Kevlar 49 fibers (diameter 1420 D, length 6338 m kg⁻¹, density 1.44 g cm⁻³, degradation temperature 482 °C, elongation at break 2.46%) were purchased from DuPont (Wilmington, DE, USA). Dimethyl sulfoxide (DMSO) was supplied by Shanghai Aladdin Biochemical Technology Co., Ltd (Shanghai, China). Aluminum chloride (AlCl₃), ferric nitrate nonahydrate (Fe(NO₃)₃·9H₂O), sodium oxalate and triethylamine, ammonium acetate (CH₃COONH₄), polyvinylpyrrolidone (PVP, M_n = 5800), potassium hydroxide (KOH), and absolute ethanol were all provided by Shanghai Macklin Biochemical Co., Ltd (Shanghai, China). All these chemicals were used without further purification.

Firstly, the aramid nanofibers (ANFs) solution was prepared according to a previous report [26]. 1.0 g of para-aramid fiber (Kevlar) was dispersed in 150 mL of DMSO, followed by the addition of 1.5 g of KOH and stirring for 7 days to obtain a dark red para-aramid nanofiber solution (ANFs solution). Subsequently, 45 mL of DMSO and 160 mL of deionized water were slowly added to 15 mL of ANFs solution under stirring, and the mixture was stirred for 24 h to obtain ANFs dispersion. The dispersion was centrifugal washed several times with deionized water and then re-dispersed into water to form 40 mL of ANFs dispersion. In this dispersion, 0.242 g of Fe(NO₃)₃·9H₂O, 0.139 g of CH₃COONH₄, and 1.0 g of PVP were dissolved and magnetically stirred at 50 °C for 10 min. The mixture underwent ultrasonic dispersion at 100 W for 20 min. The resulting solution was transferred to a 100-mL Teflon-lined autoclave and reacted at 160 °C for 24 h. After the reaction, the product was centrifugal washed with deionized water and re-dispersed into water to form α-Fe₂O₃@ANFs dispersion. The corresponding preparation process is illustrated in Fig. 1a.

Al-α-Fe₂O₃ nanosheets were synthesized following a reported method [27]. Initially, 0.100 g of AlCl₃ and 1.212 g of Fe(NO₃)₃·9H₂O were dissolved in 30 mL of deionized water with magnetic stirring to form a transparent solution, while stirring, 4.5 mL of triethylamine was added, and the mixture was stirred for 40 min. The resulting solution was then transferred to a 100-mL Teflon-lined autoclave and reacted at 160 °C for 24 h. After the reaction, the red precipitate was collected through centrifugation and washed with deionized water several times. Subsequently, the synthesized ultra-thin Al-α-Fe₂O₃ nanosheets were dispersed in water, achieving a concentration of 24 mg mL⁻¹ for further use. The corresponding preparation process is illustrated in Fig. 1b.

The α-Fe₂O₃@ANFs dispersion and 60 mL of Al-α-Fe₂O₃ solution were mixed and ultrasonically dispersed at 100 W for 20 min. The resulting dispersion was vacuum-filtered to fabricate a wet free-standing composited membrane, which was subsequently dried using supercritical CO₂. The composite aerogel membrane was finally annealed at 700 °C for 4 h in an Ar atmosphere with a heating rate of 5 °C min⁻¹, resulting in Fe₃O₄-Fe nanoparticles/carbon nanofibers/Al-Fe₃O₄-Fe nanosheets (Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe) nanocomposites. Additionally, products prepared with 20, 40, and 50 mL of Al-α-Fe₂O₃ solution were denoted as Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe_(1:1), Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe_(1:2), Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe_(1:2.5), respectively. The corresponding preparation process is illustrated in Fig. 1c.

The morphologies and size of the samples were characterized by scanning electron microscope (SEM) using JSM-7500F (JEOL, Japan) with an operating voltage of 5 kV. Transmission electron microscope (TEM) images was carried out on FEI Tecnai F20 at an acceleration voltage of 200 kV. For TEM samples, the procedure entails dispersing the appropriate amount of sample into ethanol, subjecting it to ultrasonic dispersion at 800 W for 1 min using a pressure cell disruptor, dropping the mixture onto a copper mesh, drying, and then conducting TEM analysis. X-ray diffraction (XRD) patterns were obtained using Ultima IV (Rigaku, Japan) in the 2θ range of 10-80° at scanning rate of 2° min⁻¹. The hysteresis loops were performed by vibrating sample magnetometer (VSM) using LakeShore7404 (Lake Shore, USA). Thermogravimetric analysis (TGA) of the samples was carried by STA8000 (PerkinElmer, USA) from 25 to 800 °C at a heating rate of 10 °C/min under a nitrogen atmosphere. X-ray photoelectron spectra (XPS) of the samples was carried out on K-Alpha (Thermo Scientific, USA). The complex permittivity and permeability of the samples were determined using E5071C (Agilent, USA) vector network analyzer according the coaxial-line theory in a frequency range of 2-18 GHz. The absorbent materials were prepared by mixing the sample powder into paraffin matrix with a weight of 30 wt%. The mixture was then pressed into a circular ring (outer diameter: 7.00 mm, inner diameter: 3.04 mm, thickness: ~ 2.00 mm).

The preparation procedure of the Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites is delineated in Fig. 1. Initially, the aramid nanofiber solution was generated through the deprotonation of micron-scale aramid fibers (PPTA), leading to the breakage of hydrogen bonds and the formation of a polymer solution. Water was subsequently introduced to reconstruct hydrogen bonds, leading to the dispersion of ANFs. The necklace-like α-Fe₂O₃@ANFs was synthesized through an in situ growth process of α-Fe₂O₃ on ANFs via a facile hydrothermal reaction, as illustrated in Fig. 2c. In this reaction, PVP was utilized to establish catalytically active sites on the surface of ANFs, thereby facilitating the nucleation and in situ growth of α-Fe₂O₃ nanoparticles. The composite film was generated through VAF of the Al-α-Fe₂O₃ nanosheets and α-Fe₂O₃@ANFs dispersions, and the composite aerogel (α-Fe₂O₃@ANFs/Al-α-Fe₂O₃) was successfully fabricated via supercritical drying with CO₂ (Fig. 2e). The optical image of the surface of the α-Fe₂O₃@ANFs/Al-α-Fe₂O₃ aerogel is illustrated in Fig. S1a. The aerogels, denoted as α-Fe₂O₃@ANFs/Al-α-Fe₂O_{3(1:1, 1:2, 1:2.5)}, exhibit similar morphologies (Fig. S1b-d). Furthermore, ANFs aerogel alone exhibits a 3D network structure by the supercritical dry, as depicted in Fig. 2a. Subsequently, the α-Fe₂O₃@ANFs/Al-α-Fe₂O₃ aerogel underwent high-temperature calcination, yielding in 0D/1D/2D Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites.

Following high-temperature calcination, ANFs transform into CNFs, maintaining the integral 3D network structure (Fig. 2b). The α-Fe₂O₃@ANFs was transformed into Fe₃O₄-Fe@CNFs after exposure to high temperature, maintaining 3D network appearance (Fig. 2d). The ANFs in α-Fe₂O₃@ANFs act as a skeleton, effectively separating α-Fe₂O₃ and preventing extensive ceramic sintering. The Al-α-Fe₂O₃ nanosheets exhibits an ultra-thin 2D structure (Fig. 2g). Agglomerative sintering was observed when Al-α-Fe₂O₃ nanosheets alone were calcined into Al-Fe₃O₄ nanosheets in a reducing atmosphere (Fig. S2a). The Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites present a 3D network structure, with 0D Fe₃O₄-Fe nanoparticles, 1D carbon nanofibers, and 2D Al-Fe₃O₄-Fe nanosheets clearly visible (Fig. 2f), contributing to excellent electromagnetic wave (EMW) absorption performance in multiple dimensions. Al-Fe₃O₄-Fe was affixed to the Fe₃O₄-Fe@CNFs skeleton, forming a 3D aerogel structure. With decreasing Al-Fe₃O₄-Fe content in Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe_{(1:1, 1:2, 1:2.5)}, morphological deformation and pores emerge attributed to ceramic sintering (Fig. S2b-d). The phases, structures, and components of samples were further analyzed by XRD (Fig. 2h). The primary diffraction peaks at 24.3°, 33.2°, 35.8°, 41.0°, 54.2°, and 64.1° for α-Fe₂O₃@ANFs correspond to the crystal planes (012), (104), (110), (113), (116), and (300) of α-Fe₂O₃ (JCPDS card No. 33-0664), respectively. Similarly, the primary diffraction peaks at 30.1°, 35.4°, 43.1°, and 57.0° for Fe₃O₄-Fe@CNFs and Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe correspond to the crystal planes (220), (311), (400), and (511) of Fe₃O₄ (JCPDS card No. 87-0245), respectively. Notably, there are also 44.6°, 64.9°, and 82.3° correspond to the crystal planes (110), (200), and (211) of Fe (JCPDS card No. 06-0696) in Fe₃O₄-Fe@CNFs and Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe. The major diffraction peaks align with crystal faces α-Fe₂O₃ and Fe₃O₄. α-Fe₂O₃@ANFs/Al-α-Fe₂O₃ was calcined at 700 °C for 4 h to form Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe, with Ar atmosphere during calcination. No reducing gas was introduced during calcining process of Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites, and the nanocomposites underwent self-reduction through their own components. ANFs in the α-Fe₂O₃@ANFs/Al-α-Fe₂O₃ were transformed into carbon material CNFs in an inert gas, while CNFs react with α-Fe₂O₃ in the nanocomposites, reducing it to Fe₃O₄, and some Fe₃O₄ was further reduced to Fe. The XRD pattern in Fig. 2h can provide additional insights into this self-reduction process. For comparison, Al-α-Fe₂O₃ nanosheets were reduced to Al-Fe₃O₄ nanosheets. XRD patterns of Al-α-Fe₂O₃ and Al-Fe₃O₄ nanosheets are depicted in Fig. S3. Therefore, the Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites is a hierarchical structure composed of Fe₃O₄-Fe nanoparticles, CNFs, and Al-Fe₃O₄-Fe nanosheets. The corresponding structure diagram is shown in Fig. 2i.

PPTA transformed to ANFs with a 3D network structure, as illustrated in Fig. 3a. The surface of ANFs was subsequently modified by constructing catalytic active sites using PVP, which selectively adsorbed on the ANFs surface. Then, α-Fe₂O₃ nanoparticles were grown in situ at the catalytic site through a hydrothermal reaction (Fig. 3b). The locally magnified high-resolution TEM (Fig. 3c) clearly shows a crystal plane spacing of 0.25 nm, consistent with the (110) crystal plane of α-Fe₂O₃. Figure S4a illustrates the TEM image of α-Fe₂O₃@ANFs in other region, where the sample exhibits a necklace-like structure, and particles tightly bind to fibers. After calcination, the volume of Fe₃O₄-Fe@CNFs has contracted, resulting in finer fibers (Fig. S4b). The TEM image of Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe in the corresponding region (Fig. S5) displays 0D, 1D, and 2D components, aligning with the SEM image (Fig. 2f). The hierarchical heterogeneous structure of 0D Fe₃O₄-Fe nanoparticles, 1D carbon nanofibers, and 2D Al-Fe₃O₄-Fe nanosheets (Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe) was formed, as shown in Fig. 3d. It clearly shows the necklace-like structure of Fe₃O₄-Fe nanoparticles attached to CNFs. The corresponding element mapping in Fig. 3g-j and XRD patterns in Fig. 2h confirm the presence of Fe₃O₄-Fe nanoparticles in the composites, with distributed C and N elements in the samples. The N element originates from aramid fiber, and after calcination, ANFs transform into CNFs, resulting in N-doped carbon that significantly enhances the EMW absorption performance of the materials [28]. In the high-resolution TEM image of Fe₃O₄-Fe nanoparticles in Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe (Fig. 3e), the (311) crystal faces with a crystal face spacing of 0.25 nm are clearly visible. The selected area electron diffraction (SAED) image in Fig. 3f indicates that Fe₃O₄ is a single crystal with (311) and (444) crystal face spots. Additionally, the SAED image shows Fe (110), (200), and (211) crystal face spots, confirming the reduction of some Fe₃O₄ to Fe.

The magnetic properties of as-synthesized of Fe₃O₄@CNFs, Al-Fe₃O₄ nanosheets, and Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites were measured by vibrating sample magnetometer (VSM), as illustrated in Fig. 4a. Necklace-like Fe₃O₄@CNFs, Al-Fe₃O₄ nanosheets, and Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites exhibit superparamagnetic behavior with saturation magnetization of 48, 40, and 91 emu g⁻¹, respectively. The high saturation magnetization of Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe is attributed to Fe nanocrystals existing in Al-Fe₃O₄-Fe (Fig. 2h). Additionally, the presence of carbon defects in the samples was studied through Raman spectra (Fig. 4b). In general, the D band arises from defect/disordered carbons, and the G band is attributed to in-plane stretching vibration of sp² carbons [28,29]. Therefore, the I_D/I_G ratio of as-prepared nanocomposites can be used to characterize the defect degree of carbon components by Raman spectra. In Fig. 4b, CNFs, Fe₃O₄-Fe@CNFs, and Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe have I_D/I_G values of 1.01, 1.27, and 2.10, respectively. Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe shows the highest I_D/I_G due to Fe₃O₄-Fe and Al-Fe₃O₄-Fe promoting carbon defects, favoring EMW absorption [28]. In addition, the chemical structures and surface electronic states of Fe₃O₄-Fe@CNFs, and Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe were analyzed by XPS. As shown in Fig. 4c, the XPS spectra of Fe₃O₄-Fe@CNFs and Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe show the presence of Fe, O and C, while Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe has obvious Al element signal. In the O 1s spectrum (Fig. 4d), there are chemical bonds between the iron and oxygen atoms, corresponding to Fe-O and C=O/C-O bonds at 530.3 and 531.8 eV, respectively. In Fig. 4e, the high-resolution Fe 2p spectrum of Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe can be deconvolved into eight peaks. Peaks at 711.0 and 724.5 eV correspond to Fe 2p_3/2 and Fe 2p_1/2 of Fe³⁺, while peaks at 709.7 and 722.5 eV represent Fe 2p_3/2 and Fe 2p_1/2 of Fe²⁺, and other peaks are satellite peaks resulted from shake-up excitation of high-spin oxidized Fe species [30]. These results confirm the presence of Fe₃O₄, not γ-Fe₂O₃. XPS spectra of Fe 2p of Fe₃O₄-Fe@CNFs can be divided into peaks same as Fe₃O₄-Fe@CNFs (Fig. S6). To characterize the transition from α-Fe₂O₃@ANFs/Al-α-Fe₂O₃ to Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe, the samples were analyzed by thermogravimetric analysis (Fig. 4f). In an inert gas atmosphere, pure ANFs decompose most rapidly at 556 °C, resulting in the decomposition of aramid nanofibers into carbon materials CNFs. The corresponding differential thermogravimetric (DTG) curves of samples is shown in Fig. S7. The decomposition curves of α-Fe₂O₃@ANFs and α-Fe₂O₃@ANFs/Al-α-Fe₂O₃ exhibit peaks around 410 and 550 °C, where the former is attributed to the decomposition of PVP in the system, while the latter is due to ANFs decomposition. The decomposition of aramid fibers into carbon materials CNFs, facilitates the reduction of α-Fe₂O₃ to Fe₃O₄ and Fe. The heat resistance of nanocomposites increases with Al-α-Fe₂O₃ introduction, which indicates α-Fe₂O₃@ANFs/Al-α-Fe₂O₃ nanocomposites have a higher residual weight ratio at 800 °C than pure ANFs. The introduction of inorganic materials can further improve the heat resistance of polymers [31,32]. It can better elucidate this process of the nanocomposites combining the XRD in Fig. 2h and the VSM in Fig. 4a. In this process, one component in the samples acts as a reducing agent, enabling the reduction in other components, hence termed as the self-reduction process of nanocomposites.

The relative complex permittivity (ε_r =  εʹ −  jεʺ), relative complex permeability (μ_r =  μʹ −  jμʺ), dielectric loss and magnetic loss of CNFs, Fe₃O₄-Fe@CNFs nanoparticles, Al-Fe₃O₄ nanosheets, Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe, and Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe_(1:2) are depicted in Fig. 5. The real parts of dielectric constant (εʹ) and permeability (μʹ) reflect the electrical and magnetic energy storage capacity of the materials, while the imaginary parts (εʺ and μʺ) indicate energy loss [33]. A higher tan δ_ε value implies better electrical energy absorption capability [34,35]. CNFs exhibit significantly higher dielectric constant and tan δ_ε than other materials, due to the excellent conductivity of CNFs, allowing electrons to freely propagate within them. However, CNFs lack magnetic loss, limiting their EMW absorption performance. Magnetic loss is theoretically divided into three mechanisms (exchange resonance, natural resonance, and eddy current loss) [9]. Introducing magnetic Fe₃O₄-Fe nanoparticles into CNFs can significantly reduce their dielectric real and imaginary parts, but they remain at a relatively high level, which is detrimental to EMW absorption. By adjusting the proportions of magnetic and conductive components, the composite material’s electromagnetic parameters are fine-tuned to achieve impedance matching. Further introducing low-dielectric Al-Fe₃O₄-Fe nanosheets into the nanocomposites can additionally reduce the complex permittivity of the composites, and adjusting the Al-Fe₃O₄-Fe nanosheets content maintains appropriate dielectric properties. Figure 5a illustrates that as the Al-Fe₃O₄-Fe nanosheets content increases, the real part of the dielectric constant of the nanocomposites decreases, thus adjusting the EMW absorption performance. The wide scattering area of 2D Al-Fe₃O₄-Fe nanosheets enhances EMW attenuation. The dielectric real parts of CNFs, Fe₃O₄-Fe@CNFs nanoparticles, Al-Fe₃O₄ nanosheets, Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe, and Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe_(1:2) exhibit a gradual decrease at high frequencies, which is associated with the dispersion effects present in the samples [21]. In Fig. 5d, μʹ values decrease with frequency, which may be due to relaxation of the magnetic moment process rather than hysteresis. From Fig. 4d, e, it is evident that the magnetic permeability parameters of the samples fluctuate, which is related to the occurrence of natural resonance within the samples. In Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites, 0D Fe₃O₄-Fe nanoparticles, 1D CNFs, and 2D Al-Fe₃O₄-Fe nanosheets optimize impedance matching for excellent EMW absorption.

Based on the transmission line theory [36,37], the reflection loss can be calculated by the following formulas:

where Z_in represents the normalized input impedance, Z₀ is the impedance of free space, ε_r and μ_r are the relative complex permittivity and permeability, respectively, f is the frequency of the microwave, d is the thickness of the absorber, and c is the velocity of light. In evaluating EMW absorbers, key criteria include minimum reflection loss (RL_min) and effective absorption bandwidth (EAB) [38,39,40]. In practical applications, the EAB is the frequency range where the RL value is less than − 10 dB, which signifies the absorption of 90% of electromagnetic energy [41]. The EMW absorption properties of the samples are shown in Fig. 6. Their minimum reflection losses, effective absorption bandwidths, and corresponding thicknesses are listed in Table S1. Notably, Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe has RL_min of − 59.3 dB (6.6 GHz) at 4.3 mm and an EAB of 5.6 GHz (11.8-17.4 GHz) at 2.2 mm (Fig. 6d). Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe_(1:2) shows RL_min of − 49.5 dB (14.9 GHz) at 2.1 mm and an EAB of 6.4 GHz (11.3-17.7 GHz) at 2.2 mm (Fig. 6e). Additionally, the 3D reflection loss curves maps of Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe_(1:1) and Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe_(1:2.5) are presented in Fig. S8. Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe_(1:2) offers thickness and EAB advantages compared to Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe, as depicted in the 2D contour map (Fig. 6f), while Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe has superior RL_min, as shown in the 2D reflection loss curve (Fig. 6g). With the increase in Al-Fe₃O₄-Fe nanosheets content in Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites, the minimum reflection loss decreases, corresponding to an increase in thickness, and the effective absorption bandwidth initially increases and then decreases (Table S1). Therefore, the content of Al-Fe₃O₄-Fe in the composite materials plays a role in regulating the performance of the electromagnetic parameters, thus regulating the EMW absorption performance of the materials. Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe emerges as a more competitive option, offering significant advantages in terms of both EAB and RL_min values when compared to other samples, as depicted in Fig. 6h-i. These results underscore the importance of carbon and magnetic components in electromagnetic wave (EMW) absorption. Table 1 offers a comparison of the EMW absorption capabilities of Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe with other materials mentioned in published studies. The RL_min of Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe at − 59.3 dB is stronger than other absorbing materials. Additionally, the EAB of Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe_(1:2) at 6.4 GHz is wider than other materials, and its thickness is also smaller. This is primarily attributed to the hierarchical structure of the composite material and the synergistic interaction among its components. The results demonstrate that Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe has certain advantages in terms of RL_min, effective absorption bandwidth, and thickness.

The radar cross-section (RCS) simulation is further carried out to evaluate the actual far-field EMW absorption performance. In general, the RCS value (σ) is determined by theta and phi in spherical coordinates, as shown below [46,47]:

where S, λ, E_s and E_i are the area of the simulated plate, the length of the incident EMW, the scattered field intensity of transmitting waves, and the incident field intensity of receiving waves. In general, once the shape and geometric configuration of a device are determined, RCS data can assess the device’s EMW absorption and reflection capabilities. Due to the strong electromagnetic scattering affecting metal components of military targets, traditional equipment is easily detected by radar [48]. Coating military equipment surfaces with microwave-absorbing layers reduces RCS [49]. To evaluate the practical application potential of Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe, the RCS of a 180 × 180 × 0.5 mm³ perfect electrical conductor (PEC) plate substrate coated with a 2.5 mm thick Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe layer was simulated using CST STUDIO SUITE 2023 software. Here, positive x-axis is phi 0°, positive z-axis is theta 0°, the plate is placed on the x-o-y plane in the direction of positive z-axis, and the EMW enters the model along the z-axis. The 3D results (Fig. 7a, b) maintain incident angles of EMW phi = 0° and theta = 0° at 12.5 GHz, receiving results at all angles. The 2D result (Fig. 7c) maintains incident angles of EMW phi = 0° and theta = 0° at 12.5 GHz, receiving angle phi = 0° and adjusting theta from -90° to 90° for receiving. The 3D radar wave scattering signals from the original PEC plate and the Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe-coated plate at 12.5 GHz were examined. The scattered signal from the Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe-coated PEC plate is significantly lower than that of the PEC substrate (Fig. 7a-b), indicating that Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe has an effective absorption capacity. The 2D curve of simulated RCS values in the x-o-z plane (Fig. 7c) shows a drop in RCS values from 13.4 dB m² (PEC) to 7.2 dB m² (PEC coated with Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe) when both theta and phi are 0°. RCS simulation results demonstrate Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe’s ability to effectively suppress EMW scattering and reflection from the PEC substrate surface. For a better practical application characterization, Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe is coated on the surface of an airplane with a thickness of 3 mm (Fig. 7d), and EMW absorption effects are simulated. Here, positive x-axis is phi 0°, positive z-axis is theta 0°, and the airplane is placed on the x-o-y plane and the nose of the plane is facing the direction of positive x-axis. The direction of the incident EMW is a function of theta and phi. The forward view RCS curve of the airplane (receiving angles theta = 90°, phi = 0 ~ 360°) and the overview RCS curve (receiving angles theta = -180° ~ 180°, phi = 90°) at 10 GHz were tested for horizontal polarization (HP) and vertical polarization (VP). Compared to the PEC-based model, the airplane coated with Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe absorbers reflects much weaker signals (Fig. 7e-h). These simulation results highlight Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe composites promising applications in both civil and military stealth coatings.

The possible EMW absorption mechanism of Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe is illustrated in Fig. 8. The exceptional EMW absorption performance primarily stems from the synergy of dielectric loss, conduction loss, and magnetic loss [50,51]. In Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites, 1D CNFs play a crucial role as framework materials in forming the conductive network of the nanocomposites. On the other hand, CNFs with abundant defects, oxygen-containing functional groups, and dangling bonds provide active sites for dipole polarization loss [52]. For the 2D Al-Fe₃O₄-Fe nanosheets, electromagnetic waves are repeatedly scattered by the nanosheets upon entering the absorber, thereby increasing multiple scattering losses [53]. In addition, the content of Al-Fe₃O₄-Fe nanosheets can effectively control the conductivity of hierarchical nanocomposites and optimize the electromagnetic parameters of nanocomposite materials. As for the 0D Fe₃O₄-Fe nanoparticles, the heterogeneous interface between the nanoparticles and CNFs is conducive to interface polarization, thus improving dielectric loss. Additionally, the introduction of Fe₃O₄-Fe nanoparticles and Al-Fe₃O₄-Fe nanosheets can also provide magnetic losses, thereby improving impedance matching [54]. In addition, the Fe₃O₄-Fe nanoparticles and Al-Fe₃O₄-Fe nanosheets are mixed with Fe. Because Fe exhibits strong saturation magnetization, and its presence further enhances the saturation magnetization of Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites (Fig. 4a). Fe is present within Fe₃O₄, forming heterogeneous interfaces, resulting in interface polarization. Components at heterogeneous interfaces exhibit differences in conductivity, thus under the action of an electromagnetic field, free charges can be trapped at the interface, leading to the accumulation of positive and negative charges in the interface region [55]. In addition, the hierarchical structure, artificially designed, integrating 0D/1D/2D materials was developed. The prepared hierarchical Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Fe nanocomposites exhibit excellent EMW absorption performance, closely related to multi-dimensional gradient structure, multiple loss mechanisms, and synergistic effects of each component. Moreover, Fe₃O₄-Fe@CNFs/Al-Fe₃O₄-Feʹs 3D hierarchical heterogeneous structure provides multiple interfaces for interfacial polarization and ample voids for scattering microwaves. These factors enable the composites to achieve high-efficiency EMW absorption with a low filling load and thin thickness.