Synthesis of graphene oxide-mediated high-porosity Ni/C aerogels through topological MOF deformation for enhanced electromagnetic absorption and thermal management

Abstract

Magnetic metal/carbon composite aerogels with high porosity are emerging as excellent functional materials for electromagnetic (EM) attenuation and thermal management because their unique structural advantage not only intensifies EM energy consumption, but also obstructs heat conduction. Herein, we employed Ni-MOFs as a precursor, induced its topological deformation in the presence of graphene oxide (GO) nanosheets, and finally harvested composite aerogels composed of GO nanosheets and Ni-MOF fibers. It was found that GO nanosheets can resist the thermal shrinkage of composite aerogels during high-temperature pyrolysis and endow the resulting Ni/C@rGO composite aerogels (NCGCAs) with very high porosity (96.4%) and extremely low density (0.04 g cm⁻³). Electromagnetic analyses revealed that changes in structure and composition effectively reinforce the overall loss capability of NCGCAs, and thus generate obvious improvements in broadband response with a thin thickness (12.8–18.0 GHz, 1.5 mm). The good EM absorption performance of NCGCAs is significantly superior to those of many reported Ni/C aerogels. In addition, the structural advantage of NCGCAs demonstrates their potential application in thermal management. This GO-mediated MOF deformation is a promising strategy for the formation of high-porosity magnetic metal/carbon composite aerogels, which may bring some unexpected benefits in different fields.

---

Introduction

The burgeoning use of electronic devices in the present era of wireless communication have brought great convenience to our daily life, with excessive electromagnetic (EM) radiation and interference from ubiquitous EM waves posing huge threats to human health.^1–3 Accordingly, EM protection is becoming a research hot spot attracting global attention. As the key media to attenuate EM waves, EM wave absorbing materials (EWAMs) with wide-band response and strong absorption have gradually developed as one of the most important functional materials in the past decade.^4–7

Magnetic metal/carbon composites (MCCs) have been dominant candidates for high-performance EWAMs owing to their particular advantages (i.e., dual loss mechanisms and the magneto-electric coupling effect) and have thus received widespread attention.^8–11 To better leverage the synergistic effect of magnetic and carbon components, numerous efforts have been devoted to improving the distribution of these components in MCCs with different methods.^12–14 Among them, MOF transformation has rapidly evolved into a common strategy for preparing MCCs because the periodic arrangements of metal nodes and organic ligands in crystalline MOFs have laid a solid foundation for the uniform dispersion of magnetic metal particles, favorable for the improvement of the magneto-electric coupling effect.^15–18 However, these MOF-derived MCCs still suffer from some undesirable drawbacks, especially, because their final compositions and microstructures are extremely dependent on their precursors. This situation suggests that the EM properties of these MOF-derived MCCs may not reach a desirable state.^19,20 Thus, there is still much room to upgrade their EM absorption performance. Composition optimization can be easily attained by the introduction of additional EM components,^21–23 while microstructure regulation is still very difficult to achieve. To date, only several groups have successfully fabricated MOF-derived MCCs with profitable microstructure, e.g., hollow Co/C microspheres and porous FeCo/C aerogels, by directing the assembly process of MOF units.^24–28 Although some significant achievements have been made in these unique MOF-derived MCCs, it should be noted that the regulation of the assembly of MOF units is usually carried out under precise control, which hinders the popularization and further development of this strategy.

Recent studies on EWAMs have focused on MCCs with three-dimensional (3D) aerogel structures because their exceptionally large pore size and ultra-high porosity therein are more conducive to the transmission and multiple reflections of the EM wave than those with micro-/nano-scaled microstructures.^29–32 In an effort to integrate various advantages, many groups attempted to anchor MOFs crystals on the surface of carbon aerogels to generate MCCs with 3D aerogel structures.^33,34 For example, Wang et al. attached Fe-doped ZIF-67 on the surface of a GO aerogel, and finally obtained hierarchically porous Fe–Co/C composites.³⁵ They found that the rational composition and porous structural design not only induced diverse loss mechanisms, including magnetic loss, conductive loss, and interfacial polarization, but also improved the impedance matching and intensified multiple refections of the incident EM wave, and thus the final composite could produce good absorption characteristics. Of note is that such a post-treatment method cannot ensure uniform dispersion of MOFs crystals on the surface of the carbon aerogel, and the composition segregation will inevitably offset the advantages of the MOFs transformation to some extent. More recently, our group pioneered a new strategy for the preparation of Ni/C composite aerogels, where Ni-MOFs particles were firstly converted into Ni-MOFs aerogels through topological deformation rather than being directly pyrolyzed at high temperature.³⁶ In this way, the final Ni/C composite aerogels simultaneously possessed a 3D structure and uniformly dispersed Ni nanoparticles as expected, and thus they generated overall improvements on the EM absorption, hydrophobicity, and thermal management as compared with conventional MOF-derived Ni/C composites. Zhou et al. obtained NiCo/carbon aerogels from NiCo-MOFs with a very similar method, and the minimum reflection loss (RL) intensity and the maximum effective absorption bandwidth (EAB) of the optimal sample could reach −60.7 dB and 6.2 GHz, respectively.³⁷ These results clearly indicate that MOFs aerogels are more promising precursors for MCCs than conventional MOFs particles. However, we also notice that high-temperature pyrolysis always causes serious shrinkage in the macroscopic profile of aerogels. For instance, the shrinkage ratio from the Ni-MOFs aerogels to Ni/C composite aerogels can reach up to 86.9%.³⁶ The high shrinkage ratio means a loss in porosity. Thus, the density of the final product will be drastically increased, and more importantly, the contribution from the microstructure to EM absorption may be moderately depressed. Additionally, 3D structures often enhance the thermal management performance of adiabatic media because the high porosity will effectively obstruct heat transfer.^29,32 Therefore, the improvement on the 3D structures of carbon-based aerogels from MOFs aerogels can enhance their EM absorption performance and thermal management capability at the same time. Meanwhile, how to suppress high shrinkage during the pyrolysis process is a task of great significance.

Since the successful preparation of atomically thin carbon films, graphene and graphene oxide (GO) have quickly become topics of great research interest in many fields. Very interestingly, they are also regarded as qualified structural additives to reinforce mechanical properties and suppress volume shrinkage in different aerogels, e.g., SiO₂ aerogel, resorcinol-formaldehyde aerogel, cellulose aerogel, and carbon aerogel.^38–41 Inspired by these findings, we introduce GO nanosheets in the system for topological deformation of Ni-MOFs for the first time. The presence of GO nanosheets indeed resists volume shrinkage during the pyrolysis process, which can be decreased from 81.5% of Ni/C composite aerogels (NCCAs) to 38.4% of Ni/C@rGO composite aerogels (NCGCAs). As a result, NCGCAs exhibit high porosity up to 96.3%, and their EM wave absorption performance is consequently enhanced. The optimized sample shows a minimum RL of −22.7 dB at 1.3 mm and EAB of 5.2 GHz with a very small absorber thickness of 1.5 mm. Moreover, NCGCAs also have favorable thermal management capabilities, as they retain good thermal insulation even after heating at about 200 °C for 2 h. This GO-mediated strategy for high-porosity MCCs aerogels may open a new avenue for the transformation of MOFs aerogels in the future.

Experimental section

Fabrication of Ni-MOF

In a typical synthesis, 1.90 g of Ni(NO₃)₂·6H₂O (6.5 mmol) and 1.37 g of 1,3,5-benzenetricarboxylic acid (6.5 mmol) were successively dissolved in 30.0 mL of N,N-dimethylformamide (DMF). The resultant solution was then heated at 150 °C for 3 h. At last, the products were collected by filtration, washed with DMF and ethanol for several times, and finally dried under vacuum at 60 °C overnight.

Fabrication of the Ni-MOF aerogel and Ni-MOF@GO aerogels

The Ni-MOF aerogel was prepared by following our previous report.³⁶ 1.10 g of Ni-MOF was dispersed into 3.0 mL of absolute ethanol, and then 27.0 mL of deionized water was introduced under ultrasonication. The suspension was aged at room temperature for 24 h to form a hydrogel. Finally, the hydrogel was frozen and freeze-dried for 48 h to obtain the Ni-MOF aerogel. Ni-MOF@GO aerogels were prepared in the same way, except that deionized water was replaced by an equal volume of GO aqueous suspension, and Ni-MOF@GO-1, Ni-MOF@GO-2, and Ni-MOF@GO-3 corresponded to GO concentrations of 1.5 mg mL⁻¹, 3.0 mg mL⁻¹ and 4.5 mg mL⁻¹, respectively. The GO aqueous suspension was ultrasonicated for 6 h, and there were no obvious precipitates after centrifugation at 10 000 rpm for 15 min, suggesting that the GO nanosheets were dispersed homogeneously in the solution.

Fabrication of the Ni/C composite aerogel (NCCA) and Ni/C@rGO composite aerogels (NCGCAs)

The pyrolysis of the Ni-MOF aerogel, Ni-MOF@GO-1, Ni-MOF@GO-2, and Ni-MOF@GO-3 was carried out in a tubular furnace with a constant flow of N₂. All samples were firstly heated at 100 °C for 1 h, then pyrolyzed at 700 °C for 2 h, and finally brought to room temperature via furnace cooling. The final resultant composites were sequentially named as NCCA, NCGCA-1, NCGCA-2, and NCGCA-3, respectively.

Characterization

Scanning electron microscope (SEM) images were recorded by a Phenom-World BV, and transmission electron microscope (TEM) images were taken on a Tecnai G2 F20. Powder X-ray diffractometer (XRD) data were collected on a D/MAXRC XRD with a Cu-Kα radiation source (40.0 kV, 40.0 mA). Nitrogen adsorption–desorption and mercury injection porosimetry measurements were performed on a QUADRASORB SI-KR/MP and an AutoPore IV 9510, respectively. Thermal gravimetric (TG) analyses were carried out on a SDT Q600 TGA (TA Instruments) under air atmosphere from room temperature to 800 °C with a temperature ramp of 5 °C min⁻¹, and Raman spectra were collected on a confocal Raman spectroscope equipped with a 633 nm laser. Magnetic properties were obtained by Vibrating Sample Magnetometer (VSM) (LakeShore 7404). Electromagnetic parameters were measured on an Agilent N5234A vector network analyzer, where a testing ring with an inner diameter of 3.0 mm and an outer diameter of 7.0 mm was pressed from the mixture of NCCA/NCGCAs (15 wt%) and paraffin (85 wt%). Thermal IR imaging digital images were recorded on a BY1515 heating platform with the temperature setting at 200 °C.

Results and discussion

The synthetic route for NCGCAs is schematically depicted in Fig. 1a. Ni-MOF@GO aerogels with a 3D structure assembled by GO nanosheets and crosslinked fibers derived from Ni-MOF are obtained after the aging and freeze-drying process. After high-temperature pyrolysis, Ni-MOF fibers and GO nanosheets will be converted into Ni/C fibers and rGO nanosheets, respectively, resulting in the formation of NCGCAs. As we reported previously, due to the hydrolysis of Ni-MOF and the secondary complexation between Ni²⁺ and organic ligands, Ni-MOF aerogels can be obtained by freeze-drying after topological deformation of Ni-MOF. The deformation process is closely related to the polarity of the solvent, and the final NCCA derived from individual Ni-MOF consists of crosslinked Ni/C fibers (Fig. 1b).³⁶ One can easily find rGO nanosheets in NCGCAs, and more importantly, rGO nanosheets are closely attached to the skeleton composed of these Ni/C fibers (Fig. 1c–e), suggesting that a different 3D interconnected structure has been created in these composites. It is worth noting that each rGO nanosheet can attach on many Ni/C fibers at the same time due to its larger size. Thus, the introduction of rGO nanosheets may produce significant resistance to thermal shrinkage. Fig. S1a† shows that the high-temperature pyrolysis process indeed results in an obvious size change in the macroscopic profile from the Ni-MOF aerogel to NCCA. In contrast, the shrinkage degree of NCGCAs is significantly decreased (Fig. S1b†), and the inhibitory effect is gradually enhanced upon increasing the initial concentration of the GO nanosheets (Fig. S1c†). Based on these images, the shrinkage ratios of NCCA, NCGCA-1, NCGCA-2 and NCGCA-3 are 81.5%, 68.4%, 55.8% and 38.4%, respectively. To exclude the possibility that this 3D structure is generated by GO nanosheets, we also examine the state of the individual GO suspension after aging for 24 h under the same condition as that of Ni-MOF@GO-2. The result indicates that the individual GO suspension cannot be converted into a GO aerogel (Fig. S2†). Thus, the formation of NCGCAs is confirmed to be from the synergistic assembly between the Ni-MOF fibers and GO nanosheets.

Fig. 1 Schematic illustration for the fabrication of NCGCAs (a). SEM images of NCCA (b), NCGCA-1 (c), NCGCA-2 (d) and NCGCA-3 (e). TEM images of NCCA (f) and NCGCA-2 (g). HR-TEM images of NCGCA-2 (h–j). STEM image of NCGCA-2 (k) and its corresponding EDX elemental mapping of Ni (l) and C (m), and digital photographs (n) of NCGCA-2.

TEM images manifest that as a basic structural unit in NCCA, each single fiber contains numerous Ni nanoparticles with uniform dispersion (Fig. 1f). Although the microscopic morphology of these fibers in NCGCAs is almost identical to that in NCCA (Fig. 1c–e), the introduction of the rGO nanosheets spreads the dispersion of Ni nanoparticles from carbon fibers to rGO nanosheets (Fig. 1g, marked with red circles in the inset). This is because the topological deformation of Ni-MOF refers to two necessary processes: the hydrolysis of Ni-MOF and the secondary complexation between Ni²⁺ and BTC. When the hydrolysis occurs in the initial suspension, some free Ni²⁺ will be captured by O-containing groups on GO nanosheets, and finally be converted into Ni nanoparticles on the surface of the rGO nanosheets. Moreover, these nanoparticles in NCGCAs present a typical core–shell configuration (Fig. 1h), and the lattice fringes for the core and the shell are around 0.20 and 0.34 nm (Fig. 2i and j), respectively, which correspond to the (111) plane of metal Ni with a faced-centered cubic (FCC) structure and the (002) plane of graphitic carbon, respectively. This is a common phenomenon in Ni/C composites, and ascribed to the catalytic graphitization effect of Ni nanoparticles during high-temperature process.²³ According to the results of elemental mapping, NCGCAs are mainly composed of Ni and C elements, and the profiles of the Ni and C element distribution are well consistent with that of the STEM image (Fig. 1k–m). All these results confirm that rGO nanosheets are homogeneously attached on the skeletons composed of Ni/C fibers. These nanosheets play a key role in resisting thermal shrinkage and creating larger porosity, thus further consolidating the lightweight feature (Fig. 1n).

Fig. 2 XRD patterns (a). TG curves (b) and the relative contents of carbon and Ni (c). Raman spectra (d). Magnetic hysteresis loops (e), and the histograms of the BET surface areas and pore volumes (f) of NCCA and NCGCAs.

Fig. 2a shows the XRD patterns of NCCA and NCGCAs. As observed, all samples exhibit three distinguishable diffraction peaks at 44.5°, 51.8° and 76.4°, which should be assigned to the (111), (200), and (220) lattice planes of metal Ni (JCPDS no. 04-0850), respectively (Fig. 2a1). There is also a small peak at about 26.4° indexed to the (002) plane of graphitic carbon (JCPDS no. 41-1487), which can be further discerned in the range of 2θ from 24° to 30° (Fig. 2a2). The gradually increased intensity of this peak implies that the relative graphitization degree is also reinforced from NCCA to NCGCA-3 with the introduction of rGO nanosheets. No other peaks can be detected, verifying that Ni²⁺ in the Ni-MOF aerogels and Ni-MOF@GO aerogels has been completely reduced during the high-temperature pyrolysis process.

The TG curves of NCCA and NCGCAs are shown in Fig. 2b, and they display essentially similar thermal behaviors, where each sample has a weight increase region related to the oxidation of the Ni nanoparticles and a weight loss region related to the combustion of the carbon skeleton. These TG curves will be constant with negligible weight change after the complete oxidation of Ni nanoparticles and full combustion of carbon skeletons. The relative carbon content can be calculated by the following formula:¹⁶ 1 − M(Ni)/M(NiO) wt% R = wt% carbon. As shown in Fig. 2c, the corresponding values for NCCA, NCGCA-1, NCGCA-2, and NCGCA-3 are 14.0 wt%, 27.6 wt%, 36.1 wt%, and 42.2 wt%, respectively, suggesting that the rGO nanosheets not only regulate the pore structure, but also adjust the carbon content of these composites.

Fig. 2d presents the Raman spectra of NCCA and NCGCAs. There are two distinguishable broad bands at around 1340 cm⁻¹ and 1590 cm⁻¹ for all samples, which are commonly defined as the D band and G band, respectively.^42,43 Generally speaking, the D band corresponds to the A_1g breathing mode and involves phonons close to the K zone boundary, which usually becomes active in disordered or defective carbon materials and cannot be activated in perfect graphite crystals.⁷ The G band is related to the E_2g mode that originates from the stretching vibrations of all carbon atoms with the sp² bond, but is not limited to graphitic carbon. The relative graphitization degree of carbon materials is widely assessed using the intensity ratio of the D band to G band (I_D/I_G). The I_D/I_G values of NCCA, NCGCA-1, NCGCA-2 and NCGCA-3 are 0.78, 0.87, 0.91 and 0.95, respectively. Based on the interpretation about the variation in the I_D/I_G values proposed by Ferrari and Robertson, the I_D/I_G value would not present a monotonous decrease even if defect sites were eliminated from amorphous carbon to graphitic carbon. In contrast, it would gradually increase with the enhancement of the atomically ordered arrangement in the stage from amorphous carbon to nanocrystalline graphite because the ordered degrees of the bond angle and bond bending were significantly improved.⁴⁴ Although the characteristic of graphitic carbon can be detected by TEM images and XRD patterns (Fig. 1j and 2a2), the crystallization is only limited to a small region around the Ni nanoparticles due to their catalytic graphitization degree, and the carbon skeletons are predominantly amorphous in the overall structure.¹⁴ Thus, the carbon skeletons in these samples are still in the transition stage from amorphous carbon to nanocrystalline graphite. Therefore, the incremental I_D/I_G value in our case can also be taken as a sign to indicate the consolidation in the relative graphitization degree from NCCA to NCGCA-3.

Fig. 2e displays the magnetic hysteresis loops of NCCA and NCGCAs. All samples exhibit typical ferromagnetic behaviors, which include a drastic rise in magnetization and tend to become saturated in the region of a weak applied field.^9,45 The values of saturation magnetization for NCCA, NCGCA-1, NCGCA-2 and NCGCA-3 are 33.2, 25.5, 23.8 and 21.9 emu g⁻¹, respectively. The increasing content of the non-magnetic carbon species is responsible for the decrease in saturation magnetization from NCCA to NCGCA-3 (Fig. 2c). Coercivity (H_c) can also be acquired from magnetic hysteresis loops, and the specific values are 78.6 Oe for NCCA, 69.6 Oe for NCGCA-1, 57.9 Oe for NCGCA-2 and 50.5 Oe for NCGCA-3. In our case, there is no anisotropic growth of Ni nanoparticles. Thus, the difference in these H_c values may be attributed to their average grain size rather than magneto-crystalline anisotropy or shape anisotropy.¹⁸ Typically, when the size of the magnetic particles is less than their critical size, the value of H_c is positively correlated with the particle size. Conversely, if the size of the magnetic particles exceeds the critical size, they will be negatively correlated. According to our previous study, the Ni nanoparticles in NCCA from 700 °C are obviously less than 55 nm.³⁶ The introduction of rGO nanosheets further disperses the Ni nanoparticles from carbon fibers. Thus, the Ni nanoparticles in NCGCAs may have a smaller size, which results in the decrease of H_c values from NCCA to NCGCA-3.

Fig. 3a exhibits the N₂ adsorption–desorption isotherms of NCCA and NCGCAs. All samples generate long and narrow hysteresis loops when the relative pressure P/P₀ is between 0.4 and 1.0. These isotherms can be categorized as IV-type based on the IUPAC standard, and are suggestive of abundant micropores and mesopores in the carbon skeletons.⁴⁶ Compared to NCCA, the hysteresis loops of NCGCAs will appear at a higher P/P₀ value over 0.7, suggesting the presence of a lamellar structure in the skeleton. As calculated by the Brunauer–Emmett–Teller method, the specific surface areas and the pore volumes of NCCA and NCGCAs are shown in Fig. 2f. It is found that the specific surface area and pore volume of NCGCA are lower than those of NCCA (145.7 m² g⁻¹ and 0.17 cm³ g⁻¹). By considering that the N₂ adsorption–desorption technique only evaluates the textural information from micropores and mesopores, the cliff-like drops of the specific surface area and pore volume may be attributed to the fact that a substantial part of the micropores and mesopores are covered by rGO nanosheets and inaccessible to N₂ molecules. However, with the increase of rGO nanosheets, some additional micropores and mesopores are created in the rGO nanosheets due to the removal of surface groups during high-temperature pyrolysis, which causes the specific surface area and pore volume to be increased gradually from NCGCA-1 to NCGCA-3. To collect the information on those 3D macropores, mercury intrusion-extrusion measurements are further carried out. Different from the N₂ adsorption–desorption technique, an external pressure is always required to push mercury into the void of macroporous materials. Thus, with a smaller pore size, a greater external pressure is required.⁴⁷ As shown in Fig. 3b, the mercury intrusion-extrusion isotherms of NCCA and NCGCAs reveal that the mercury cumulative intrusion of these curves rapidly increase in the low-pressure range (0–100 psia), indicating that these Ni/C composite aerogels have abundant macropores. The intrusion/extrusion curve of NCCA almost reaches a saturation state at around 4000 pisa, and its specific saturation volume and porosity are 3.81 mL g⁻¹ and 91.4%, respectively. As for NCGCAs, the corresponding pressure for the saturation state decreases to around 1000 pisa, and their specific saturation volumes and porosities are 6.74 mL g⁻¹ and 94.0% for NCGCA-1, 12.24 mL g⁻¹ and 94.7% for NCGCA-2, and 19.45 mL g⁻¹ and 96.3% for NCGCA-3, respectively. These results firmly demonstrate that rGO nanosheets indeed play a crucial role in maintaining the 3D network during high-temperature pyrolysis. The densities for NCCA, NCGCA-1, NCGCA-2 and NCGCA-3 are deduced as 0.24, 0.14, 0.07 and 0.04 g cm⁻³, respectively, indicating that the preservation of the macroscopic profile favors small density, so that a Pennisetum can easily support an aerogel monolith without any deformation (Fig. 1n). Fig. 3c shows the pore size distribution curves of NCCA and NCGCAs. It can be seen that all curves have peaks in the range of 1000–4000 nm, further suggesting that there are abundant macropores in both NCCA and NCGCAs. These peaks gradually increase from NCCA to NCGCA-3, which is consistent with the above results. In addition, another peak over 100 μm can be observed in all curves, and these ultra-large pores may be related to the fact that fibrous or lamellar units in the aerogel cannot create completely close macropores. Overall, the number of macropores and ultra-large macropores in NCCA and NCGCAs gradually increases with the content of rGO nanosheets, and this situation will be more conducive to multiple reflections and energy attenuation of EM waves in NCGCAs.

Fig. 3 N₂ adsorption–desorption isotherms (a). Mercury intrusion-extrusion isotherms (b), and pore size distributions curves (c) of NCCA and NCGCAs.

Generally speaking, the relative complex permittivity and relative complex permeability of EWAMs are considered as two critical factors when determining their EM absorption performance.^17,48–50 Frequency-dependent and curves of NCCA and NCGCAs in the range of 2.0–18.0 GHz are shown in Fig. 4a and b. As observed, all samples exhibit typical frequency dispersion behaviors, while the corresponding values of and are both gradually increased along with the incremental content of rGO. For example, NCCA displays the lowest and values among these samples, and its and values decline from 7.3 and 1.3 at 2.0 GHz to 5.9 and 1.1 at 18.0 GHz, respectively. By comparison, NCGCAs present relatively larger and values, and their values at 2.0 GHz are 12.9 and 3.0 for NCGCA-1, 19.1 and 9.8 for NCGCA-2, and 32.9 and 23.0 for NCGCA-3, respectively. The corresponding values at 18.0 GHz are 10.3 and 2.7 for NCGCA-1, 11.5 and 4.9 for NCGCA-2, and 19.5 and 10.8 for NCGCA-3, respectively. A dielectric loss tangent is usually utilized to evaluate the dielectric loss capability of EWAMs.⁵¹ Fig. S3† shows the tan δ_e curves of the mixture with NCCA and NCGCAs, and the result confirms that the tan δ_e sequence of these samples is NCGCA-3 > NCGCA-2 > NCGCA-1 > NCCA, which is consistent with that of the relative complex permittivity (Fig. 4a and b). According to the Debye theory, the dielectric loss capability of EWAMs mainly depends on two aspects: conductive loss and polarization loss.⁵² It is well acknowledged that the conductive loss is closely related to the conductivity of EWAMs. Thus, we test the conductivities of the wax-based mixture with NCCA and NCGCAs by the four-probe resistivity.⁵³ The conductivities of the mixtures are 2.93 × 10⁻³ with NCCA, 8.01 × 10⁻³ with NCGCA-1, 1.69 × 10⁻² with NCGCA-2 S cm⁻¹, and 6.15 × 10⁻² with NCGCA-3 S cm⁻¹, respectively. Although both carbon and nickel are conventionally good conductors, the Ni nanoparticles in NCCA and NCGCAs are nanoscaled in size, and the small size effect dictates that they will not produce a significant contribution to the conductivity of these composites.¹⁶ Therefore, the 3D carbon skeleton can be taken as the primary source of conductive loss. In general, the conductivities of carbon-based composites are highly dependent on their carbon content and relative graphitization degree. In our case, they are both gradually increased from NCCA to NCGCA-3 (Fig. 2c and d). Thus, the composition is one of the reasons for the incremental conductivity. Additionally, the specific volume of carbon-based EWAMs is positively correlated with their conductivities because a large specific volume will favor the formation of conductive networks in an organic matrix (wax herein). Since the presence of rGO nanosheets effectively inhibits the thermal shrinkage of the Ni/C composite aerogels and their porosities progressively increase from NCCA to NCGCA-3, their specific volumes gradually increase as well, which is also an important reason for the enhancement in conductivity.

Fig. 4 (a), (b), (c), and (d) of NCCA and NCGCAs. 3D RL maps of NCCA (e), NCGCA-1 (f), NCGCA-2 (g), and NCGCA-3 (h). The histograms of the maximum EAB and corresponding thickness (i). The attenuation constants (j), and the intrinsic wave impedance curves (k) of NCCA and NCGCAs.

Among various polarization relaxation modes, the interfacial polarization and dipole orientation polarization are thought to be primarily responsible for the energy attenuation of EM waves in the range of 2.0–18.0 GHz.⁵⁴ The introduction of rGO nanosheets not only creates new heterogeneous interfaces between the rGO nanosheets and Ni/C fibers, but also promotes the distribution of Ni nanoparticles from carbon fibers to rGO nanosheets. Therefore, it can be predicted that the interfacial polarization will get increasingly stronger from NCCA to NCGCA-3.⁵⁵ When it comes to the dipole orientation polarization, the Raman spectra reveal that the relative graphitization degree of the carbon skeletons in these composites presents an increasing tendency from NCCA to NCGCA-3, which means a reduction in the number of defects, as well as boundary charges and dipoles of carbon component, thereby leading to the diminished dipole orientation polarization from NCCA to NCGCA-3. Although the conductive loss, interfacial polarization and dipole orientation polarization may all contribute to the dielectric loss of these composites, it is apparent that conductive loss and interfacial polarization will be two more important modes because they exhibit similar trends with the dielectric tangents (Fig. S3†).

Fig. 4c and d display the and values of NCCA and NCGCAs. Despite some slight fluctuations, the and values of these samples are both gradually decreased from NCCA to NCGCAs. Meanwhile, the difference in the values and magnetic loss tangents of these samples is pretty small, and also in the sequence of NCCA < NCGCA-1 < NCGCA-2 < NCGCA-3 (Fig. S4†). This phenomenon can be attributed to the fact that the magnetic coupling among Ni nanoparticles in these composites is gradually weakened as the carbon content of the non-magnetic component increases, thus leading to the decreases in the and values, as well as the magnetic loss ability. Typically, a magnetic loss in the range of 2.0–18.0 GHz can be attributed to the eddy current effect and ferromagnetic resonance. If the magnetic loss only benefits from the eddy current effect, the values of C₀ will be constant in the studied frequency range.¹³ It is evident that the C₀ values of NCCA and NCGCAs continuously change in the 2.0–18.0 GHz (Fig. S5†), implying that both ferromagnetic resonance and eddy current contribute to magnetic loss in NCCA and NCGCAs.

The EM absorption properties of EWAMs can be described by RL intensities at different frequency points with the following equations:¹⁰

where Z_in represents the normalized input impedance, c is the velocity of the EM wave in free space, and d is the thickness of EWAMs. Fig. 4e–h illustrate the 3D RL maps of NCCA and NCGCAs, which reveal the variation of the RL intensities of these Ni/C composites with the applied thickness from 1.0 to 5.0 mm in the frequency range of 2.0–18.0 GHz, and the RL intensities are deliberately truncated to −20.0 dB for clarity. It is obvious that the RL intensities of NCCA and NCGCAs are able to reach below −10 dB, indicating that they are all able to satisfactorily absorb the incident EM wave. As shown, the strongest RL intensities of NCCA, NCGCA-1, NCGCA-2, and NCGCA-3 are −19.2 dB (f = 18.0 GHz, d = 5.0 mm), −20.5 dB (f = 4.3 GHz, d = 5.0 mm), −22.7 dB (f = 18.0 GHz, d = 1.3 mm), and −12.4 dB (f = 12.4 GHz, d = 1.0 mm), respectively. Compared to NCCA, the RL intensities of NCGCAs are improved after the introduction of rGO. Especially for NCGCA-2, its relatively strong RL value (−22.7 dB) is achieved upon decreasing the thickness. The effective absorption bandwidth (EAB) is regarded as a more important index for evaluating the EM absorption performance of EWAMs. Fig. 4i exhibits the maximum EAB and corresponding thickness of all samples, and their values are 2.2 GHz for NCCA (d = 5.0 mm), 3.4 GHz for NCGCA-1 (d = 1.5 mm), 5.2 GHz for NCGCA-2 (d = 1.5 mm) and 2.9 GHz for NCGCA-3 (d = 1.0 mm), respectively. After the introduction of rGO, the maximum EAB is effectively broadened, and the corresponding thickness is greatly reduced. In view of the above results, one can conclude that the presence of rGO nanosheets can effectively reinforce the EM wave absorption performance of these Ni/C composite aerogels, and NCGCA-2 is the best candidate among these samples. Meanwhile, we also tested the relative complex permittivity and complex permeability of NCGCA-2 from the different sites of the as-obtained aerogel monolith (NCGCA-2-a and NCGCA-2-b) and calculated their EM wave absorption properties (Fig. S6†). As observed, NCGCA-2-a and NCGCA-2-b exhibit similar , , and values, as well as EM wave absorption performance to those of NCGCA-2. These results again illustrate the homogeneous dispersion of rGO nanosheets in the obtained aerogels. In order to validate the EM wave absorption performance of NCGCA-2 more objectively, we have summarized the RL intensities and EAB values of some previously reported Ni/C composite aerogels with similar components (Table S1†).^56–62 In comparison, although the RL intensity of NCGCA-2 is not the best among these composites, it does exhibit a broader EAB value with a decreased thickness. Overall, the introduction of rGO nanosheets into the MOF-derived aerogels to resist thermal shrinkage is an effective strategy to improve the EM wave absorption performance of the Ni/C composite aerogels.

It is worth noting that NCGCA-2 does not generate the strongest dielectric loss or magnetic loss among the three samples, while it produces the best EM absorption performance. To disclose the underlying reasons, we further analyze and discuss the attenuation constant (α) of NCCA and NCGCA, which can provide a more comprehensive evaluation in the overall dissipation capability than the tan δ_e and tan δ_m values, and be deduced from the following equation:⁶³

Fig. 4j demonstrates the calculated α values of NCCA and NCGCAs. One can easily find that the α values gradually increase with the increase of the rGO content, indicating the enhancement of the overall attenuation capacity from NCCA to NCGCA-3. Very interestingly, NCGCA-2 still fails to generate the strongest attenuation capability. This phenomenon is because there is another important factor, i.e., characteristic impedance, that can affect the EM wave absorption performance. Only by combining sufficient attenuation capability with good impedance matching can favorable EM wave absorption performance be realized. The impedance matching characteristics of EWAMs can usually be evaluated using the values of their intrinsic wave impedance (η_in) through the following equation:⁶⁴

In general, the closer the η_in value is to 377 Ω, the less EM wave will be reflected at the interface between the transmission medium and the free space. This means that the impedance matching is better, indicating that the intrinsic attenuation capability of EWAMs can consume more EM wave. Fig. 4k demonstrates the calculated η_in values of NCCA and NCGCAs. Obviously, the η_in values gradually decrease with the increase of the rGO content, and the average η_in values are 154 Ω for NCCA, 114 Ω for NCGCA-1, 99 Ω for NCGCA-2, and 74 Ω for NCGCA-3, indicating that the high loading of rGO and the relatively high graphitization degree of the carbon skeleton are not beneficial to good impedance matching. NCGCA-2 has moderate attenuation capability and characteristic impedance, which balance the relationship between the incidence and loss of EM wave, thus bringing good EM absorption performance. The potential EM wave absorption mechanisms of NCGCA-2 are schematically shown in Fig. 5a. First, NCGCA-2 consists of Ni/C fibers coupled with 2D rGO nanosheets, and these nanosheets are loaded on the skeleton consisting of crosslinked fibers to form a new crosslinked porous structure. They effectively alleviate the thermal shrinkage of the carbon skeleton. This crosslinked porous structure is advantageous for electron hopping and migration. Thus, conductive loss occurs when an alternating EM field is applied to generate an induced current. Second, the carbon species that constitute fibers in NCGCA-2 is still mainly in the amorphous state. Thus, there are residual functional groups and defects in the carbon skeleton. They can play as dipoles to generate dipole reorientation polarization, as the direction of the applied EM field is continuously changed. Third, numerous Ni nanoparticles presenting in both carbon fibers and rGO nanosheets can provide magnetic loss through intrinsic ferromagnetic resonance and eddy current effects around them. Fourth, the introduction of rGO nanosheets provides abundant heterogeneous interfaces that include Ni nanoparticles and graphite carbon shells, 1D Ni/C fibers and 2D rGO nanosheets, as well as the graphite carbon shells around the Ni nanoparticles and amorphous carbon skeletons, which generate a powerful interfacial polarization. Final, the rGO nanosheets and Ni/C fibers create a new crosslinked-coupled 3D conductive network with extremely high porosity, which facilitates multiple reflections of the incident EM wave.

Fig. 5 Schematic illustration of EM wave absorption mechanisms in NCGCA-2 (a). The far-field response based on the plane wave theory of PEC (b), NCCA (c), NCGCA-1 (d), NCGCA-2 (e) and NCGCA-3 (f). RCS simulated curves from different scanning angles of PEC, NCCA and NCGCAs (g–j).

The radar cross-section (RCS), as an essential parameter for radar stealth technology to evaluate the EM wave absorption characteristics of EWAMs under actual far-field conditions, can be obtained by simulation with CST STUDIO software. Fig. 5b–f presents the CST simulation results of NCCA and NCGCAs. The simulation model is carried out on a perfect electric conductor (PEC) coated with NCCA or NCGCAs film, and the intensities of the reflected signals from the pure PEC line and PEC plates with the NCCA or NCGCAs film are monitored for the whole detection angle under a horizontal incident EM wave with the thickness and frequency setting at 2.0 mm and 12.5 GHz, respectively. It can be seen that the horizontally reflected intensity of NCGCA-2 is much weaker than those of other samples, indicating that more EM energy can be consumed by NCGCA-2. Fig. 5g–j show the 2D projections of the RCS distributions of PEC, NCCA and NCGCAs in the X–O–Y plane. It can be seen that NCGCA-2 exhibits superior EM absorption performance over the angle range of −90 to 90°. In conclusion, strong EM wave signals are effectively suppressed when a metal plane is covered with the NCGCA-2 film. These CST simulation results are consistent with the previously discussed EM wave absorption properties.

In addition to the EM wave absorption performance, the 3D porous structure also has an important impact on the thermal management performance of EWAMs. NCCA and NCGCAs are placed on the heating platform with a constant setting temperature of 200 °C, and thermal infrared images are collected at 1 min, 10 min, 30 min, 60 min, 90 min and 120 min (Fig. 6a–d). The heating platform is warmed up for a period of time before the testing, and the ambient air temperature is relatively high. Thus, the surface temperature of the NCCA and NCGCAs will rise rapidly at the beginning of the measurements. After heating for 120 min, the temperature changes of the upper surface for NCCA, NCGCA-1, NCGCA-2, and NCGCA-3 are 10.6, 9.8, 8.8, and 7.2 °C, respectively. The time-dependent temperature difference between the heating platform and the upper surface is also plotted in Fig. 6e. Although the temperature difference is gradually narrowed with the heating time, NCGCAs always display a larger temperature difference than NCCA, confirming the superiority of NCGCAs in thermal management. Generally speaking, the thermal insulation performance of a medium is mainly determined by its intrinsic properties and microstructure.⁶⁵ From NCCA to NCGCA-3, the carbon content is incremental and the relative graphitization degree is reinforced. Thus, the intrinsic thermal conductivity will be gradually enhanced. Meanwhile, the porosity is also enlarged from 91.4% to 96.3%. It is well accepted that the porous microstructure not only significantly increases the electron transfer distance and retards the heat transfer, but also blocks the upward transmission of heat energy by introducing a large amount of air. Thus, it is reasonable to say that the contribution of the microstructure to thermal insulation in NCCA and NCGCAs is positively correlated with their porosities.³² Based on these results, one can conclude that the microstructure herein plays a more important role in thermal management. Moreover, when NCGCA-2 comes into contact with hand skin, the corresponding thermal insulation image also demonstrates an obvious difference in temperature between the upper surface of the sample and hand skin (Fig. 6e, inset), further confirming the good thermal management capacity of NCGCA-2.

Fig. 6 Thermal insulation images of NCCA (a), NCGCA-1 (b), NCGCA-2 (c) and NCGCA-3 (d) at different heating times, and temperature difference curves between the heating platform temperature and upper surface temperature of NCCA and NCGCAs (e).

Conclusions

A series of Ni/C@rGO composite aerogels (NCGCAs) with very high porosity have been successfully constructed through GO-mediated topological deformation of Ni-MOF and subsequent high-temperature pyrolysis. GO nanosheets play an extremely important role in resisting the thermal shrinkage of Ni-MOF skeletons during the pyrolysis process. The low shrinkage ratio, as well as the conversion from GO to rGO, make a solid contribution to the dielectric loss capabilities of NCGCAs. When the dosage of GO nanosheets is rationally regulated, the resultant composite aerogel, NCGCA-2, will possess powerful attenuation ability and matched impedance. Thus, it can exhibit good electromagnetic (EM) absorption performance, especially for its broad EM response with a decreased thickness. In addition, the high-porosity structure also renders NCGCAs as promising candidates for thermal management because they show remarkable improvements in thermal insulation. It is believed that GO-mediated topological MOFs deformation may open a new avenue for the design of magnetic metal/carbon composite aerogels, and upgrade their performance in various application fields.

Author contributions

Pan Wang: conceptualization, methodology, investigation, validation, and writing. Dingge Fan: validation and visualization. Lixue Gai: visualization, software. Bo Hu: software. Ping Xu: resources. Xijiang Han: supervision, funding acquisition, resources. Yuchen Du: supervision, project administration, writing, and editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work is financially supported by the National Natural Science Foundation of China (52373262). We appreciate the help of Dr Long Xia (Xiamen University) for the discussion of N₂ adsorption–desorption.

References

1. Y. Zhang, Y. Huang, T. F. Zhang, H. C. Chang, P. S. Xiao, H. H. Chen, Z. Y. Huang and Y. S. Chen, Adv. Mater., 2015, 27, 2049–2053.
2. H. L. Lv, Z. H. Yang, H. G. Pan and R. B. Wu, Prog. Mater. Sci., 2022, 127, 100946.
3. L. L. Liang, W. H. Gu, Y. Wu, B. S. Zhang, G. H. Wang, Y. Yang and G. B. Ji, Adv. Mater., 2022, 34, 2106195.
4. M. S. Cao, X. X. Wang, M. Zhang, J. C. Shu, W. Q. Cao, H. J. Yang, X. Y. Fang and J. Yuan, Adv. Funct. Mater., 2019, 29, 1807398.
5. B. T. Yang, J. F. Fang, C. Y. Xu, H. Cao, R. X. Zhang, B. Zhao, M. Q. Huang, X. Y. Wang, H. L. Lv and R. C. Che, Nano-Micro Lett., 2022, 14, 170.
6. X. M. Guan, Z. H. Yang, M. Zhou, L. Yang, R. Peymanfar, B. Aslibeiki and G. B. Ji, Small Struct., 2022, 3, 2200102.
7. H. Zhao, X. Xu, Y. Wang, D. Fan, D. Liu, K. Lin, P. Xu, X. Han and Y. Du, Small, 2020, 16, 2003407.
8. Z. Zhang, Z. M. Xiong, Y. Yao, D. R. Wang, Z. Q. Yang, P. Zhang, Q. Zhao and W. K. Zhou, Adv. Funct. Mater., 2022, 32, 2206053.
9. T. Gao, R. Z. Zhao, Y. X. Li, Z. Y. Zhu, C. L. Hu, L. Z. Ji, J. Zhang and X. F. Zhang, Adv. Funct. Mater., 2022, 32, 2204370.
10. P. Wang, D. W. Liu, L. R. Cui, B. Hu, X. J. Han and Y. C. Du, J. Phys. D: Appl. Phys., 2021, 54, 473003.
11. Y. Fei, M. Liang, Y. Chen and H. W. Zou, Ind. Eng. Chem. Res., 2019, 59, 154–165.
12. Y. Wang, X. Li, X. Han, P. Xu, L. Cui, H. Zhao, D. Liu, F. Wang and Y. Du, Chem. Eng. J., 2020, 387, 124159.
13. P. T. Xie, Y. Liu, M. Feng, M. Niu, C. Z. Liu, N. N. Wu, K. Y. Sui, R. R. Patil, D. Pan, Z. H. Guo and R. H. Fan, Adv. Compos. Hybrid Mater., 2021, 4, 173–185.
14. F. Wang, Y. Liu, H. Zhao, L. Cui, L. Gai, X. Han and Y. Du, Chem. Eng. J., 2022, 450, 138160.
15. S. Gao, G. Z. Zhang, Y. Wang, X. P. Han, Y. Huang and P. B. Liu, J. Mater. Sci. Technol., 2021, 88, 56–65.
16. D. Liu, Y. Du, P. Xu, N. Liu, Y. Wang, H. Zhao, L. Cui and X. Han, J. Mater. Chem. C, 2019, 7, 5037–5046.
17. H. H. Zhao, F. Y. Wang, L. R. Cui, X. Z. Xu, X. J. Han and Y. C. Du, Nano-Micro Lett., 2021, 13, 208.
18. L. R. Cui, Y. H. Wang, X. J. Han, P. Xu, F. Y. Wang, D. W. Liu, H. H. Zhao and Y. C. Du, Carbon, 2021, 174, 673–682.
19. Q. W. Zeng, L. Wang, X. Li, W. B. You, J. Zhang, X. H. Liu, M. Wang and R. C. Che, Appl. Surf. Sci., 2021, 538, 148051.
20. N. N. Wu, D. M. Xu, Z. Wang, F. L. Wang, J. R. Liu, W. Liu, Q. Shao, H. Liu, Q. Gao and Z. H. Guo, Carbon, 2019, 145, 433–444.
21. R. W. Shu, W. J. Li, Y. Wu, J. B. Zhang and G. Y. Zhang, Chem. Eng. J., 2019, 362, 513–524.
22. K. Zhang, F. Wu, J. J. Li, M. X. Sun, A. M. Xie and W. Dong, Mater. Chem. Phys., 2018, 208, 198–206.
23. D. Liu, Y. Du, P. Xu, F. Wang, Y. Wang, L. Cui, H. Zhao and X. Han, J. Mater. Chem. A, 2021, 9, 5086–5096.
24. Z. N. Li, X. J. Han, Y. Ma, D. W. Liu, Y. H. Wang, P. Xu, C. L. Li and Y. C. Du, ACS Sustainable Chem. Eng., 2018, 6, 8904–8913.
25. Y. Xiong, L. l. Xu, C. X. Yang, Q. F. Sun and X. J. Xu, J. Mater. Chem. A, 2020, 8, 18863–18871.
26. Z. H. Yang, H. L. Lv and R. B. Wu, Nano Res., 2016, 9, 3671–3682.
27. Y. J. Sun, N. Wang, H. Y. Yu and X. Z. Jiang, J. Mater. Sci., 2020, 55, 7304–7320.
28. D. Ding, Y. Wang, X. D. Li, R. Qiang, P. Xu, W. L. Chu, X. J. Han and Y. C. Du, Carbon, 2017, 111, 722–732.
29. L. Liang, Q. Li, X. Yan, Y. Feng, Y. Wang, H. B. Zhang, X. Zhou, C. Liu, C. Shen and X. Xie, ACS Nano, 2021, 15, 6622–6632.
30. H. B. Zhao, J. B. Cheng, J. Y. Zhu and Y. Z. Wang, J. Mater. Chem. C, 2019, 7, 441–448.
31. B. Zhao, Y. Li, H. Y. Ji, P. W. Bai, S. Wang, B. B. Fan, X. Q. Guo and R. Zhang, Carbon, 2021, 176, 411–420.
32. Y. H. Wang, M. H. Zhang, X. S. Deng, Z. G. Li, Z. S. Chen, J. M. Shi, X. J. Han and Y. C. Du, Int. J. Miner., Metall. Mater., 2023, 30, 536–547.
33. L. L. Xu, Y. Xiong, B. K. Dang, Z. N. Ye, C. D. Jin, Q. F. Sun and X. H. Yu, Mater. Des., 2019, 182, 108006.
34. W. J. Yu, B. Liu and X. J. Zhao, Nanomaterials, 2022, 12, 655.
35. S. S. Wang, Y. C. Xu, R. R. Fu, H. H. Zhu, Q. Z. Jiao, T. Y. Feng, C. H. Feng, D. X. Shi, H. S. Li and Y. Zhao, Nano-Micro Lett., 2019, 11, 76.
36. P. Wang, L. X. Gai, B. Hu, Y. L. Liu, F. Y. Wang, P. Xu, X. J. Han and Y. C. Du, Carbon, 2023, 212, 118132.
37. Z. H. Zhou, Q. Q. Zhu, Y. Liu, Y. Zhang, Z. R. Jia and G. L. Wu, Nano-Micro Lett., 2023, 15, 137.
38. S. Dervin, Y. Lang, T. Perova, S. H. Hinder and S. C. Pillai, J. Non-Cryst. Solids, 2017, 465, 31–38.
39. W. Sun, A. Du, B. Zhou, J. Shen, S. M. Huang and J. Tang, J. Sol-Gel Sci. Technol., 2016, 80, 68–76.
40. S. Z. Zhang, J. P. He, S. X. Xiong, Q. Xiao, Y. Y. Xiao, F. Ding, H. Ji, Z. Y. Yang and Z. Q. Li, Biomacromolecules, 2021, 22, 3216–3222.
41. X. Y. Jiang, Z. X. Zhao, S. T. Zhou, H. W. Zou and P. B. Liu, ACS Appl. Mater. Interfaces, 2022, 14, 45844–45852.
42. N. Wang, W. J. Ma, Z. Q. Ren, Y. C. Du, P. Xu and X. J. Han, J. Mater. Chem. A, 2018, 6, 884–895.
43. Y. X. Li, Y. J. Liao, L. Z. Ji, C. L. Hu, Z. H. Zhang, Z. Y. Zhang, R. Z. Zhao, H. W. Rong, G. W. Qin and X. F. Zhang, Small, 2022, 18, 2107265.
44. A. Ferrari and J. Robertson, Phys. Rev. B, 1999, 61, 14095–14107.
45. F. Y. Wang, Y. L. Liu, R. D. Feng, X. Wang, X. Han and Y. Du, Small, 2023, 19, 2303597.
46. B. B. Wang, W. Y. Zhang, J. M. Sun, C. H. Lai, S. B. Ge, H. W. Guo, Y. Liu and D. H. Zhang, J. Materi. Chem. A, 2023, 11, 8656–8669.
47. X. T. Ma, Y. Lou, X. B. Chen, Z. Shi and Y. Xu, Chem. Eng. J., 2019, 356, 227–235.
48. P. Das, A. Katheria, J. Nayak, A. Pal, B. Roy, S. Paul, S. Biswas and N. C. Das, J. Alloys Compd., 2024, 976, 173011.
49. P. Das, A. Katheria, A. Jana, M. Das, B. Roy, J. Nayak, K. Nath, S. K. Ghosh, A. De and N. C. Das, ACS Appl. Eng. Mater., 2023, 1, 1186–1200.
50. P. Das, A. Katheria, S. Ghosh, B. Roy, J. Nayak, K. Nath, S. Paul and N. C. Das, Synth. Met., 2023, 294, 117304.
51. L. J. Rao, L. Wang, C. D. Yang, R. X. Zhang, J. C. Zhang, C. Y. Liang and R. C. Che, Adv. Funct. Mater., 2023, 33, 2213258.
52. M. Qin, L. M. Zhang and H. J. Wu, Adv. Sci., 2022, 9, 2105553.
53. Y. L. Liu, C. H. Tian, F. Y. Wang, B. Hu, P. Xu, X. J. Han and Y. C. Du, Chem. Eng. J., 2023, 461, 141867.
54. B. L. Wang, Q. Wu, Y. G. Fu and T. Liu, J. Mater. Sci. Technol., 2021, 86, 91–109.
55. Z. Liu, C. H. Liu, Y. Wang, M. p. Song, J. C. Guo, W. Wang and X. P. Gao, Carbon, 2024, 217, 118655.
56. X. L. Ye, J. X. Zhang, X. M. Ma, X. Li and H. Yu, J. Phys. Chem. C, 2023, 127, 1251–1259.
57. H. B. Zhao, Z. B. Fu, X. Y. Liu, X. C. Zhou, H. B. Chen, M. L. Zhong and C. Y. Wang, Ind. Eng. Chem. Res., 2017, 57, 202–211.
58. S. J. Han, S. Y. Wang, W. H. Li, Y. R. Lai, N. Zhang, N. Yang, Q. H. Wang and W. Jiang, Ceram. Int., 2018, 44, 10352–10361.
59. H. B. Zhao, Z. B. Fu, H. B. Chen, M. L. Zhong and C. Y. Wang, ACS Appl. Mater. Interfaces, 2016, 8, 1468–1477.
60. T. Wu, X. S. Hu, C. S. Shao, Z. Y. Cao, C. C. Li, S. Gao, X. Z. Ren, C. Y. Ding, G. W. Wen, X. X. Huang and S. S. Wu, J. Colloid Interface Sci., 2023, 629, 44–52.
61. Y. P. Zhao, X. Q. Zuo, Y. Guo, H. Huang, H. Zhang, T. Wang, N. X. Wen, H. Chen, T. Z. Cong, J. Muhammad, X. Yang, X. N. Wang, Z. Fan and L. J. Pan, Nano-Micro Lett., 2021, 13, 144.
62. Q. Y. Wang, X. F. Wu, J. Huang, S. Y. Chen, Y. Zhang, C. J. Dong, G. Chen, L. H. Wang and H. T. Guan, J. Alloys Compd., 2022, 890, 161887.
63. Q. Li, Z. Zhang, L. P. Qi, Q. L. Liao, Z. Kang and Y. Zhang, Adv. Sci., 2019, 6, 1801057.
64. L. X. Gai, H. H. Zhao, F. Y. Wang, P. Wang, Y. L. Liu, X. J. Han and Y. C. Du, Nano Res., 2022, 15, 9410–9439.
65. J. Wang, D. S. Yuan, P. Y. Hu, Y. J. Wang, J. Wang and Q. W. Li, Adv. Funct. Mater., 2023, 33, 2300441.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ta00125g

---