Y.-T. Wu, X. Liu and Y. Zhang, Rare Met., 2023, 42, 2670–2678.
30. Z. Liu, S. Cheng, E. Schulman, W. Chen, D. G. Vlachos, Y. Shu, D. T. Tran and D. Liu, Catal. Today, 2023, 416, 113873.
31. D. Zhao, S. Ge-Zhang, Z. Zhang, H. Tang, Y. Xu, F. Gao, X. Xu, S. Liu, J. Zhou, Z. Wang, Y. Wu, X. Liu and Y. Zhang, ACS Appl. Mater. Interfaces, 2022, 14, 54662–54669.
32. S. Wang, J. Liu, Y. Huang and N. Yang, Appl. Surf. Sci., 2020, 530, 147255.
33. X. Li, H. Luo, C. Yang, F. Wang, X. Jiang, R. Guo and D. Zhang, ACS Appl. Mater. Interfaces, 2023, 15, 41828–41838.
34. F. Wang, J. Cai, C. Yang, H. Luo, X. Li, H. Hou, G. Zou and D. Zhang, Small, 2023, 19(26), 2300510.
35. Y. Li, J. Yin, Y. Feng, J. Li, H. Zhao, C. Zhu, D. Yue, Y. Liu, B. Su and X. Liu, Chem. Eng. J., 2022, 429, 132228.
36. H. Hui, Y. Li, Y. Feng, J. Li, H. Zhao, C. Zhu, D. Yue, J. Yin and X. Liu, Polym. Adv. Technol., 2022, 33, 1685–1694.
37. D. T. Nguyen, L. D. T. Nguyen, Q. T. Pham, T. M. Le, B. Q. G. Le, N. X. D. Mai, T. L. H. Doan and L. H. T. Nguyen, Microporous Mesoporous Mater., 2021, 327, 111445.
38. M. Safak Boroglu and A. B. Yumru, Sep. Purif. Technol., 2017, 173, 269–279.
39. C. Yan, Y. Wan, H. Long, H. Luo, X. Liu, H. Luo and S. Chen, Adv. Funct. Mater., 2024, 34(8), 2312238.
40. P. Ma, J. Wang, Y. He and X. Duan, J. Mater. Eng. Perform., 2024, 33, 14329–14334.
41. Z. Liu, M. Yang, Z. Wang, Y. Zhao, W. Wang and Z.-M. Dang, J. Phys. Chem. Lett., 2023, 14, 11550–11557.
42. Z. Wang, Y. Zhou, M. Luo, Y. Zhang, X. Cao, Z. Zhang, R. Wang and X. Zhang, Compos. Sci. Technol., 2024, 248, 110440.
43. J. Di-wu, W. Zhou, Y. Wang, F. Kong, Y. Wang, S. Zhao and X. Liu, FlatChem, 2025, 50, 100813.
44. X. Meng, W. Zhou, X. Chen, F. Kong, J. Zhao, W. Li, Y. Zhang, F. Wang and M. Yuan, Mater. Today Chem., 2025, 43, 102492.
45. S. Li, G. Yin, G. Chen, J. Li, S. Bai, L. Zhong, Y. Zhang and Q. Lei, IEEE Trans. Dielectr. Electr. Insul., 2010, 17, 1523–1535.
46. J.-W. Ren, R.-C. Zeng, J. Yang, Z. Wang, Z. Wang, L.-H. Zhao, G.-L. Wang and S.-L. Jia, J. Appl. Phys., 2024, 136(4), 045101.
47. Y. Li, C. Liu, W. Zhou, Z. Hou, Q. Shi, C. Gong and Y. Wu, Mater. Today Commun., 2021, 29, 102792.
48. W. Zhou, T. Yao, M. Yuan, Y. Yang, J. Zheng and J. Liu, IET Nanodielectr., 2023, 6, 165–181.
49. H. Chen, Z. Pan, W. Wang, Y. Chen, S. Xing, Y. Cheng, X. Ding, J. Liu, J. Zhai and J. Yu, Composites, Part A, 2021, 142, 106266.
50. L. Ren, L. Yang, S. Zhang, H. Li, Y. Zhou, D. Ai, Z. Xie, X. Zhao, Z. Peng, R. Liao and Q. Wang, Compos. Sci. Technol., 2021, 201, 108528.
51. Y. Zhou, C. Yuan, S. Wang, Y. Zhu, S. Cheng, X. Yang, Y. Yang, J. Hu, J. He and Q. Li, Energy Storage Mater., 2020, 28, 255–263.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta00458f

---