Three-dimensional thermal network structured GnPs&MWCNTs@PBO/PEEK composites integrating high thermal conductivity and electromagnetic shielding†

Yageng Bai^a, Hongxia Qian^b, Xueling Cao^c, Fengyu Wen^a, Yashu He^a, Jierun Ma^a, Lin Cheng^a, Yifan Wang^a, Haoyuan Tan^a, Yuxuan Gu^a, Pengbo Lian^a, Rui Chen*^d and Jianxin Mu*^a
aKey Laboratory of High Performance Plastics, Ministry of Education, National & Local Joint Engineering Laboratory for Synthesis Technology of High Performance Polymer, College of Chemistry, Jilin University, Changchun 130012, P. R. China. E-mail: Jianxin_mu@jlu.edu.cn
^bSiping Experimental Middle School, Siping 136000, P. R. China
^cCollege of Science, Qiongtai Normal University, Haikou 571100, P. R. China
^dLight Industry and Chemical Engineering, Guangdong University of Technology, Canton 510006, P. R. China. E-mail: jrchen1136@163.com

First published on 12th August 2024

---

---

1 Introduction

The rapid advancement in next-generation portable electronic devices necessitates the use of dual-functional materials that exhibit both effective thermal management and exceptional electromagnetic interference (EMI) shielding capabilities.^1,2 Poly (ether ether ketone) (PEEK) is extensively employed in the construction of 5G base stations on account of its remarkable chemical and thermal stability, dielectric characteristics, and exceptional processability, among several other beneficial attributes.^3,4 However, the insulating properties of PEEK, originating from its lack of free electrons, contribute to its poor electromagnetic shielding performance. Furthermore, the irregularities in the non-crystalline region and lattice defects in the crystalline region result in extensive non-simple harmonic coupling vibrations during phonon transmission, leading to interfacial scattering and a significant reduction in thermal conductivity. These limitations significantly restrict the potential applications of PEEK-based composites in terms of both thermal conductivity and electromagnetic shielding capabilities. Researchers have attempted to improve the electromagnetic shielding and thermal conductivity of these composites by incorporating conductive and thermal fillers into PEEK. However, achieving the desired enhancement at low filler concentrations remains a considerable challenge.^5,6 Similarly, Chen and colleagues employed a polybenzoxazine (PBZ) modified filler-PEEK interface and used non-covalent bond modification to reduce interfacial thermal resistance (ITR), thereby reducing the interface scattering of phonons. The TC of (GnPs&MWCNTs)@PBZ/PEEK was 6.27 W m⁻¹ K⁻¹ (in-plane) and 3.38 W m⁻¹ K⁻¹ (through-plane) at a 19.84 vol% filler loading.⁷ Although the interface between the filler and matrix is significantly enhanced, the enhancement in thermal conductivity remains minimal, necessitating extensive filler integration. Large quantities of filler make it challenging to achieve the desired increase in thermal conductivity, primarily due to the interfacial thermal resistance (ITR) between polymers and inorganic nanofillers.⁸ To reduce the ITR, various research strategies have been developed to enhance the thermal conductivity of PEEK-based composites. One promising strategy involves modifying the chemical structure of PEEK, which can affect its macroscopic properties, including thermal stability.⁹ Therefore, the primary objective in enhancing the macroscopic thermal conductivity of composites is to identify methods for constructing an optimal filler network, increasing the bridging effect between fillers, and improving the interface contact state to decrease the thermal resistance at the interface.

Considerable research effort has been devoted to enhancing the thermal conductivity of polymers by integrating high-conductivity heat fillers. Carbon-based fillers and their blends, ranging in size from nanometers to micrometers, offer exceptional TC.^10–12 Carbon-based hybrid fillers, including carbon fibers, MWCNTs, and GnPs, can substantially increase TC. This phenomenon results from the synergistic effects produced by the interconnected network of multiple fillers.^13,14 MWCNTs are one-dimensional (1D) fillers possessing a high aspect ratio.¹⁵ GnPs represent a novel material composed of tightly packed single-layer two-dimensional (2D) honeycomb lattice structures formed by sp² hybridized carbon atoms.¹⁶ MWCNTs and GnPs were combined to create 3D hybrid fillers, leveraging the π–π interaction between them to prevent aggregation and promote optimal morphological dispersion.¹⁷ Liu et al. prepared a 3D continuous network of multiwall carbon nanotube/graphene/silicone rubber (SR) elastomers via the hydrothermal reaction and graphitization.¹⁸ The composites achieved an ideal thermal conductivity of 1.30 W m⁻¹ K⁻¹ at a low loading of 2.77 wt%, which is 465% greater than the TC of pure silicone rubber (0.23 W m⁻¹ K⁻¹). Lin et al. developed a GNP/MWCNTs/polypyrrole (PPy)/polyurethane (PU) nanohybrid using a straightforward and scalable method.¹⁹ This hybrid exhibited exceptional EMI SE and demonstrated significant potential for commercial applications. Therefore, the incorporation of hybrid fillers comprising GnPs and MWCNTs is anticipated to be highly advantageous for the production of composites with outstanding TC and EMI characteristics.

Poly (p-phenylene benzobisoxazole) (PBO), a liquid crystal aromatic heterocyclic polymer, demonstrates an exceptionally high thermal decomposition temperature of 650 °C. Despite the potential for single-bond rotations, the planarity of all atoms within the PBO molecule is maintained due to the presence of conjugation, resulting in a linear configuration. The graphitization of PBO fibers leads to the formation of highly ordered graphite structures originating from the conjugated aromatic backbone.²⁰ Wang et al. used PBO-derived graphite as the graft layer to enhance the thermal conductivity of PAN-CF and its epoxy resin composites.²¹ These inherent features play a significant role in maintaining the exceptional rigidity of PBO molecules, effectively reducing phonon scattering caused by side group rotation and interchain vibrations.^22–24 PBO fibers can be degraded into nanofibers (PBONFs) using a strong protic acid, and the nanofibers can maintain the original strength and thermal stability of PBO fibers. Yu et al. fabricated a nacre-like layered structure fluorinated graphene (FG)/PBONF composite film with a TC of 12.13 W m⁻¹ K⁻¹ utilizing the “sol–gel–dissociation” method.²⁵ Additionally, the coplanar and linear arrangement promotes dense packing, whereas the π–π stacking interactions among molecules increase intermolecular interactions, resulting in exceptional mechanical and thermal properties of PBO fibers.²⁶

This study presents the fabrication of a 3D thermally conductive filler skeleton with a “line-to-surface” structure, composed of MWCNTs and GnPs. This was achieved by combining self-assembled PBO fiber balls with hot-pressing techniques. Furthermore, a novel approach is introduced which involves incorporating PBO fiber balls, known for their high TC and stability, into PEEK to construct the PBO@PEEK matrix. This method not only retains the excellent thermal characteristics of PEEK but also significantly enhances the TC of the polymer matrix. The integration aims to enhance heat flow and minimize electromagnetic wave (EW) loss. Additionally, incorporating PBO enhances the filler-PEEK interface, improving interfacial compatibility, facilitating phonon migration, and reducing ITR. To achieve a balance between high TC and exceptional EMI capability, detailed simulations were conducted to evaluate the impact of constructing a 3D thermal network structure on the overall TC and EMI performance of the composites.

2 Experimental

2.1 Materials

The ultrafine PEEK powder was supplied by Changchun High-Performance Plastics Engineering Research Co., Ltd, in Changchun, China. PEEK powder had a melt index of 29 g per 10 min. PBO fibers (HM type) were obtained from Toray Industries Co. Ltd, Japan. MWCNTs and GnPs were purchased respectively from Mitsubishi Japan Ltd and Guoheng Technology Co. Ltd (China). Acetone and ethanol were provided by Xiya Chemical Reagent Co., Ltd. All chemicals were used as received.

2.2 Synthesis of GnPs&MWCNTs@PBO/PEEK composites

To remove the adhesive, the commercial PBO fibers were immersed in acetone at 80 °C and allowed to flow back for 24 h. Following this, these were washed with ethanol and water several times, separately, and then dried at 120 °C for 6 h. The resultant material, referred to as PBO fiber balls, was pretreated to reduce the impact of commercial beam concentration on fiber beams, making it easier for fillers to penetrate the fibers. MWCNTs (2.58 g), GnPs (6.02 g, 7:3), and PBO fiber balls (14 g) were added to a 3D dynamic mixer at 30 rpm for 3 h to obtain GnPs&MWCNTs@PBO composites. Subsequently, PEEK ultrafine powder (6 g) was added to the mixture, which was then mixed at 30 rpm for an additional 1 h. This process resulted in the formation of 19.31 vol% GnPs&MWCNTs@PBO/PEEK composites (detailed information is provided in Table S1†). Fig. 1 shows the schematic representation of the synthesis of 3D filler network structured GnPs&MWCNTs@PBO/PEEK composites. The GnPs&MWCNTs@PBO/PEEK composites were prepared via hot pressing. The process involved preheating the mixture at 380 °C for 10 min, followed by applying a pressure of 50 MPa for 15 min. The composite sheets were then formed by releasing the pressure and allowing the composites to cool to room temperature. Composites with varying filler contents (2.85 vol%, 5.84 vol%, 8.97 vol%, 12.24 vol%, 15.68 vol%, and 19.31 vol%) were prepared using this method.

2.3 Characterization

The composition of the modified carbon fillers was examined using X-ray photoelectron spectroscopy (XPS, Shimadzu/Krayos AXIS Ultra DLD). X-ray diffraction (XRD, Panalytical Empyrean) analysis was carried out using Cu Kα radiation between 2θ = 10– 80° to examine the crystal structures of the fillers. Raman spectroscopy (Raman, Horiba LabRAM HR Evolution) was used to examine the crystal integrity of the carbonaceous materials. The surface profile and appearance of the resulting hybrid fillers and composites were examined using various advanced imaging techniques, including scanning electron microscopy (SEM, HITACHI SU8020), optical super depth-of-field microscopy (OSM, VHX-7000, KEYENCE), transmission electron microscopy (TEM, JEM-F200, JEOL), and energy dispersive spectroscopy (EDS). The Archimedes method (SD-200L) was used to examine the density of the composites. The laser flash method (LFA-467, NETZSCH) was employed to examine the thermal diffusion coefficient in both through-plane and in-plane directions. Samples had a diameter of 12.5 mm, a thickness of 600 μm, and a temperature test range between 25 and 200 °C. TC values were determined using eqn (1):

λ = α × Cp × ρ	(1)

where λ, α, C_p, and ρ represent TC (λ, W m⁻¹ K⁻¹), thermal diffusivity coefficient (α, mm² s⁻¹), heat capacity (C_p, J g⁻¹ K⁻¹), and density (ρ, g cm⁻³), respectively.²⁷ Furthermore, the electrical conductivity of the composite was assessed using a four-probe resistivity tester (2450 SourceMeter, KEITHLEY). The EMI shielding potential of the composites, measuring 22.84 mm × 10.16 mm × 2 mm, was evaluated using the waveguide method with an Agilent N5244A PNA-X (America) across a frequency range from 8.2 to 12.4 GHz.

The EMI SE_T of the GnPs&MWCNTs@PBO/PEEK composites was computed using eqn (2)–(4):²⁸ In eqn (2), SE_T denotes the total shielding effectiveness, SE_A signifies the sum of absorption, SE_R represents the sum of reflection, and SE_M signifies multiple reflections. The shielding effectiveness (SE) is the logarithm of the ratio between the incident power (P_i) and the transmitted power (P₀).

	(2)

	(3)

	(4)

The determination of specific heat capacity was carried out by differential scanning calorimetry (DSC, Mettler Toledo DSC821e), while thermal stability was examined using thermogravimetric analysis (TGA, METTLER) under a N₂ environment with a flow rate of 50 mL min⁻¹ and a heating rate of 10 °C min⁻¹. The coefficient of linear thermal expansion (CTE) of the composites was investigated using a DIL806 (America) in the temperature range of 30 °C to 260 °C at a heating rate of 5 °C min⁻¹. Composite surface temperatures were measured using an infrared thermograph (FOTRIC-226).

3 Results and discussion

3.1 Chemical and morphology analysis of GnPs&MWCNTs@PBO composites

The elemental bonding and chemical composition of GnPs&MWCNTs@PBO were determined by XPS. The XPS spectra, as illustrated in Fig. 2a, demonstrate the existence of the elements C, O, and N. The presence of a distinct carbon C 1s peak at 285.0 eV and a weak oxygen O 1s peak at 531.3 eV in the neat GnPs and MWCNTs, respectively, indicates the existence of defects in the oxygen-containing functional groups. The stronger pyridinic-N(–CN) and the weaker pyrrolic-/pyridinic-N(–NH) peaks at 398.4 and 399.7 eV, respectively, correspond to the PBO, as depicted in Fig. 2b of the XPS N 1s spectrum of the GnPs&MWCNTs@PBO.^29,30 These findings demonstrate the formation of non-covalent interactions between carbon fillers and PBO.

Raman spectroscopy was employed to analyze the composites. Raman spectra of GnPs, MWCNTs, PBO, and GnPs&MWCNTs@PBO were acquired using 514 nm laser excitation, as depicted in Fig. 2c. GnPs and MWCNTs exhibit a prominent peak at 1581 cm⁻¹ (G-band), characteristic of sp² hybridized carbon materials in Raman spectra. Additionally, a weak absorption peak at 1356 cm⁻¹ (D band) and a strong absorption peak at 2730 cm⁻¹ (G′ band) correspond to disordered carbon in the fillers. The intensity of the D band is proportional to the degree of defects and disorder in the carbon fillers and is used to characterize the degree of sp³ hybridization. The Raman spectrum of the pure PBO polymer exhibits multiple bands attributed to amorphous carbon and disordered or defective structures. Structural units of PBO are characterized by distinct bands at 1628 cm⁻¹, 1279 cm⁻¹, and 1180 cm⁻¹.³¹ For GnPs&MWCNTs@PBO, no significant band displacements were observed. However, the gap between the A and B peaks of the PBO polymer gradually fills upon the addition of carbon fillers, confirming their presence and uniform dispersion within the PBO matrix. This phenomenon is attributed to the π–π interactions between the carbon fillers and the conjugated PBO backbone.³² The slightly lower I_D/I_G ratio observed in GnPs&MWCNTs@PBO further confirms the incorporation of the PBO polymer and the retention of its crystalline structure.^24,33

The XRD patterns of MWCNTs, GnPs, PBO and GnPs&MWCNTs@PBO are depicted in Fig. 2d. The (002) diffraction peak corresponding to carbon fillers prominently appeared in the XRD patterns of all samples (2θ = 26°), indicating the highly ordered crystalline structure of the carbon nanomaterials.³⁴ Significantly, the primary characteristic peak was observed at 16.2° (200), with additional peaks at 26.3° and 27.5° corresponding to diffractions in the (010) and (210) crystalline planes, respectively.^35,36 There is no significant shift in the diffraction positions of the carbon fillers following their incorporation into PBO, indicating that their structural integrity and crystallinity are not affected by non-covalent modifications, which helps maintain thermal and electrical conductivity. The addition of PBO introduces distinct crystalline peaks in the XRD pattern of GnPs&MWCNTs@PBO. Compared to MWCNTs and GnPs alone, the (200) diffraction peak in the XRD pattern of GnPs&MWCNTs@PBO shifts slightly to lower angles, indicating that the presence of PBO chains enhances the intercalation effect and expands the interlayer spacing of GnPs.^37,38 Furthermore, no diffraction peaks arising from structural features caused by aggregation or phase separation were observed among the components, indirectly indicating the uniform and stable dispersion of carbon nanofillers in the PBO matrix.

Fig. 3 displays transmission electron micrographs (TEM) of the GnPs&MWCNTs@PBO composites and carbon fillers. The pure polished MWCNTs and GnPs showed dense aggregation, as depicted in Fig. 3a and b, which can presumably be attributed to van der Waals forces that induce the formation of interconnected bundles and entanglements.^39,40 In contrast, Fig. 3c and d depict GnPs&MWCNTs@PBO as a loosely packed arrangement with uneven and rough surfaces. This can be attributed to the enhanced interaction between conjugated PBO backbones and MWCNTs/GnPs, facilitated by the formation of stable π–π interactions.^41,42 The elemental mapping and medium-resolution dark-field TEM images are shown in Fig. 3e and f. The non-uniform dispersion of C, which may be related to the spatial arrangement of the graphite, is illustrated in Fig. 3f. The results also demonstrate the incorporation of a new element N into the GnPs&MWCNTs@PBO. As observed from Fig. S1(a–d),† the PBO fibers thus facilitated GnPs and MWCNTs to develop a strong thermal conductivity network. The parallel alignment of sheet-like GnPs increases the contact area between MWCNTs and GnPs, enhancing the interconnected fillers' ability to facilitate heat transmission. Well-organized and aligned 3D thermal conductive networks formed by these fillers effectively reduce phonon scattering and promote efficient heat transfer. These findings further confirm the successful preparation of uniformly dispersed GnPs&MWCNTs@PBO composites.

3.2 Morphology analyses of GnPs&MWCNTs@PBO/PEEK composites

The 3D thermal conductive network structured GnPs&MWCNTs@PBO/PEEK composites were systematically investigated using an optical super depth-of-field microscope (OSM) and a scanning electron microscope (SEM). The OSM of the composites with a filler content of 5.84 vol% is shown in Fig. 4a₁–a₃. The filler phase was predominantly localized at the polymer interfaces rather than dispersed randomly. Simultaneously, the confined carbonaceous fillers established an interconnected thermal conduction network, distinct from the PEEK phase, facilitating efficient heat transfer (white fuzzy lines indicate the filler direction). Increasing the hybrid filler content improved both the width and integrity of this interconnected thermally conductive network. Fig. 4b₁–b₃ demonstrate that the 3D thermal conductive network structure becomes more firmly stacked in layers as the filler concentration increases from 2.85 vol% to 15.68 vol%. Fig. 4c₁–c₃ illustrate denser 3D thermal conductive networks with increasing filler concentration up to 19.81 vol%. Sheet-like GnPs primarily interconnected horizontally, while linear MWCNTs tend to form vertical connections from top to bottom (detailed images in Fig. S1(e and f)†). The increased density of the thermal conductive network achieved through hot-pressing techniques establishes stable channels and efficient routes for phonon transport and allows fillers with high aspect ratios to significantly enhance thermal conductivity.^43,44 Upon incorporation into PEEK composites, low-carbon fillers can only contribute to the formation of an “island” structure.^45,46 Optimization of the internal heat transmission network and thereby increase in the TC of the composites can be achieved through the increase of the contact area between the fillers. Fig. 4d₁–d₃ demonstrate that the incorporation of PBO significantly improved the dispersion of carbon fillers in the PEEK substrate. The π–π conjugation facilitated the uniform deposition of carbon fillers and aligned them along the PBO fibers, thereby enhancing the heat diffusion network and increasing the pathways for heat flow.^47,48

3.3 Thermal conductivity

The in-plane and through-plane TC characteristics of GnPs&MWCNTs@PBO/PEEK composites were investigated in detail. Fig. 5a and b show the TC as well as the thermal conductivity enhancement of the GnPs&MWCNTs@PBO/PEEK and GnPs&MWCNTs/PEEK composites. The in-plane TC curves of GnPs&MWCNTs/PEEK and GnPs&MWCNTs@PBO/PEEK composites display a similar tendency, with an initial gradual increase, followed by a dramatic growth upon the addition of fillers. In contrast, the through-plane TC curves of GnPs&MWCNTs@PBO/PEEK and GnPs&MWCNTs/PEEK showed a gradually increasing trend. Compared to the GnPs&MWCNTs/PEEK, the GnPs&MWCNTs@PBO/PEEK composites showed a remarkable increase in TC. The thermal conductivity enhancement factor (η) was calculated from eqn (5), where κ_PPGM and κ_P are the thermal conductivity of GnPs&MWCNTs@PBO/PEEK composites and PEEK, respectively; η is either through or in-plane thermal conductivity enhancement factor.⁴⁹ The thermal conductivity of pure PEEK was 0.23 W m⁻¹ K⁻¹. By contrast, the GnPs&MWCNTs@PBO/PEEK composites showed higher TC as the filler loading increased. The in and through-plane TC of the GnPs&MWCNTs/PEEK composites were measured to be 2.91 and 1.65 W m⁻¹ K⁻¹, respectively, at a filler fraction of 19.31 vol%. In contrast, the maximum values for the TC of GnPs&MWCNTs@PBO/PEEK were 22.17 and 3.19 W m⁻¹ K⁻¹, at the same filler content, which were 9540% and 1289% greater than those of pure PEEK. It was initially proven that the non-covalent bond modification approach and the assembly of a 3D thermal conduction network structure significantly improved TC.

	(5)

The phonon transport capability of the target composites is influenced by temperature. Therefore, the TC characteristics of the GnPs&MWCNTs@PBO/PEEK composites at different temperatures were systematically investigated. Fig. 5c displays the measured TC of the GnPs&MWCNTs@PBO/PEEK composites at elevated temperatures. The in-plane thermal conductivity at a filler loading of 19.31 vol% reaches 25.26 W m⁻¹ K⁻¹ at 100 °C, marking a 14% improvement compared to room temperature. This enhancement is primarily due to the elevated temperature, which promotes more efficient phonon transmission across lattice interfaces. However, as the temperature increases, the enhanced Umklapp scattering at the interfaces can diminish the intrinsic thermal conductivity of the fillers, thereby potentially reducing the overall thermal conductivity.⁵⁰ However, there is an overall increase in TC of all composites.⁵¹ The findings indicate that the MWCNTs&GnPs@PBO/PEEK composites remain highly promising for high-temperature applications. The Agari model was employed to fit the TC curves of GnPs&MWCNTs@PBO/PEEK composites facilitating further theoretical investigation. As illustrated in Fig. 5d, the TC of the GnPs&MWCNTs/PEEK and GnPs&MWCNTs@PBO/PEEK composites increased nonlinearly as the filler content increased. The Agari model is described in eqn (6), where TC of the composites, polymer matrix, MWCNTs, and GnPs is respectively denoted as λ, λ₁, λ₂, and λ₃. V₂ and V₃ are the ratios of the volume of two fillers to the volume of the total filler, and C₁ is the factor influencing the particle size and crystallinity of the polymer. Additionally, C₂ and C₃ are the free factors that can form thermal chains, with values lying between 0 and 1, and φ denotes the volume fraction of the filler.^52,53 C₃ was set to 1 to avoid the over-referencing phenomenon caused by the close relationship between the C₂ and C₃ fitting parameters. As the value approaches 1, the formation of thermal chains becomes more feasible, resulting in a notable improvement in thermal conductivity. Given the significant interdependence between C₂ and C₃, it is inferred that GnPs can establish a fully interconnected thermal conductive network, hence assigning a value of 1 to C₃. The parameter C₂, ranging from 0 to 1, reflects the degree of independence of the thermal chain constituents. In the GnPs&MWCNTs@PBO/PEEK composites, the in-plane and through-plane C₂ values measured 1.783 and 1.401, respectively, significantly exceeding 1. This indicates a robust filler network formation. Additionally, these values surpass those of randomly blended composites (C₂: 0.913 and 0.829), confirming the development of additional thermal chains and the establishment of a more efficient TC network. The data comparison highlights a significant finding: GnPs&MWCNTs@PBO/PEEK composites exhibit a propensity to form a highly conductive pathway characterized by enhanced physical contact. The 3D conductive network structure enhances TC by establishing efficient pathways for heat transfer, thereby increasing the average phonon propagation distance and minimizing phonon scattering.^54,55

lgλ = ϕ(V2C2lgλ2 + V3C3lgλ3 + …) + (1 − ϕ)lg(V1C1λ1)	(6)

Interfacial thermal resistance is pivotal in influencing the performance of thermally conductive composites and their practical applications. It essentially comprises the ITR observed at interfaces between filler–polymer matrix and filler–filler interfaces. Minimizing ITR enhances the TC of the composites.^56,57 The presence of a 3D thermally conductive network structure in the GnPs&MWCNTs@PBO/PEEK composites has been confirmed. The TC of the composites was studied using the Foygel model, as described by eqn (7)–(9). In these equations, K represents the TC of the composites, and K_m indicates the TC of the polymer matrix. K₀ represents a prefactor that is determined by the composite filler made up of GnPs and MWCNTs, β is the conductivity index, and V_C and V_f indicate the critical percolation and volume fraction values of the composite filler, respectively. L denotes the particle size of GnPs and MWCNTs, which is approximately 8 μm in this investigation. R_C and R_it are the ITR of the filler–filler and filler–matrix interfaces, respectively. S represents the average overlapping area between the MWCNTs/GnPs thermally conductive fillers.^58,59 The GnPs or MWCNTs surface is estimated to contribute 1/100th of its area to heat conduction, with a value of around 2.64 × 10⁻¹² m² K W⁻¹.

	(7)

	(8)

Rit = Rc × S	(9)

The fitting ITR curves of the GnPs&MWCNTs/PEEK and GnPs&MWCNTs@PBO/PEEK composites for the TC are displayed in Fig. 5e (additional details have been presented in Table S2†). The V_C values of the filler for GnPs&MWCNTs@PBO/PEEK were calculated by fitting the TC using the tangent technique. In-plane and through-plane fitted values of V_C for GnPs&MWCNTs@PBO/PEEK were 6.69 vol% and 13.49 vol%, whereas those for GnPs&MWCNTs/PEEK were 6.34 vol% and 13.03 vol%, respectively (details in Fig. S2†). The ITR between two compound fillers was fitted and determined using the Foygel model eqn (7)–(9) in combination with V_C. The ITR values for GnPs&MWCNTs@PBO/PEEK composites were measured to be 4.58 × 10⁻⁸ m² K W⁻¹ (in-plane) and 2.09 × 10⁻⁸ m² K W⁻¹ (through-plane), compared to 1.63 × 10⁻⁶ m² K W⁻¹ (in-plane) and 1.60 × 10⁻⁶ m² K W⁻¹ (through-plane) for GnPs&MWCNTs/PEEK composites. These results indicate a substantial reduction in ITR for GnPs&MWCNTs@PBO/PEEK composites, highlighting the importance of establishing a 3D thermal conductive network that facilitates efficient phonon transport pathways. The decrease in ITR can be attributed to three factors: firstly, the presence of intermolecular forces like π–π interactions. This is contribute to the fact that carbon nanomaterials well-deposited on the surface of PBO fiber balls, which are composed of PBO ultra-short fibers arranged in a spherical structure. This three-dimensional network formed by carbon nanomaterials and PBO fiber balls enhances the probability of contact between MWCNTs and GnPs, facilitating the establishment of interconnected thermal pathways. This results in a significant enhancement of phonon transfer efficiency and overall thermal conductivity. Secondly, π–π interactions between PBO and carbon nanomaterials improve the dispersion of thermally conductive fillers within the polymer, effectively reducing heat transfer resistance caused by filler agglomeration and promoting rapid heat flow.²⁵ Finally, the facile orientation of carbon fillers along PBO fibers leads to a dense and continuous thermal network, providing a stable pathway for heat transfer and thereby reducing interfacial thermal resistance.²⁹ Fig. 5f demonstrates the remarkable advancement in thermal conductivity of the GnPs&MWCNTs@PBO/PEEK composites compared to previous studies (additional details are provided in Fig. S3†).

Samples with identical filler contents and dimensions (20 × 20 × 2 mm) were simultaneously placed on a hot plate maintained at 130 °C to assess the thermal conductivity of composites. The surface temperatures of the composites GnPs&MWCNTs/PEEK and GnPs&MWCNTs@PBO/PEEK were monitored using an infrared camera, as illustrated in Fig. 6a. The GnPs&MWCNTs@PBO/PEEK composites achieved a surface temperature of 67.8 °C within 10 s. After a passage of 60 s, the composites had a surface temperature of 125.8 °C, surpassing that of the GnPs&MWCNTs/PEEK composites by 38.4 °C. Infrared thermal images are depicted in Fig. S4.† The heat transmission characteristics of the two composites in a parallel-to-the-plane direction during a typical heating process were analyzed using finite element simulations in COMSOL Multiphysics 6.1. The ESI† provides specific information regarding the parameters employed in the study. The temperature distribution of GnPs&MWCNTs/PEEK and GnPs&MWCNTs@PBO/PEEK composites with similar filler content is shown in Fig. 6b. The temperature distribution gives insights into the heat transfer rate and conduction mechanisms of the composites. The GnPs&MWCNTs@PBO/PEEK composites featured extensive hot regions and a few cold regions, indicating rapid and uniform heat transfer from left to right. In contrast, the GnPs&MWCNTs/PEEK composites showed a significantly reduced thermal area compared to the GnPs&MWCNTs@PBO/PEEK composites, suggesting lower heat transfer efficiency. The in-plane heat transfer in GnPs&MWCNTs@PBO/PEEK composites was faster than that in GnPs&MWCNTs/PEEK composites. The incorporation of PBO facilitated the formation of a 3D filler network structure, enhancing the heat transfer process within the matrix and establishing a strong thermal diffusion network.

The fitting results from the Agari model confirm the improved TC of GnPs&MWCNTs@PBO/PEEK composites. Low packing loads result in limited interfacial contact and minimal interaction between the fillers. PBO serves to bridge the isolated fillers, enhancing heat flow conveyance within the matrix and establishing a favorable heat diffusion network. Compared to control group composites with randomly blended fillers, the incorporation of PBO promotes a more uniform distribution within the polymer matrix, reducing the formation of isolated “island” structures (as depicted in Fig. S5a†), thereby significantly enhancing the thermal conductivity of the polymer matrix. Fig. S5b† illustrates that at a filler content of 19.31 vol%, the thermal conductivity of GnPs&MWCNTs@PBO/PEEK composites surpasses that of GnPs/MWCNTs@PBO/PEEK composites. This improvement is attributed to the synergistic effect of hybrid fillers with high aspect ratios, which establish continuous and interconnected thermally conductive pathways, referred to as the “nano-microbridge” effect. The effect reduces the contact area between the filler and the polymer, thereby promoting rapid heat flux while minimizing phonon scattering.^60–62 Furthermore, a strong π–π interaction was observed between the conjugated PBO backbones and carbon fillers, which effectively reduced the phonon mismatch effect. This resulted in enhanced interfacial compatibility and a reduction in the ITR between the fillers and polymer.^48,63 These methods effectively reduced the adverse effects of ITR at the filler–polymer interface. The fabrication of a 3D network structure with PBO facilitates a rapid and unobstructed transmission pathway for phonons, as depicted in Fig. 6c.

3.4 EMI shielding performance and electrical conductivity

The electrical conductivity has a significant impact on the EMI shielding capabilities of the conductive composites. There was a gradual increase in electrical conductivity followed by a substantial and abrupt rise from 407.6 to 3375.4 S m⁻¹ in Fig. 7a. This increase was more than 17 orders of magnitude greater compared to pure PEEK, which typically exhibits a conductivity in the range of 10⁻¹⁴ S m⁻¹. The conductive percolation threshold of the composites was simulated according to the classical Kirkpatrick-Zallen eqn (S1).†^64,65 The simulation confirmed the formation of an initial conductive network, enabling the passage of electrons through the insulating layers and facilitating carrier mobility at a rate below 1%. An analysis of the fitted image in Fig. 7b shows that a sudden and significant increase in conductivity occurred at a filler loading of 0.76 vol%, confirming the occurrence of percolation phenomena.⁶⁶ The fitted value of t = 2.8 (t = 1.9 denotes the construction of a 3D conductive network) demonstrated that the fillers successfully constructed a highly effective conductive network.^67–70 It was observed that as the network structure progressively improves, the energy barrier for electrons to traverse the insulating layer decreases, thereby facilitating carrier migration.

The wave guide method was utilized to assess the EMI SE of composites. Fig. 7c displays the variation in SE_T, SE_R, and SE_A effectiveness of GnPs&MWCNTs@PBO/PEEK composites at 8.2 GHz, as the filler content increased from 2.85 vol% to 19.31 vol%. Both SE_T and SE_A show progressive strengthening as the filler content increases. SE_A provides 80–90% of the contribution in relation to SE_R. However, the value of SE_R demonstrated a slight upward trend. Fig. 7d shows the power coefficients for EMI shielding absorption (A), reflection (R), and transmission (T) of the composites. The value of R consistently exceeds that of A and T, indicating that the primary shielding mechanism of the composites is R.^71,72 The three-dimensional carbon network structure generates multiple conductive surfaces, thereby extending the transmission paths of electromagnetic waves and increasing their scattering through numerous internal reflections. This process facilitates a distinctive “absorption-reflection-reabsorption” mechanism, which is essential in applications aimed at reducing secondary electromagnetic interference.⁷³ Fig. 7e and f display the SE_T and SE_A of GnPs&MWCNTs@PBO/PEEK composites at different filler concentrations within the X-band (8.2–12.4 GHz) frequency range. The EMI SE of the composites exceeded 20 dB when the filler loading reached 2.85 vol%. The SE_T value reaches up to 128.4 dB at 19.31 vol% filler, resulting in a shielding efficiency of 99.99999999993%, which meets the minimum requirement of 100 dB for military-shielded rooms.^74,75

Fig. 7g is a schematic representation of potential decay mechanisms occurring during the electromagnetic shielding process. According to the principle of impedance mismatch, EWs experience reflection loss when encountering interfaces between fillers and the matrix. This results in a significant impedance mismatch between the composites and the surrounding air.⁷⁰ The incoming EWs are categorized into three sections, with one section being immediately reflected. Additional EWs infiltrate the conductive filler network by penetrating the polymer matrix. Furthermore, the strong interaction between the electric dipoles of fillers and the EWs is responsible for dielectric loss and dielectric relaxation.^76–78 The EWs undergo repeated reflection and conversion into heat until they are completely dissipated. Any remaining EWs are then transferred to the subsequent point within the filler network, initiating a cycle of further reflections and transmissions.^79,80 PBO facilitates the formation of a stable conductive network with carbon nanomaterials, enabling the conductive fillers to induce intense reflection and scattering of incoming electromagnetic waves (EWs). Moreover, this conductive interface causes multiple reflections of EWs, widening transfer channels and increasing energy consumption within the composites. Additionally, non-covalent modification of carbon fillers and PBO enhances interface compatibility, reducing the occurrence of defects and diminishing EW transmission. The combined effects of absorption and reflection synergistically contribute to the exceptional EMI shielding characteristics of the GnPs&MWCNTs@PBO/PEEK composites.

3.5 Thermal stability performance

The thermal stability of polymers and the GnPs&MWCNTs@PBO/PEEK composites was assessed using DSC, TGA, and CTE, as illustrated in Fig. 8. Both pure PEEK and GnPs&MWCNTs@PBO/PEEK composites demonstrated distinct melting peaks. As indicated in Fig. 8a, the T_c and T_m of pure PEEK were measured to be 176 °C and 334 °C, respectively.⁸¹ As the filler content increased in GnPs&MWCNTs@PBO/PEEK composites, the morphology of the cold crystallization peak transitioned from a sharp and narrow shape to a broader and more rounded shape, indicating a reduction in the PEEK content. The T_m values of the resulting composites ranged from 324 °C to 350 °C. The minor shift in T_m can be attributed to the incorporation of thermoplastic PBO, which has a negligible effect on the melting process. The onset temperature (T₁₀) values of the GnPs&MWCNTs@PBO/PEEK composites exceeded 608 °C, demonstrating exceptional thermal stability and promising potential for applications in harsh environments, as illustrated in Fig. 8b. The residual carbon content of the MWCNTs&GnPs@PBO/PEEK composites demonstrated a steady increase as high-temperature-resistant fillers were continuously added. The enhanced “heat resistance index” further substantiates the outstanding thermal stability of MWCNTs&GnPs@PBO/PEEK composites, rendering them suitable for research applications requiring high temperatures (details in Table S3†).⁸² As illustrated in Fig. 8c, the GnPs&MWCNTs@PBO/PEEK composites had a CTE of 2.12 × 10⁻⁵ °C⁻¹ at 19.31 vol% filler content, which was below 50% of the value of the pure polymer. The CTE undergoes a gradual decline as the filler content increases. The addition of PBO and the establishment of a 3D thermal network serve to decrease the average distance between polymer molecules and restrict their movement.⁸³ Furthermore, employing composites with a 3D network structure leverages the inherent thermal stability of the PBO matrix, thereby emphasizing the exceptional thermal stability of these materials.

3.6 Thermal management capability

Fig. 9 illustrates the practical uses of 3D network structured GnPs&MWCNTs@PBO/PEEK composites for effective heat dissipation. Furthermore, random blend control composites containing 5.84 vol% filler were evaluated for comparison. A 20 W LED, measuring approximately 40 mm × 40 mm × 3 mm, was employed as the heat source. It was affixed to the test sample using thermally conductive silicone rubber, and the temperature of its surface was continuously monitored using an IR camera. The LED chips in all devices were heated to an initial temperature of 125 °C and then the power was turned off. The heat produced by the LED chips was effectively dispersed when GnPs&MWCNTs/PEEK and GnPs&MWCNTs@PBO/PEEK composites were employed as heat spreaders. The GnPs&MWCNTs@PBO/PEEK composites with 5.84 vol% filler within a three-dimensional network structure undergo cooling from 125 °C and record a temperature decrease to 77.3 °C within 30 seconds, representing a 47.7 °C reduction. In comparison, the randomly blended GnPs&MWCNTs/PEEK composites with an equivalent filler content exhibit a cooling from 125 °C to 84.2 °C at the same time, achieving a temperature decrease of 40.8 °C. Such a marked disparity in cooling efficacy highlights the advanced thermal dissipation properties of composites with a 3D network structure. In addition, the GnPs&MWCNTs@PBO/PEEK composites with a 19.31 vol% filler content were cooled from 125 °C to 33.3 °C in 100 s, indicating that the composites are effective in dissipating heat and have the potential to be applied in high-performance electronic devices.

4 Conclusion

In summary, novel 3D network structured GnPs&MWCNTs@PBO/PEEK composites were prepared to achieve high TC and effective EMI shielding performance. PBO fiber balls were employed to construct a three-dimensional thermal conductivity network structure, which enhances interfacial contact and reduces interfacial thermal resistance. Moreover, the non-covalent modification of PBO improves the interface compatibility between PEEK and fillers, minimizing interfacial phonon scattering resulting from interfacial voids. Consequently, the in-plane thermal conductivity of GnPs&MWCNTs@PBO/PEEK composites increases to 22.17 W m⁻¹ K⁻¹ at a filler content of 19.31 vol%, representing a 9540% increase over PEEK. The results from Agari and Foygel modeling demonstrate that the composites showed exceptional anisotropic TC and low ITR. Furthermore, the electrical conductivity of GnPs&MWCNTs@PBO/PEEK reached 3375.4 S m⁻¹, and the EMI SE was up to 128.4 dB in the X-band, attributed to its 3D network structure. This study represents a novel approach to creating efficient and environmentally friendly dual-functional materials for electromagnetic interference shielding and heat dissipation. These materials hold promise for applications in wearable devices, precision electronics, and aerospace technologies.

Data availability

Date available on request from the authors.

Author contributions

Yageng Bai: conceptualization, methodology, writing-original draft. Hongxia Qian: writing-review & editing. Xueling Cao: resources. Fengyu Wen: methodology&project administration. Yashu He: formal analysis & visualization. Jierun Ma: data curation. Lin Cheng: formal analysis. Yifan Wang: resources. Haoyuan Tan: visualization. Yuxuan Gu: formal analysis & visualization. Pengbo Lian: investigation & software. Rui Chen: writing-review & editing. Jianxin Mu: supervision.

Conflicts of interest

There are no conflicts of interest to declare.

Acknowledgements

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Notes and references

1. L. Chen, T. Liu, X. Wang, Y. Wang, X. Cui, Q. Yan, L. Lv, J. Ying, J. Gao, M. Han, J. Yu, C. Song, J. Gao, R. Sun, C. Xue, N. Jiang, T. Deng, K. Nishimura, R. Yang, C. Lin and W. Dai, Adv. Mater., 2023, 35, 2211100–2211113 CrossRef CAS PubMed .
2. S. Cheng, X. Guo, W. Cai, Y. Zhang and X. Zhang, J. Mater. Chem. A, 2024, 12, 8640–8662 RSC .
3. R. Selzer and K. Friedrich, Composites, Part A, 1997, 28A, 595–604 CrossRef CAS .
4. Q. Xu, W. Xu, Y. Yang, X. Yin, C. Zhou, J. Han, X. Li, Y. Shang and H. Zhang, Addit. Manuf., 2022, 55, 102852–102863 CAS .
5. M. Niu, Z. Zhao, B. Wang, C. Yu, M. Li, J. Hu, L. Zhu, X. Hao, S. Wan, M. Yue, W. Xuan, Q. Lu, W. Cao, K. Chen and Q. Wang, J. Mater. Chem. A, 2023, 11, 23787–23797 RSC .
6. Y. Lin, P. Li, W. Liu, J. Chen, X. Liu, P. Jiang and X. Huang, ACS Nano, 2024, 18, 3851–3870 CrossRef CAS PubMed .
7. R. Chen, Q. He, X. Li, F. Wen, L. Cheng, L. Li, Y. He, X. Liu and J. Mu, Chem. Eng. J., 2022, 431, 134049–134061 CrossRef CAS .
8. Z. Yao, J. Song, Y. Lu, R. Yang, Y. Zhang and K. Zhang, J. Mater. Chem. A, 2024, 12, 9113–9123 RSC .
9. F. Wen, S. Li, R. Chen, Y. He, L. Li, L. Cheng, J. Ma and J. Mu, Polymers, 2022, 14, 1328–1345 CrossRef CAS PubMed .
10. Z. Peng, Q. Lv, J. Jing, H. Pei, Y. Chen and E. Ivanov, Composites, Part B, 2023, 251, 110491–110491501 CrossRef CAS .
11. E. A. O'Rear, S. Onthong and T. Pongprayoon, Nanomaterials, 2023, 14, 80–100 CrossRef PubMed .
12. N. Zhang, D. Huang, X. Chen, C. Ye, B. Li, S. Zhu, Z. Fan, H. Liu and J. Liu, Composites, Part B, 2023, 252, 110509–110519 CrossRef CAS .
13. C. Li, X. Cao, Y. Tong, Z. Yang, D. Gao, Y. Ru and G. He, ACS Appl. Polym. Mater., 2022, 4, 7152–7161 CrossRef CAS .
14. W. Yang, Y. Wang, Y. Li, C. Gao, X. Tian, N. Wu, Z. Geng, S. Che, F. Yang and Y. Li, Composites, Part B, 2021, 224, 109168–109177 CrossRef CAS .
15. M. Feng, Y. Pan, M. Zhang, Q. Gao, C. Liu, C. Shen and X. Liu, Compos. Sci. Technol., 2021, 206, 108666–108674 CrossRef CAS .
16. S. Cheng, X. Guo, P. Tan, M. Lin, J. Cai, Y. Zhou, D. Zhao, W. Cai, Y. Zhang and X. Zhang, Composites, Part B, 2023, 264, 110916–110924 CrossRef CAS .
17. H. Jia, Q. Kong, X. Yang, L. Xie, G. Sun, L. Liang, J. Chen, D. Liu, Q. Guo and C. Chen, Carbon, 2021, 171, 329–340 CrossRef CAS .
18. D. Liu, Q. Kong, H. Jia, L. Xie, J. Chen, Z. Tao, Z. Wang, D. Jiang and C. Chen, Carbon, 2021, 183, 216–224 CrossRef CAS .
19. C. Lin, J. Li, Y. Chen, J. Chen, C. Cheng and C. Chiu, ACS Omega, 2022, 7, 45697–45707 CrossRef CAS .
20. X. Chen, K. Wu, Y. Zhang, D. Liu, R. Li and Q. Fu, Adv. Mater., 2022, 34, 2206088–2206097 CrossRef CAS .
21. L. Tang, Y. Tang, J. Zhang, Y. Lin, J. Kong, K. Zhou and J. Gu, Sci. Bull., 2022, 67, 2196–2207 CrossRef CAS .
22. M. Hao, X. Qian, Y. Zhang, J. Yang, C. Li, H. Gong, X. Wang, P. Wang, L. Liu and Y. Huang, Compos. Sci. Technol., 2023, 231, 109800–109812 CrossRef CAS .
23. Y. Liu, W. Zou, M. Yang, H. Luo, S. Yang, J. Xu and N. Zhao, Adv. Funct. Mater., 2023, 33, 2303561–2303570 CrossRef CAS .
24. T. Sun, W. Cao, K. Zhao, X. Wang, Z. Wang, G. Gao, Z. Ye, K. Zhao, Z. Su, B. Dai, M. Zhang, J. Han and J. Zhu, Chem. Eng. J., 2023, 474, 145916–145929 CrossRef CAS .
25. Z. Yu, S. Wu, C. Li, Y. Xiao, J. Liu and B. Zhang, ACS Appl. Nano Mater., 2022, 5, 18247–18255 CrossRef CAS .
26. L. Wang, Z. Ma, Y. Zhang, H. Qiu, K. Ruan and J. Gu, Carbon Energy, 2022, 4, 200–210 CrossRef CAS .
27. R. Chen, X. Li, J. Ma, L. Cheng, F. Wen, L. Li, Y. Bai, Y. He and J. Mu, Composites, Part A, 2023, 173, 107633–107645 CrossRef CAS .
28. P. Wang, T. Mai, W. Zhang, M. Qi, L. Chen, Q. Liu and M. Ma, Small, 2023, 20, 2304914–2304929 CrossRef .
29. L. Tang, K. Ruan, X. Liu, Y. Tang, Y. Zhang and J. Gu, Nano-Micro Lett., 2023, 16, 38–53 CrossRef .
30. B. Song, Z. Liu, L. Chen, L. Ma and Y. Huang, Polym. Compos., 2022, 43, 454–466 CrossRef CAS .
31. M. Hao, Z. Hu, Y. Zhang, X. Qian, L. Liu, J. Yang, X. Wang, J. Zhi, Y. Huang and X. Shi, Polym. Degrad. Stab., 2022, 199, 109896–109909 CrossRef CAS .
32. W. Zhao, J. Kong, H. Liu, Q. Zhuang, J. Gu and Z. Guo, Nanoscale, 2016, 8, 19984–19993 RSC .
33. S. Wang, D. Feng, Z. Zhang, X. Liu, K. Ruan, Y. Guo and J. Gu, Chin. J. Polym. Sci., 2024, 42, 897–906 CrossRef CAS .
34. M. Qin, X. Zhang, J. Zhu, Y. Yang, Z. Ti, Y. Shen, X. Wang, X. Liu and Y. Zhang, J. Mater. Chem. A, 2023, 11, 10612–10627 RSC .
35. P. Hu, F. Wu, B. Ma, J. Luo, P. Zhang, Z. Tian, J. Wang and Z. Sun, Adv. Mater., 2024, 36, 2310023–2310035 CrossRef CAS PubMed .
36. S. Wang, D. Feng, Z. Zhang, X. Liu, K. Ruan, Y. Guo and J. Gu, Chin. J. Polym. Sci., 2024, 42, 897–906 CrossRef CAS .
37. Z. Yan, X. Zhang, Y. Gao, Z. Kong, X. Ma, Q. Gou, H. Liang, X. Cai, H. Tan and J. Cai, J. Appl. Polym. Sci., 2023, 140, e54541–e54551 CrossRef CAS .
38. B. Xue, S. Yang, X. Sun, L. Xie, S. Qin and Q. Zheng, J. Mater. Chem. A, 2020, 8, 14506–14518 RSC .
39. Y. Li, W. Li, Z. Wang, P. Du, L. Xu, L. Jia and D. Yan, Carbon, 2023, 211, 118096–118104 CrossRef CAS .
40. L. Kong, Y. Zhu, P. J. Williams, M. Kabbani, F. R. Brushett and J. L. M. Rupp, J. Mater. Chem. A, 2024, 12, 4299–4311 RSC .
41. A. Navidfar and L. Trabzon, Composites, Part B, 2019, 176, 107337–107348 CrossRef CAS .
42. J. Wang, X. Jin, H. Wu and S. Guo, Carbon, 2017, 123, 502–513 CrossRef CAS .
43. K. Ruan, X. Shi, Y. Zhang, Y. Guo, X. Zhong and J. Gu, Angew. Chem., Int. Ed., 2023, 62, e202309010–e202309021 CrossRef CAS .
44. H. Yu, P. Guo, M. Qin, G. Han, L. Chen, Y. Feng and W. Feng, Compos. Sci. Technol., 2022, 222, 109406–109414 CrossRef CAS .
45. B. Ghosh, F. Xu and X. Hou, J. Mater. Sci., 2021, 56, 10326–10337 CrossRef CAS .
46. S. Gul, S. Arican, M. Cansever, B. Beylergil, M. Yildiz and B. S. Okan, ACS Appl. Polym. Mater., 2022, 5, 329–341 CrossRef .
47. H. Duan, C. Wang, Y. Yi, X. Mu, H. Ding, Z. Bi, Y. Hu and B. Yu, Chem. Eng. J., 2024, 483, 149302–149311 CrossRef CAS .
48. X. Chen, K. Wu, Y. Zhang, D. Liu, R. Li and Q. Fu, Adv. Mater., 2022, 34, 2206088–2206097 CrossRef CAS PubMed .
49. Y. Lin, Q. Kang, H. Wei, H. Bao, P. Jiang, Y.-W. Mai and X. Huang, Nano-Micro Lett., 2021, 13, 180–194 CrossRef CAS PubMed .
50. H. Duan, C. Wang, Y. Yi, X. Mu, H. Ding, Z. Bi, Y. Hu and B. Yu, Chem. Eng. J., 2024, 483, 149302–149308 CrossRef CAS .
51. Y. Wu, Z. Chen, P. Nan, F. Xiong, S. Lin, X. Zhang, Y. Chen, L. Chen, B. Ge and Y. Pei, Joule, 2019, 3, 1276–1288 CrossRef CAS .
52. D. Pan, G. Yang, H. M. Abo-Dief, J. Dong, F. Su, C. Liu, Y. Li, B. Xu, V. Murugadoss, N. Naik, S. M. El-Bahy, Z. M. El-Bahy, M. Huang and Z. Guo, Nano-Micro Lett., 2022, 14, 118–137 CrossRef CAS .
53. M. Dong, G. Hou, J. Zhang, L. Liu, G. Liang, X. Hao, Y. Guo and M. Wang, Composites, Part B, 2022, 242, 110033–110044 CrossRef CAS .
54. H. He, W. Peng, J. Liu, X. Chan, S. Liu, L. Lu and H. L. Ferrand, Adv. Mater., 2022, 34, 2205120–2205131 CrossRef CAS PubMed .
55. S. Yelishala, C. Murphy and L. Cui, J. Mater. Chem. A, 2024, 12, 10614–10658 RSC .
56. X. Wang, J. Zhou and S. Yang, Chem. Eng. J., 2022, 447, 137508–137519 CrossRef CAS .
57. F. Xu, D. Bao, Y. Cui, Y. Gao, D. Lin, X. Wang, J. Peng, H. Geng and H. Wang, Adv. Compos. Hybrid Mater., 2021, 5, 2235–2246 CrossRef .
58. B. Wan, X. Li, X. Zeng and J. Zha, Compos. Commun., 2024, 45, 101803–101810 CrossRef .
59. J. Yang, X. Shen, W. Yang and J. K. Kim, Prog. Mater. Sci., 2023, 133, 101054–101102 CrossRef CAS .
60. X. Jiang, C. Wang, G. Li, Y. Yu and X. Yang, Composites, Part B, 2024, 273, 111238–111150 CrossRef CAS .
61. J.-U. Jang, S. H. Lee, J. Kim, S. Y. Kim and S. H. Kim, Composites, Part B, 2021, 222, 109072–109081 CrossRef CAS .
62. Q. Zhang, W. Tian, J. Zhou, Y. Li and L. Qi, J. Magnesium Alloys, 2023, 21, 31–41 Search PubMed .
63. M. Hao, Z. Hu, Y. Huang, X. Qian, Z. Wen, X. Wang, L. Liu, F. Lu and Y. Zhang, Composites, Part B, 2022, 229, 109468–109481 CrossRef CAS .
64. L. Wang, P. Song, C. Lin, J. Kong and J. Gu, Research, 2020, 2020, 4093732 CAS .
65. L. Wang, H. Wang, B. Li, Z. Guo, J. Luo, X. Huang and J. Gao, J. Mater. Sci., 2020, 55, 11727–11738 CrossRef CAS .
66. L. Yan, T. Xiong, Z. Zhang, H. Yang, X. Zhang, Y. He, J. Bian, H. Lin and D. Chen, J. Polym. Res., 2021, 28, 350–362 CrossRef CAS .
67. K. Bilisik and M. Syduzzaman, Polym. Compos., 2021, 42, 1670–1697 CrossRef CAS .
68. Y. Liu, H. He, G. Tian, Y. Wang, J. Gao, C. Wang, L. Xu and H. Zhang, Compos. Sci. Technol., 2021, 214, 108956–108966 CrossRef CAS .
69. Y. Shi, J. Li, Y. Tan, Y. Chen and M. Wang, Compos. Sci. Technol., 2019, 170, 70–76 CrossRef CAS .
70. G. Wang, L. Wang, L. H. Mark, V. Shaayegan, G. Wang, H. Li, G. Zhao and C. B. Park, ACS Appl. Mater. Interfaces, 2017, 10, 1195–1203 CrossRef PubMed .
71. Y. Li, Y. Chen, X. He, Z. Xiang, T. Heinze and H. Qi, Chem. Eng. J., 2022, 431, 133907–133918 CrossRef CAS .
72. Y. Zhan, Y. Cheng, N. Yan, Y. Li, Y. Meng, C. Zhang, Z. Chen and H. Xia, Chem. Eng. J., 2021, 417, 129339–129352 CrossRef CAS .
73. W. Cao, F. Chen, Y. Zhu, Y. Zhang, Y. Jiang, M. Ma and F. Chen, ACS Nano, 2018, 12, 4583–4593 CrossRef CAS PubMed .
74. D. Munalli, G. Dimitrakis, D. Chronopolous, S. Greedy and A. Long, Composites, Part B, 2019, 173, 106906–106918 CrossRef .
75. H. Duan, H. Zhu, J. Gao, D. Yan and Z. Li, J. Mater. Chem. A, 2020, 8, 9146–9159 RSC .
76. B. Deng, Z. Xiang, J. Xiong, Z. Liu, L. Yu and W. Lu, Nano-Micro Lett., 2020, 12, 55–71 CrossRef CAS .
77. G. Zhang, H. Wang, W. Xie, S. Zhou, Z. Nie, G. Niwamanya, Z. Zhao and H. Duan, J. Mater. Chem. A, 2024, 12, 5581–5605 RSC .
78. Y. Zhang, K. Ruan, K. Zhou and J. Gu, Adv. Mater., 2023, 35, 2211642–2211655 CrossRef CAS .
79. X. Hao, D. Li, X. Peng, W. Lan and C. Liu, Chem. Eng. J., 2024, 479, 147681–147690 CrossRef CAS .
80. J. Yang, Y. Chen, C. Liu, H. Wang, X. Yan, X. Chai, Z. Chen, Y. Xia, H. Gao, H. Zhang and X. Liao, J. Mater. Res. Technol., 2023, 23, 5115–5126 CrossRef CAS .
81. X. Liu, H. Zhou, Z. Wang, X. Han, Z. Zhao, Y. Guo, W. Liu, J. Wang and T. Zhao, Compos. Sci. Technol., 2022, 220, 109289–109298 CrossRef CAS .
82. Y. Han, K. Ruan, X. He, Y. Tang, H. Guo, Y. Guo, H. Qiu and J. Gu, Angew. Chem., Int. Ed., 2024, 63, e202401538–e202401545 CrossRef CAS PubMed .
83. M. F. Arif, H. Alhashmi, K. M. Varadarajan, J. H. Koo, A. J. Hart and S. Kumar, Composites, Part B, 2020, 184, 107625–107635 CrossRef CAS .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ta03002h

This journal is © The Royal Society of Chemistry 2024