Introduction

Due to the popularization of electronic equipment, computer network and mobile phone, electromagnetic wave became ubiquitous and brought a series of problems inevitably [1,2,3]. For example, electromagnetic wave pollution would have detrimental effects on human health and precision instruments [4,5,6,7,8,9,10,11], so that the world health organization ranked it as the fourth pollution after water pollution, air pollution and noise pollution [12,13,14,15]. MA materials could convert electromagnetic waves into joule heat to solve electromagnetic wave pollution [16]. At the same time, MA materials could also be used in the military anti-reconnaissance field to prevent the aircraft by enemy radars. Therefore, the research of MA materials had become a hotspot in recent years.

It was well known that electromagnetic wave was composed of oscillating electric and magnetic fields. Therefore, the loss of MA materials was mainly composed of dielectric loss and magnetic loss. The reflection loss (RL) values were calculated by the following equations at a given frequency and thickness layer according to the transmit line theory [17, 18]:

$${Z}_{\mathrm{in}}={Z}_{0}\sqrt{\frac{{\mu }_{r}}{{\varepsilon }_{r}}}\mathrm{tanh}\left[j\left(\frac{2\pi fd}{c}\right)\sqrt{{\mu }_{r}{\varepsilon }_{r}}\right]$$
$$\mathrm{RL}\left(\mathrm{dB}\right)=20\mathrm{lg}\left|\frac{{Z}_{\mathrm{in}}-{Z}_{0}}{{Z}_{\mathrm{in}}+{Z}_{0}}\right|$$

where Z0 was the impedance of free space, Zin was the input impedance of the absorber, c was the velocity of EM wave in free space, f was the frequency of electromagnetic wave, and d was the thickness of the absorber. Therefore, MA materials needed to achieve strong absorption of microwaves under the synergistic effect of dielectric and magnetic losses.

Common magnetic loss materials were mainly Fe [19], Co [20], Ni [21] and their compounds because of high saturation magnetization and high Snoek’s limit22. Unfortunately, these conventional magnetic loss materials had high density and poor chemical resistance. Compared with magnetic loss materials, carbon materials had the advantages of low density, chemical corrosion resistance and good dielectric properties23, 24. According to the expression of skin depth (δ = \(\sqrt{1/\pi {\mu }_{0}{\mu }_{i}\sigma f}\)), the large conductivity of carbon would lead to low skin depth, so that most microwaves could not enter the interior of the MA materials 25. The researchers combined magnetic particles with carbon to improve MA capabilities. Magnetic materials could adjust the impedance matching of carbon materials and increase the phase interfaces. Furthermore, the carbon materials could enhance the chemical corrosion resistance of magnetic materials and reduce their density.

In addition to the composition of MA materials, the microstructure also deeply affected the properties of MA materials. Among all kinds of structures, 1D structures had attracted much attention because of their unique shape anisotropy and spatial constraint effect. Moreover, 1D carbon nanofibers had the quantum size effect and macroscopic quantum tunneling effect, and the split energy level spacing was just within the energy range of microwave to increase the absorption pathway of microwave26–28. The synergistic interaction between 1D materials and other structures (flower-like structure, core–shell structure and porous structure) could also further improve the microwave absorption performance. In addition, the preparation of 1D microwave absorption materials was also very convenient, and electrospinning and hydrothermal method were common preparation methods.

Due to the excellent properties of 1D carbon-based magnetic composites, the number of publications had increased significantly in the last five years, as shown in Fig. 1. Herein, the research progress of 1D carbon-based magnetic composites for MA was reviewed. The results showed that a reasonable microstructure could increase the interface polarization and microwave reflection. Next, 1D carbon-based magnetic composites with different structures were discussed to explore the influence of structure on MA properties. Meanwhile, challenges for the development of 1D carbon-based magnetic composites were pointed out to design better MA materials.

Figure 1
figure 1

Number of the publications on a "MA" and "1D or nanofibers or nanochain or nanowire" and "Fe or Co or Ni" b "MA and carbon" and "1D or nanofibers or nanochain or nanowire" and " Fe or Co or Ni" searched by Web of Science

Full size image

Magnetic carbon fiber composites

Carbon fibers were widely used in MA due to their high dielectric loss and low density29. By combining carbon fibers with other magnetic loss materials, the composite absorbers would have complex interfaces and produce interfacial polarization to improve the MA performance30. Through the modification of carbon fiber, the strong absorption of microwaves by 1D magnetic carbon materials was achieved. The structure of common 1D carbon fiber materials is shown in Fig. 2.

Figure 2
figure 2

Reproduced with permission from Ref. [31,32,33,34]

Various structures of carbon-based magnetic composite fibers.

Full size image

Porous composites

The porous MA materials could make the microwave reflect several times to extend the transmission path of microwaves within the material, thus accelerating the attenuation of microwaves. Proper porosity could improve the impedance matching and reduce the dielectric constant and the density of MA materials [35]. In addition, a higher gas–solid interface enhanced the interfacial polarization of MA materials and then improves the attenuation ability of microwaves.

With the increase in the papers on porous carbon fiber, various preparation methods of porous carbon fiber were reported [36,37,38]. Hard template method was often used for the preparation of macroporous carbon fibers. Liu et al. [39, 40] used SiO2 as a template to prepare 1D materials by electrospinning and then etched away SiO2 with alkaline solution to obtain porous structures. MA materials with large and graded pore structures could be prepared by this method. Chen et al. [41] obtained mesoporous carbon fiber (MCF) by electrostatic spinning and subsequent HF etching. The pore structure of MCF could be adjusted by controlling the amount of tetraethyl orthosilicate. When the amount of tetraethyl orthosilicate was 0.50 mL, the pore structure of the prepared MCF was the most suitable for MA. When the thickness of MCF was 2.6 mm, the minimum reflection loss (RLmin) was  −61.2 dB and the maximum effective absorption bandwidth (EAB) was 4.9 GHz.

High-temperature degradation materials were added to the raw materials for electrospinning, and the porous structure was formed by subsequent heat treatment. Abdalla et al. [42] successfully prepared 1D MA materials with microporous distribution by using PVP as a pore-forming agent. Using the thermal decomposition properties of PMMA, Zuo et al. [43] prepared porous magnetic carbon nanofibers (P-CNF/Fe) to effectively absorb low-frequency microwave, and explored the effect of porous nanostructures on MA. BET results showed that the pore size ranges from 30 to 100 nm. When the thickness of the materials was 4.1 mm, the RLmin was – 44.86 dB and the maximum EAB was 3.28 GHz in the range of 12.96–16.24 GHz. Soft template method did not need complex equipment and had the advantages of simple method, convenient operation, low cost and so on, so it had been widely concerned. Using vegetable oil as a pore-forming agent, different evaporation rates of vegetable oil and DMF were used to form pore structure. Li et al. [44] prepared Ni/C porous nanofibers by using this method, and BET results showed that the pore sizes were mainly mesopores and micropores.

MOF materials had the advantages of low density, high porosity and large specific surface area, and its porous structure could be retained after carbonization [45]. Therefore, MOF materials were often used to prepare porous materials. Chen et al. [46] successfully prepared 1D porous FeIII-MOF-5-derived/carbon fibers composites (FMCFs) by using the method of electrospinning, and the synthesis process of FMCFs is shown in Fig. 3a. As can be seen in Fig. 3b, c, the FeIII-MOF-5 in the fiber still retained the polyhedral structure. BET results showed that the specific surface area and pore volume of FMCF increased with the addition of FeIII-MOF-5 (Fig. 3f, g). MOF material was an important method to prepare porous MA materials. Using CoZn-MOF as the precursor, Liao et al. [47] prepared porous Co/ZnO/C (CZC) microrods by pyrolysis with N2. Figure 3d, e shows that CZC microrods had a high porosity structure. In addition, BET results showed that the surface area increases sharply with the increase in pyrolysis temperature (Fig. 3h, i), which was attributed to the release of 2-methylimidazole and some small molecules. Using ZIF-67-coated carbon fibers as precursors, Quan et al. [48] embedded Co3O4-modified nitrogen-doped carbon array (triangular nanoplates) on the surface of carbon fibers in the carbon paper (NC-Co3O4/CP). The carbon fiber had a porous structure with pore size of micropores. When the thickness was 1.5 mm, the RLmin of NC-Co3O4/CP could reach – 34.34 dB. Therefore, the porosity of MA materials could be controlled by adjusting the MOF amount and pyrolysis temperature to achieve strong MA.

Figure 3
figure 3

Reproduced with permission from Ref. [46, 47]

a, b BET information of FMCFs, c N2 absorption isotherms and d pore size distributions of CZC-600, CZC-700 and CZC-800.

Full size image

Flower-like composites

The flower-like structure was usually fabricated by the irregular growth of the nanosheets to enhance microwave attenuation, which had the following advantages: (1) 1D flower-like structure could form conductive network to provide more conductive paths for electron transport. (2) The specific surface area of the MA materials could be greatly increased by coating the flower-like structure on the surface of the 1D materials. (3) Rough surfaces formed by the flower-like structure could induce multiple reflections and scatterings of microwave [49, 50].

The electromagnetic parameters of 1D flower composites could be adjusted by changing the composition and structure of nanoflowers. Tong et al. [51] synthesized Ni nanowire by magnetic field induction method and then prepared Ni@Co3O4/RGO by annealing and hydrothermal reaction. Flower-like structure was formed by a large number of Co(OH)2 nanosheets. After annealing, Co3O4 nanosheets with flower-like structure were obtained, as illustrated in Fig. 4a. When the thickness was 1.6 mm, the maximum EAB was 4.62 GHz, as shown in Fig. 4c. In a different attempt, Tong et al. [52] grew flower-like MoS2 on the surface of Fe3O4/Fe@C nanofibers, and the microscopic morphology is investigated in Fig. 4b. Fe3O4/Fe provided magnetic loss, and C provided dielectric loss. MoS2 was a kind of semiconductor material with non-centrosymmetric structure, and its flower-like structure increased the microwave attenuation ability of microwave absorbers. When the thickness of the absorber was 2.24 mm, the RLmin of Fe3O4/Fe@C@MoS2 nanofibers was -53.79 dB at 11.12 GHz, and the corresponding EAB reached 4.4 GHz, as illustrated in Fig. 4d.

Figure 4
figure 4

Reproduced with permission from Ref. [51, 52]

SEM images of a Ni@Co3O4 nanowires, b Hierarchical Fe3O4/Fe@C@MoS2 core–shell nanofibers. 3D RL plots of c Ni@Co3O4/RGO composites, d Hierarchical Fe3O4/Fe@C@MoS2 core–shell nanofibers.

Full size image

The multicomponent MA material had a larger specific surface area than the single-component MA material, and its electromagnetic parameters were easier to adjust. Ma et al. [53] prepared 1D flower-like ZnFe2O4@SiO2@C@NiCo2O4 (ZSCNC) nanochains with core–shell structure to achieve broadband MA, as shown in Fig. 5a. Figure 5b, c reveals the unique aspect ratio and shape of the ZSCNC. At a loading of 30 wt% and a thickness of 2.39 mm, the RLmin was -54.29 dB at 11.14 GHz and the EAB was 5.66 GHz (11.94–17.60 GHz), as investigated in Fig. 5d. Huang et al. [54] prepared ZnFe2O4@SiO2@C@NiO (ZSCN)/rGO composites by a multi-step synthesis method, and the flower-like structure was formed by NiO nanosheets. When the thickness of ZSCN/rGO was 2.45 mm, the RLmin was – 66.38 dB at 9.58 GHz and the EAB was 3.44 GHz (8.12–11.56 GHz). The magnetic field induction method greatly simplified the difficulty of synthesizing magnetic nanochains, and the flower-like structures on the magnetic chains greatly increased the absorption area. The flower-like structure was caused by the irregular growth of NiO nanosheets. Magnetic field induction was a common method for preparing 1D magnetic nanochains. Liao et al. [55] field-induced distillation-precipitation polymerization and high-temperature carbonization. The materials had a RLmin of – 52.5 dB at 13.2 GHz with a thickness of 2.23 mm, and the EAB was 4.98 GHz (10.10–15.08 GHz).

Figure 5
figure 5

Reproduced with permission from Ref. [53]

a Schematic illustration of the formation process of 1D flower-like ZSCNC nanochains, SEM images of b 1D flower-like ZSCNC nanochains, c TEM images of 1D flower-like ZSCNC nanochains, (d) RL curves of the samples loaded with 1D flower-like ZSCNC nanochains.

Full size image

Core–shell composites

Core–shell structure was an assembly structure that orderly combined one material with another through chemical bonds or other forces. Recent studies had also shown that core–shell heterostructures played an important role in improving MA by means of space charge polarization and quantum confinement effect [56, 57]. In addition, core–shell nanocomposites also helped to combine the advantages of different materials to obtain new chemical and physical properties [58].

The common MA mechanism of carbon-based magnetic composites with 1D core–shell structure is investigated in Fig. 6. 1D core–shell microfiber composites attenuated microwave energy by forming a continuous conductive network, inducing current and triggering multiple reflections. Moreover, the interface between the core and shell of the MA fiber materials became the polarization center to provide the interfacial polarization [59, 60].

Figure 6
figure 6

Reproduced with permission from Ref [60]

Schematic illustration of the possible absorbing electromagnetic wave mechanism in C/NiP microfibers composites.

Full size image

Carbon fibers could be easily prepared by electrostatic spinning and carbonization technology. Therefore, the preparation of 1D core–shell materials based on electrostatic spinning technology had been widely reported. Liu et al. [61] prepared MA materials with carbon fibers and carbonyl iron composites by chemical vapor deposition. The carbonyl iron layer was uniformly adhered to the surface of carbon fiber as a shell. The RLmin of the absorber was – 21.5 dB at a thickness of 2 mm. The core–shell structure of carbonyl iron@carbon fiber was too single, so that the MA performance was not ideal. On this basis, magnetic nanoparticles could be used to modify the core–shell materials to obtain better MA properties. Qiao et al. [62] prepared Ni-embedded TiO2/C core–shell terpolymer nanofibers using a combination of electrospinning and heat treatment, with carbon fiber as the core and Ni-coated TiO2 layer as the shell (Fig. 7a, b). The extremely large interface, conductive network and rough surface structure were the main sources of electromagnetic wave attenuation. Figure 7c shows that the RLmin of Ni-embedded TiO2/C could reach – 74.5 dB when the thickness was 2 mm.

Figure 7
figure 7

Reproduced with permission from Ref. [62]

a, b HR-TEM images of T2N1, c 3D RL representation of T2N1.

Full size image

The dielectric shell materials had good chemical corrosion resistance to protect the magnetic core materials, and the core–shell structure combining with the magnetic core and dielectric shell was often used to prepare MA materials. Cui et al. [63] successfully prepared FeCo/PVP core–shell nanochains@GO composites (FCPG) by reduction method. It could be clearly seen that PVP was uniformly coated on FeCo magnetic nanochains. This core–shell structure enhanced the interfacial polarization of FeCo nanocrystals and the stability of magnetic materials to acquire good MA properties. The RLmin of FCPG was – 52.5 dB at 13.2 GHz with a thickness of 2.23 mm, and the maximum EAB was 4.9 GHz.

In addition to the above structures, carbon nanofiber@core–shell nanoparticles were also important 1D core–shell materials. The multiple components and well-designed microstructures account for their excellent performance. Meng et al. [64] prepared C/Co nanofibers by simple electrostatic spinning and in situ reduction. Co nanospheres promoted the graphitization of carbon and formed core–shell Co/C on the surface of carbon fibers. Therefore, C/Co nanofibers had excellent corrosion resistance. When the thickness of C/Co nanofibers was 3.5 mm, the RLmin was – 58.8 dB at 7.4 GHz, and the EAB was 12.9 GHz.

Other composites

Core@void@shell structure had the advantages of low density, large specific surface area and adjustable gaps, so core@void@shell was a highly efficient electromagnetic absorbing structure. Qiao et al. [65] prepared Fe3O4@void@SiO2 and Fe3O4@void@SiO2@PPy (FVSP) MA materials with yolk-shell structure and compared their absorbing properties. Because PPy was a good dielectric loss material, FVSP nanochain had stronger MA performance than FVS nanochain.

Coating magnetic nanoparticles inside carbon nanotubes could greatly increase the corrosion resistance and interface area of materials. Xu et al. [66] successfully coated Fe/Fe3C nanospheres inside nitrogen-doped CNTs (Fe/Fe3C@NCNTs), resulting in a unique pea-like structure. The materials were prepared by heating FeCl3·6H2O and melamine in an inert atmosphere at a ratio of 1:1, and its chemical composition and microstructure were affected by pyrolysis temperature. Therefore, the electromagnetic parameters of the materials could be adjusted by controlling the pyrolysis temperature, and the rule was that the dielectric loss decreased with the increase in pyrolysis temperature. When the pyrolysis temperature was 600 °C and the thickness was 4.97 mm, the RLmin was – 46 dB at 3.6 GHz. Rough surfaces could increase the absorption properties of MA materials [67]. In Fig. 8a, Wang et al. [68] prepared magnetic Ni@C nanofibers with branched multi-walled CNTs (MWCNTs) by a two-step method of electrostatic spinning and in situ pyrolysis. In Fig. 8b, the branch of carbon nanofiber was MWCNTs grown under the catalysis of Ni crystal. This multi-branched 1D structure not only increased the surface area of MA materials, but also easily formed 3D conductive network. When the thickness was 1.5 mm, the RLmin of Ni@C nanofibers was – 53.2 dB and the EAB was 5.6 GHz (Fig. 8c).

Figure 8
figure 8

Reproduced with permission from Ref. [68]

a Schematic diagram of the preparation process of Ni@C nanofibers, b SEM images of Ni@C nanofibers, c 3D-contour maps of RL values of Ni@C nanofibers.

Full size image

The 1D MA materials with unique spiral shape could be crosslinked into networks to increase cross-linked polarization and conduction losses. Using Fe/Sn-nanoparticles as catalysts, Zuo et al. [69] prepared nanohelical coils by chemical vapor deposition. ZIF-67 was coated on their surface, and then Co@N-Carbon/CNCs was synthesized by pyrolysis of ZIF-67/CNCs at different temperatures. The results showed that the EAB of Co@N-Carbon/CNCs reached 5.5 GHz when the filling rate was only 6 wt%.

With the development of 1D carbon fibers, more and more 1D carbon fibers had been used to prepare MA materials. The RLmin and the EAB of these similar materials are shown in Table 1.

Table 1 The MA performance of magnetic carbon fiber composites
Full size table

Magnetic CNTs composites

CNTs were seamless tubular structures consisting of 2D graphite sheets coiled around a central axis, which could be single-walled or multi-walled [72]. Among them, single-wall CNTs had higher conductivity than multi-wall CNTs at the same content due to their smaller pipe diameter and larger aspect ratio [73, 74]. As a typical 1D nanomaterials, CNTs had the advantages of corrosion resistance, high electrical conductivity, great high-temperature strength, and low density [75]. Magnetic CNTs composites retained the advantages of CNTs to a certain extent and improved the impedance matching of CNTs. The electromagnetic parameters of magnetic CNTs composites could be easily adjusted by combining them with different magnetic materials [76]. Therefore, magnetic CNTs composites showed great application potential in MA.

CNTs/alloy composites

Only using CNTs as MA materials lacked magnetic loss capability, so CNTs needed to be combined with magnetic alloy materials to improve impedance matching [76]. Magnetic alloy materials could generate a large number of phase interfaces, and the magnetic loss capability of microwave absorbers could be tuned by adjusting the composition of the alloy [77, 78]. Some of the nanoalloys were distributed inside the CNTs, so the CNTs could reduce the agglomeration of nanoalloys very well. In addition, CNTs provided more electron transport pathways for 1D MA materials.

Nitrogen doping could promote interfacial and dipole polarization, and nitrogen-doped CNTs (N-CNT) composites had been widely reported in recent years [79, 80]. Zhou et al. [81] designed and prepared a novel porous iron-based nanocomposite containing Fe3C, Fe3N, Fe (FexCyNz) and N-CNT. Some of the magnetic nanoparticles were encapsulated inside the CNTs to reduce the agglomeration of magnetic nanoparticles. The RLmin of FexCyNz/N-CNT-1000 obtained under the pyrolysis condition of 1000 °C was – 25.1 dB at a thickness of 4 mm. When the thickness range was 2–5 mm, the EAB was 8.2 GHz. Magnetic metal particles were often filled into the interior of CNTs to form rich interfaces between metal particles and NCNTs. In order to explore the MA properties of different magnetic metal particles combined with NCNTs, Ning et al. [82] successfully prepared the M@NCNTs composite materials (Fig. 9a–c), where M was Fe, Co, Ni nanoparticles. The curlier CNTs effectively avoided agglomeration of metal nanoparticles to obtain good MA performance. When the thickness was 3.2 mm, the RLmin of Fe@NCNTs composite was – 30.43 dB, and the EAB reached 5.7 GHz (Fig. 9g). Co@NCNTs and Ni@NCNTs also had satisfactory EAB of 4.08 GHz and 4.72 GHz, respectively. Similarly, Yang et al. [83] successfully prepared γ-FeNi/CNT nanocomposite particles by the wet chemical method, and γ-FeNi nanoparticles were uniformly distributed on the inside and outside of CNTs (Fig. 9d, e). When the sample thickness was 1.6 mm and the loading was 2 wt%, the RLmin of γ-FeNi/CNTs at 16.5 GHz was – 15.4 dB. Qi et al. [84] synthesized Fe/CNTs with core–shell structure by chemical vapor deposition, using hollow Fe2O3 particles as catalyst. The experiment used C2H2 to reduce Fe2O3 to Fe and then carbonized to form Fe3C. Fe3C could effectively catalyzed the growth of core–shell carbon nanohybrid. As shown in Fig. 9f, Fe nanoparticles were uniformly encapsulated in CNTs. Due to the special structure of Fe/CNTs, the RLmin was about – 40.15 dB at the thickness of 1.5 mm, as shown in Fig. 9h. These excellent properties demonstrated that CNTs/alloy composites with 1D structures were efficient, sustainable MA materials.

Figure 9
figure 9

Reproduced with permission from Ref [82,83,84]

TEM images of a Fe@NCNTs, b Co@NCNTs and c Ni@NCNTs, d Local CNTs well coated with γ-FeNi alloys, e Local CNTs filled in with γ-FeNi alloys, f Fe/CNT hybrid. g RLs of Fe@NCNTs/wax with 10 wt% loadings under different thicknesses. h Absorption properties of Fe/CNTs composite with different thicknesses.

Full size image

Using CNTs as templates, MOF materials were grown on CNTs and further carbonized to prepare CNTs/alloy composites. Wang et al. [85] successfully designed CNTs fiber/Ni-MOF interpenetrating structure by solvothermal method to enhance carrier mobility. At a thickness of 5 mm, the RLmin was – 24.32 dB. CNTs had a large number of surface atoms and unsaturated suspended bonds, which were favorable for interfacial polarization and multiple scattering. Multi-walled CNTs (MWNTS) were often used as fillers in MA materials because of their larger specific surface area. Yin et al. [86] used MWCNTs as raw materials to prepare Co–C/MWCNTs composites through pyrolysis precursor ZIF-67/MWCNTs. Under the induction of magnetic field, Co–C/MWCNTs had certain orientation in the paraffin matrix, and the effective dielectric constant was improved. When the filling rate of Co–C/MWCNTs was 15 wt%, the RLmin was – 48.9 dB.

Among the traditional preparation methods of CNTs, Fe [87], Co [88] and Ni [89] had been proved to be good metal catalysts for the preparation of CNTs. In addition, MOF materials (ZIF-8 [90], ZIF-67 [91], etc.) were easily reduced to metallic particles during carbonization, and these metallic particles acted as catalysts to convert MOF-derived carbon into CNTs. Therefore, researchers tended to use the method of in situ generation of CNTs to prepare MA materials. Wang et al. [92] prepared NiCo alloy/carbon nanorods decorated with CNTs by pyrolysis, and CNTs were grown in situ under the catalysis of NiCo nanoparticles. The rod-like structure and uniform mesoporous structure of NiCo-MOF-74 were well preserved in the NiCo alloy/carbon composites. Thus, by controlling the Ni/Co ratio in the precursor, the microstructure of the MA materials could be easily tuned. In addition, CNTs on the fiber surface formed a conductive network to increase the attenuation capacity of materials. When MOFs were used as templates for the preparation of CNTs, dicyandiamide or melamine was often added as an additive [93]. Liu et al. [94] prepared Fe/Fe3C@NCNTs by placing a ceramic vessel loaded with dicyandiamide in the air inlet during Fe-MOF pyrolysis. Bamboo-like NCNTs were coated on Fe/Fe3C nanospheres. Fe/Fe3C@NCNTs had a large number of defects, voids and interfaces, which greatly improved the polarization loss. When the thickness was 1.6 mm, the RLmin was – 50.8 dB. Xue et al. [95] successfully prepared Co/MnO/CNTs composites by pyrolysis of core–shell MnO2/ZIF template. Under the catalysis of metal particles, CNTs grew in large quantities, and Co/MnO/CNTs composites showed caterpillar lamellar structure. The CNTs were intertwined with each other to form a complex mesh structure, which was favorable to generate multiple polarizations and resonances. When the thickness was 1.97 mm and the filling amount was 35 wt%, the RLmin was – 58 dB and the EAB was 5.36 GHz. Without the use of templates, Wu et al. [96] self-assembled MXene and CoNi-bimetallic MOF into 1D heterostructural MA materials (MXene fibers@MOF-derived CNTs, MMC). By controlling the loading of MOFs on MXene fibers, the growth of CNTs could be effectively controlled, and the electromagnetic parameters could be regulated. The 1D structure of MMC was entangled to form a spatial network, which was beneficial to the conductivity of matrix. The thickness of MMC was only 1.6 mm, the RLmin was –  51.6 dB, and the maximum EAB was 4.6 GHz. In recent years, absorbers related to CNTs/alloy composites are shown in Table 2.

Table 2 The MA performance of CNTs/alloy composites
Full size table

CNTs/ferrite composites

Ferrite was widely used as MA materials with high permeability, high resistivity and good impedance matching performance [97, 98], mainly through self-polarization, hysteresis loss, domain wall resonance and natural resonance to lose electromagnetic wave. Impedance matching of microwave absorbers could be improved by compositing ferrite and CNTs[99]. Therefore, the combination of CNTs and ferrite was an effective method to improve MA performance.

Spinel ferrite, as a common crystal type, had high magneto-crystal anisotropy and saturation magnetization. Therefore, it was widely used in MA field [100,101,102,103]. Its molecular formula was MeFe2O4, and Me was metal ion [104]. Fang et al. [105] mass-produced NiCo-SWCNTs in helium by DC arc discharge and prepared spinel CoFe2O4 nanocrystals by sol–gel spontaneous combustion. Then NiCo-SWCNTs/CoFe2O4 nanocomposites were synthesized by mechanical mixing. As illustrated in Fig. 10a, b, many nanoparticles were uniformly attached to the long filamentous carbon. When the thickness of NiCo-SWCNTs/CoFe2O4 was 1.80 mm, the EAB was 7.10 GHz (Fig. 10e). In addition to straight CNTs, spiral CNTs were also synthesized by the researchers. Feng et al. [106] prepared CoFe2O4 (CFO) nanoparticles coated with coiled CNTs (CCNTs) to form CCNTs/CFO composites. As shown in Fig. 10c, d, CFO was grown on the surface of helical CNTs with a large number of interfaces that are favorable to dipole polarization and interfacial polarization. At a thickness of 3 mm, the EAB of CCNTs/CFO was 4.4 GHz, as shown in Fig. 10f.

Figure 10
figure 10

Reproduced with permission from Ref. [105, 106]

A typical a SEM image and b TEM image of the as-grown NiCo-SWCNTs. The inset in (b) is a HR-TEM image taken from the square region marked in (b). c SEM image of CCNTs composite with ωCCNT/ωCFO of 0.6. d TEM image of CCNTs. e The RL for the NiCo-SWCNTs nanocomposites at different thicknesses. f The RL of CCNTs/CFO composite with ωCCNT/ωCFO of 0.6.

Full size image

The distribution of magnetic particles on the fiber surface also affected the MA performance. Li et al. [107] prepared nano-Fe3O4 compact-coated CNTs (FCCs) and Fe3O4 loose-coated CNTs (FLCs) by the solvothermal method. Fe3O4 nanoparticles were evenly distributed on the surface of the two types of MA materials. In addition, CNTs and Fe3+ had the appropriate mass ratio in FCCs and FLCs to achieve both strong dielectric and magnetic losses. The MA performance of FCCs was superior to that of FLCs, and the RLmin of FCC was – 43 dB at a thickness of 1.5 mm. Fe3O4 was a traditional magnetic loss material with high saturation magnetization, so it had been widely used in MA field [108]. Ren et al. [109] synthesized CNTs loaded with Fe3O4 by chemical vapor deposition method. The MA capacity of a single iron oxide on CNTs was limited. When the thickness of the composites was 3 mm, the RLmin was  −35.9 dB at 7.12 GHz. When the thickness was 1.5 mm, the EAB was 4.32 GHz. Similarly, Fe3O4 nanoparticles were used to enhance the magnetic loss capacity of microwave absorber. Alanagh et al. [110] prepared Fe3O4/MWCNTs@CFs nanocomposites by introducing MWCNTs and Fe3O4 nanoparticles into carbon fibers through electrophoretic co-deposition process. Fe3O4/MWCNTs@CFs nanocomposites had a large number of mesoporous structures. Compared with the two-component MA materials, the three-component Fe3O4/MWCNTs@CFs had more interfaces, producing more interfacial polarization. The EAB was about 8 GHz at a thickness of 2 mm and a filling amount of 20 wt%. Direct contacts between MWCNTs and Fe layers were prevented by amorphous carbon layers to inhibit inverse radiation and eddy currents. Wu et al. [111] first prepared MWCNTs@C by in situ polymerization and high-temperature carbonization and then obtained the MWCNTs@C@FexOy (Fe/α-Fe2O3) composite by subsequent heat treatment. MWCNTs interweaved into conductive networks both to increase the electron conduction pathway and to form a large number of interfaces with magnetic nanoparticles. The materials had an EAB of 4.6 GHz at a thickness of 1 mm. More examples of CNTs/ferrite composites are summarized in Table 3.

Table 3 The MA performance of CNTs/ferrite composites
Full size table

Magnetic biomass-derived carbon composites

Biomass resources were natural carbon-based carriers, which had the characteristics of various types, wide sources, good porosity, and rich carbon content [115, 116]. The preparation of carbon-based MA materials from biomass had become a new sustainable and low-cost method [117]. Natural biomass usually had fine microtubule channels and heterostructures, which was difficult to achieve by conventional chemical synthesis. Using natural and unique microstructure characteristics to optimize the MA performance was an effective way to deal with the growing electromagnetic interference problem. Cotton was often used as a precursor of carbon fiber in the field of MA due to its wide availability and low cost. Li et al. [118] successfully prepared Co/C fiber by impregnation and calcination using cotton as carbon source. The MA properties of Co/C fibers with different Co contents were investigated. The maximum EAB of Co/C fibers reached 6.7 GHz at a thickness of 2.5 mm. Using cotton as the precursor of carbon fiber, Yin et al. [119] prepared CFs@H-Fe3O4/CoFe hybrids using a simple two-stage method of hydrothermal and calcination. The Fe3O4 nanospheres on the surface of carbon fiber had hollow structure. CFs@H-Fe3O4/CoFe had good low-frequency absorption capacity. The RLmin of CFs@H-Fe3O4/CoFe was  −40.85 dB with a thickness of 3.5 mm.

However, biomass-derived 1D carbon-based magnetic composites prepared by impregnation and pyrolysis of metal ions often produce agglomeration. To explore more complex structures, the researches tried to combine the MOFs with carbon fibers. Using cotton and MOF as raw materials, Yang et al. [120] prepared hierarchical carbon fiber coated with Co/C dodecahedron nanoparticles and villus-like CNTs (HCF@CZ-CNTs). The apparent density of the material was as low as 0.0198 g/cm3. When the thickness was 2.0 mm, HCF@CZ-CNTs exhibited the RLmin value of -53.5 dB at 7.8 GHz and the EAB of 8.02 GHz. Li et al. [121] prepared Fe-MOFs/biomass-cotton-derived Fe@nano-porous carbon@carbon fiber (Fe@NPC@CF) composites by in situ synthesis and thermal decomposition. The results showed that the composites were well combined with the Fe@NPC composites uniformly dispersed on the carbon fiber. The composite material had a strong MA performance. When the matching thickness was 2.5 mm, the RLmin was – 46.2 dB, and the EAB was 5.2 GHz. In recent years, ZIF-67 had been used to prepare Co/C composites because of its high porosity, simple preparation and large pore size. Zhao et al. [122] prepared carbon-cotton/Co@NPC composites by pyrolysis cotton/ZIF-67, as shown in Fig. 11a. The carbon-cotton/Co@NPC retained the dodecahedron shape of ZIF-67 well (Fig. 11b, c), and its unique hierarchical and porous structure made the materials have good MA properties. When the thickness was 1.65 mm, the EAB was 4.4 GHz and the RLmin was – 51.2 dB (Fig. 11d).

Figure 11
figure 11

Reproduced with permission from Ref. [122]

a Schematic illustration of the synthesis process of carbon-cotton/Co@NPC composites, b carbon-cotton/Co@NPC composites, c Amplified image of (b), d The RL values versus different thickness of carbon-cotton/Co@NPC.

Full size image

A large amount of biomass waste was generated every year around the world, and these biomass waste could be used as raw materials to prepare various carbon-based composite materials, such as anode materials [123], biofuels [124], and microwave absorbers [125, 126]. At present, biomass waste as green and cheap carbon precursor had received extensive attention. Cellulose was a cheap and readily available natural material on the earth and was found in many agricultural residues (rice husks, wheat straw, bagasse, etc.). Using these agricultural residues as precursors of carbon materials, waste could be turned into treasure. Wei et al. [127] prepared Ni/C 1D MA materials by carbonization treatment with cellulose as raw materials. The RLmin of Ni/C composites at a thickness of 4.5 mm was – 53.8 dB (at 4.72 GHz). Biomass was used as a precursor for 1D carbon materials mainly due to the existence of natural structure (hollow structure [128], porous structure [129], layered fiber structure [130]). Zhang et al. [128] selected raw cattail tube with natural hollow structure as carbon precursor and synthesized MoS2@Fe3O4@CT hybrid by hydrothermal method, carbonization, surface modification and electrostatic self-assembly. Fe3O4 nanoparticles were surrounded by MoS2 nanosheets to form a flower-like core–shell structure. The RLmin of 1D MoS2@Fe3O4/CT hybrids was – 62 dB at 2.09 mm, and its EAB was 6.8 GHz at 2.35 mm.

The porous structure of biomaterials promoted the entry of metal ions into the MA materials and provided a large number of active sites for the growth of magnetic materials [131]. Liu et al. [129] prepared lightweight porous Fe3O4/CF composites by in situ growth and graphitization using sugarcane bagasse as a low-cost carbon source. The Fe3O4/CF composites retained the natural porous structure of bagasse to obtain good MA capacity. Fan et al. [132] impregnated reed with Fe(NO3)3 and carbonized to prepare Fe@Fe3C/C nanocomposites. Due to the capillary effect of the porous microstructure, Fe3+ was better impregnated into reed stalks. After calcination, the interfacial reaction between Fe and C formed Fe@Fe3C with a core–shell structure. When the thickness was only 1.43 mm, the EAB was 4.57 GHz. In addition, the biological waste fibers produced by the plane tree polluted the living environment of human beings and also caused waste of biomass resources. Wu et al. [133] prepared ultra-light magnetic helical porous carbon fibers (MHPFs) using biomass-inspired platanus (plane tree) fibers as the matrix (Fig. 12a). They developed a catalytic self-deposition (CSD) technique to achieve the confined growth of Ni nanoparticles in CNTs. Figure 12b, c reveals the carbon fibers were 1D hollow bamboo node-like morphology, and its surface was smooth with twisted texture. When the thickness of MHPFs-2.5 was 3.4 mm, the RLmin was – 55.39 dB.

Figure 12
figure 12

Reproduced with permission from Ref. [133]

a Synthesis diagram of MHPFs. SEM images of b MHPFs-5, c HR-TEM of CNTs deposited on MHPFs-2.5.

Full size image

In addition to plant fibers, natural collagen fibers extracted from animal bodies could also be used as precursors of carbon materials. Using natural collagen fiber as biological template, Wang et al. [130] successfully prepared ferromagnetic hierarchical carbon nanofiber bundles (FHCNBs). Furthermore, the EAB and absorption intensity of FHCNBs could be effectively adjusted by changing the ferromagnetic materials (Fe3O4, Fe3N, α-Fe, FeTiO3/Fe3O4 and ZrO2/Fe3O4) embedded on FHCNBs. This hierarchical nanofibrous structure of FHCNBs not only provided a smooth channel for microwave penetration, but also induced multiple reflections and scattering, thus obtaining excellent MA performance. Considering the superior properties of MOF materials, Zhao et al. [134] coated CoNi-bimetallic MOFs on bamboo carbon fiber and then obtained magnetic porous bamboo fiber/CoNi alloy (CN-ABF) by carbonizing. The unique porous structure of organisms greatly increased the specific surface area of carbon materials to increase the microwave attenuation of materials. Fang et al. [135] successfully synthesized carbon microtube and Co particle composite (CMT/Co) through Co salt impregnation and carbonization process. Porous carbon microtube/Co particle composite (PCMT/Co) were synthesized by treating carbon fibers with H3PO4 and Co salt. PCMT/Co and CMT/Co still maintained the hollow microstructure of cotton fiber after the carbonization process. Co nanoparticles were uniformly distributed on the cotton-derived hollow carbon tubes. The surface of PCMT/Co was rougher and more porous than CMT/Co due to the addition of phosphoric acid. When the thickness was 1.4 mm, the RLmin of PCMT/Co was -36.8 dB. The RLmin of CMT/Co was – 31.2 dB. Magnetic biomass-derived carbon composites showed great potential in MA, as listed in Table 4.

Table 4 The MA performance of magnetic biomass-derived carbon composites
Full size table

Conclusions and perspectives

Recent development of 1D carbon-based magnetic composites as a promising absorption candidate was proposed in this work. The outstanding advantages of different 1D materials are summarized systematically, which provides a basis for the future study of 1D magnetic carbon-based microwave absorption materials. Common 1D carbon included carbon fiber, CNTs, and biomass-derived carbon, which had unique advantages in combination with magnetic materials. In the case of carbon fiber, the 1D porous structure reduced the density of MA materials and increased the times of microwave reflection. The irregular growth of the nanosheet formed a 1D flower-like structure, which greatly increased the specific surface area and the roughness of the absorbing materials. Core–shell heterostructure could generate space charge polarization to increase the polarization loss of MA materials. The preparation method of magnetic CNTs composites was flexible, and cross-linked CNTs provided abundant loss pathways for MA. The carbon fiber derived from biomass could retain the structure of biological materials itself, so it was suitable for preparing some structures that were difficult to be synthesized artificially. Different 1D structures had their own unique advantages for attenuating microwave energy. Therefore, a reasonable 1D structure was an important guarantee of MA performance.

Although significant progress had been achieved in 1D carbon-based magnetic composites, they still faced many challenges. Firstly, current MA materials tended to composite various materials. However, due to the lack of detailed research on the significance of these materials combinations, there were blind spots in the preparation of multicomponent composite MA materials. The preparation of multicomponent MA materials was complicated and involved some toxic or polluting reagents. Secondly, the EAB of MA materials was usually concentrated in the high frequency region, while the study of low-frequency MA materials was not enough. Thirdly, most 1D carbon-based magnetic composites could not be used in harsh environments such as strong acid, high humidity and high temperature.

In order to solve the above problems, it was necessary to make full use of 1D carbon-based magnetic composites. The satisfactory microwave absorbers could be prepared by optimizing hybridization and constructing network structures. Furthermore, we believed that 1D carbon-based magnetic composites as outstanding absorbers would show greater potentiality. This review could provide guidance for the design of better 1D carbon-based magnetic composites in the future.