Preparation of the carbon nanotubes by catalytic pyrolysis of waste plastic and application for electromagnetic shielding

2.1. Experimental materials

LDPE, MgO (analytical grade), Fe (NO₃)₂⋅9H₂O (analytical grade), and Ni(NO)₂⋅6H₂O (analytical grade) were purchased from Sinopharm Reagent Co., LTD.

2.2. Preparation of Fe/MgO and Ni/MgO catalysts

The Fe/MgO and Ni/MgO catalysts were synthesized via a conventional wet impregnation method. Initially, 7.21 g of Fe(NO₃)₃·9H₂O or 4.91 g of Ni(NO₃)₂·6H₂O was dissolved in 50 mL of ethanol (95%), followed by ultrasonic dispersion for 30 min to form a homogeneous solution (Figure 1a). Subsequently, 10 g of the MgO support was introduced into the respective metal salt solution. The resulting suspension was continuously sonicated for an additional 1 h to ensure the uniform distribution of metal ions on the support. After complete dispersion, the mixture was dried and finely ground into a homogeneous powder. Finally, the precursor powder was placed in a tube furnace and heated to 800°C at a heating rate of 5°C/min. The sample was then calcined under an air atmosphere for 2 h and naturally cooled to room temperature, yielding Fe/MgO or Ni/MgO catalyst.

Close

Figure 1. (a) The preparation process of Fe/MgO and Ni/MgO catalysts and (b) CNTs synthesis via the catalytic pyrolysis of LDPE.

Export to PPT

2.3 Preparation of CNTs

Figure 1(b) illustrates the schematic flowchart of CNTs synthesis via the catalytic pyrolysis of LDPE. The catalytic pyrolysis experiments were conducted in a single-zone tube furnace using a corundum boat as the reaction vessel. Specifically, a mixture of LDPE and the prepared catalyst (with varying catalyst loadings of 0.5 wt%, 1.0 wt%, and 1.5 wt%) was loaded into the corundum boat and placed inside the tube furnace. Prior to heating, nitrogen gas was purged through the system at a flow rate of 40 mL/min for 30 min to establish a strictly inert atmosphere. Subsequently, the system was heated to the target pyrolysis temperature (ranging from 700°C to 900°C) at a heating rate of 5°C/min and maintained at this temperature for 2 h with N₂ flow rate of 40 mL/min. Upon completion of the reaction, the tube furnace was naturally cooled to room temperature under N₂ atmosphere. The resulting solid products collected from the furnace were designated as X-CNTs-Y, where X represents the pyrolysis temperature and Y denotes the mass fraction of the catalyst. The characterizations of the sample are put in the supplementary material.

2.4. Characterization

The crystal structures of the Fe/MgO and Ni/MgO catalysts were characterized by X-ray diffraction (XRD) within a 2θ scanning range of 10° to 90°. The pore structures and specific surface areas of the catalysts were analyzed via Brunauer-Emmett-Teller (BET) measurements, and the strong metal-support interactions were investigated using X-ray photoelectron spectroscopy (XPS). The composition of the pyrolysis gas generated from LDPE was analyzed utilizing a gas chromatography-mass spectrometry (GC-MS) system. The surface morphologies and microstructures of the prepared catalysts and synthesized carbon nanotubes (CNTs) were observed via scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Thermogravimetric analysis (TGA) was employed to quantitatively evaluate the purity and thermo-oxidative behavior of the CNTs. Furthermore, the degree of graphitization, structural defects, and purity of the carbon products were assessed using Raman spectroscopy and an organic elemental analyzer. Finally, the electromagnetic interference (EMI) shielding performance of the CNT composites was evaluated using a vector network analyzer (VNA).

3.1. Characterization of the Fe/MgO and Ni/MgO catalysts

The (XRD) patterns of the as-prepared Fe/MgO and Ni/MgO catalysts are shown in Figure 2(a). The strong diffraction peaks observed at 2θ = 36.93°, 42.83°, 62.30°, 74.63°, and 78.63° are assigned to the (111), (200), (220), (311), and (222) planes of MgO (JCPDS No. 45-0946), respectively, confirming the high crystallinity of the support. The diffraction peak detected at 2θ = 35.36° is attributed to the (311) plane of magnesioferrite (MgFe₂O₄), indicating a solid-state reaction between the Fe species and the MgO matrix during preparation process and forming a stable spinel structure. Notably, no diffraction peaks of Ni-related oxides are detected, likely due to their overlap with the MgO peaks. As illustrated in Figure 2(b), both catalysts exhibit typical type-IV isotherms with distinct hysteresis loops, indicating well-developed mesoporous structure. The specific surface area of the Fe/MgO catalyst is 45.62 m₂/g, which is slightly higher than that of the Ni/MgO catalyst (42.47 m₂/g). Furthermore, the corresponding pore size distribution curves (Figure 2b inset) demonstrate that the mesopores of both catalysts are primarily concentrated in the 5-15 nm range. These mesoporous features facilitate the high dispersion of active metal sites for the reactions.

Close

Figure 2. (a) XRD patterns, (b) N₂ adsorption–desorption isotherms and pore size distributions of the Fe/MgO and Ni/MgO catalysts, yields of (c) Fe/MgO-CNTs and (d) Ni/MgO-CNTs synthesized at various temperatures and catalyst loadings.

Export to PPT

The chemical states of the catalysts are analyzed by XPS, as shown in Figure S1(a) and Figure S1(b). For the Fe/MgO catalyst, the Fe 2p spectrum reveals the coexistence of Fe₃⁺ (66.98%) and Fe₂⁺ (35.02%). Furthermore, the main Fe 2p_3/2 peak is located at 712.3 eV, exhibiting a positive binding energy shift of 1.4 eV compared to standard bulk Fe₂O₃ (710.9 eV). This result indicates a strong metal-support interaction (SMSI) between the Fe₃⁺ species and the MgO support [11] is formed. Conversely, the Ni 2p spectrum of the Ni/MgO catalyst shows that Ni predominantly exists as Ni₂⁺ (81.51%), accompanied by a minor proportion of Ni₃⁺ (18.49%). Additionally, the Ni 2p_3/2 peak is centered at 855.3 eV, displaying a positive shift of 1.3 eV relative to standard bulk NiO (854.0 eV). This substantial shift provides robust evidence for SMSI and the formation of a stable NiO-MgO solid solution [12].

3.2. Effect of catalytic pyrolysis temperature and catalyst content on CNTs yield

Figures 2(c and d) illustrate the CNTs yields using the Fe/MgO and Ni/MgO catalysts with varying mass fractions at 700-900°C. The yield initially increases and then decreases, peaking at 800°C, which aligns with the findings of [13]. At 800°C, a maximum CNTs yield of 21.04% is obtained with a 1.0 wt% Fe/MgO loading (Figure 2c). The reduced yield at higher loadings (> 1.0 wt%) is likely attributed to reduced catalytic efficiency resulting from particle agglomeration [14,15]. Similarly, Figure 2(d) shows a similar trend for the Ni/MgO catalyst. However, its CNTs yield remains consistently lower than that of the Fe/MgO catalyst. Increasing the Ni/MgO catalyst loading from 1.0 to 1.5 wt% resulted in negligible improvement in CNTs yield, suggesting that Ni nanoparticles are more prone to agglomeration than Fe nanoparticles. The SEM analysis will further confirm this hypothesis.

3.3. Effect of catalytic pyrolysis temperature and catalyst content on the graphitization degree of CNTs

Figure 3(a) illustrates the thermogravimetric analysis (TG) curves of the synthesized CNTs using the Fe/MgO and Ni/MgO catalysts (0.5-1.5 wt%) at 800°C. The weight loss occurring before 550°C is attributed to the oxidation of amorphous carbon. While, the weight loss after 550°C is ascribed to the oxidation of the CNTs. The primary oxidation temperatures for all samples are well above 550°C, confirming a high degree of graphitization. Notably, both systems exhibit their maximum weight loss at a catalyst loading of 1.0 wt%, reaching approximately 59.84% for Fe/MgO and 38.45% for Ni/MgO. The derivative thermogravimetry (DTG) curves (Figure 3b) display oxidation peaks predominantly above 550°C, indicating the high purity of the products.

Close

Figure 3. (a) TGA and (b) DTG curves of CNTs synthesized with different Fe/MgO and Ni/MgO catalyst loadings, (c) Raman spectra of CNTs prepared with various Fe/MgO and (d) Raman spectra of CNTs prepared with various Ni/MgO catalyst loadings.

Export to PPT

Based on DTG calculations, the graphitic carbon contents of the CNTs grown on the Fe/MgO catalysts with loadings of 0.5, 1.0, and 1.5 wt% are 85.93%, 91.55%, and 84.69%, respectively. Meanwhile, at the same loading levels, the graphitic carbon contents on the Ni/MgO catalysts are 36.45%, 89.58%, and 96.18%, respectively. As the metal loading increases from 0.5 wt% to 1.0 wt%, the oxidation peaks for both catalysts shift toward high temperatures, demonstrating a reduction in structural defects. However, at a loading of 1.5 wt%, the Fe/MgO peak shifts toward low temperatures, which is likely due to structural defects caused by the agglomeration of Fe nanoparticles [16]. In contrast, the Ni/MgO peak continues to shift toward high temperatures at 1.5 wt%, suggesting that a high Ni concentration favors the formation of highly graphitized CNTs.

Raman spectroscopy was employed to further evaluate the degree of graphitization. As illustrated in Figure 3(c), the characteristic D-band (1350 cm⁻₁) and G-band (1580 cm⁻₁) correspond to disordered carbon structures and the tangential vibrations of ordered sp₂ carbon atoms, respectively. A lower intensity ratio of these peaks (I_D/I_G) signifies a higher degree of graphitization [17]. For the Fe/MgO system (Figure 3c), the I_D/I_G ratio reached a minimum of 0.857 at 1.0 wt% loading, indicating the optimal degree of graphitization at this concentration. At a lower catalyst content of 0.5 wt%, the I_D/I_G ratio peaked at 0.975. This is attributed to the scarcity of Fe active sites; when LDPE pyrolysis generates a large volume of gaseous hydrocarbon fragments (e.g., ethylene, methane), a portion of these fragments cannot attach to the active sites for adsorption and catalytic growth. Instead, these free carbon fragments undergo free-radical polymerization in the gas phase, forming amorphous carbon that encapsulates the catalyst. This leads to increased structural defects and a higher I_D/I_G ratio.

Increasing the catalyst loading to 1.0 wt% resulted in the minimum I_D/I_G ratio (0.857), suggesting a reduction in crystalline defects. This indicates that Fe particles were uniformly dispersed on the MgO support, forming a high density of effective active sites that facilitated the ordered growth of carbon atoms into carbon nanotubes (CNTs). However, further increasing the loading to 1.5 wt% caused the I_D/I_G ratio to rise to 0.904, indicating partial agglomeration of the catalyst at high temperatures. The resulting larger Fe metallic particles were less effective in catalyzing CNT synthesis, leading to carbon encapsulation on the catalyst surface and a subsequent decline in the quality of the produced CNTs. In contrast, for the Ni/MgO system (Figure 3d), the I_D/I_G ratio decreased monotonically with increasing loading (1.223 > 1.124 > 1.012). This trend is consistent with the DTG results, suggesting that the Ni/MgO catalyst possesses superior anti-sintering stability compared to the Fe/MgO catalyst.

As shown in Table S1, the elemental analysis reveals the carbon contents of 93.52% and 94.47% for the acid-washed CNTs prepared using the Fe/MgO and Ni/MgO catalyst, respectively. Furthermore, the hydrogen contents (0.59 wt% and 0.68 wt%) confirm that the LDPE feedstock underwent catalytic deep cracking and extensive dehydrogenation. These collective results demonstrate that the prepared CNTs possess an excellent degree of graphitization and high purity, rendering them highly suitable for subsequent EMI shielding applications. Comparison of synthesis strategies and EMI shielding performance is shown in Table S2.

3.4. Effect of catalyst content on CNTs growth

To establish the correlation between the CNTs microstructure and catalyst loading, the morphology of the CNTs synthesized at 800°C is analyzed using SEM. Figure 4(a) reveals sparse filamentous CNTs on the 0.5 wt% Fe/MgO catalyst, accompanied by numerous underdeveloped CNTs precursors. This result indicates incomplete carbon source utilization. The carbon supply rate exceeds the conversion capacity, leading to rapid carbon encapsulation and subsequent catalyst deactivation [18]. Upon increasing the Fe/MgO loading to 1.0 wt% (Figure 4b), the dense and uniform elongated CNTs are formed on the catalyst surface, indicating improved carbon utilization efficiency and suppressed catalyst encapsulation. However, as the Fe/MgO loading further increases to 1.5 wt% (Figure 4c), the localized aggregation and underdeveloped CNTs precursors reappear. Correlating this with the yield reduction observed in Figure 2(c), it is evident that the concentration of Fe nanoparticles exceeds the dispersion limit of the MgO support. Consequently, the high-temperature sintering induces metal agglomeration, thereby inhibiting catalytic activity. Thus, for the Fe/MgO system, carbon encapsulation dictates catalyst deactivation at low loadings, whereas sintering-induced agglomeration dominates at high loadings.

Close

Figure 4. SEM images of CNTs synthesized at 800°C with various catalyst loadings: (a-c) Fe/MgO and (d-f) Ni/MgO at (a, d) 0.5 wt%, (b, e) 1.0 wt%, and (c, f) 1.5 wt%.

Export to PPT

In contrast, the Ni/MgO system exhibits a distinctly different morphological evolution. At 0.5 wt% Ni/MgO (Figure 4d), short filamentous CNTs are distributed across the surface. When the loading content is increased to 1.0 wt% (Figure 4e), the length of the CNTs increases. At a Ni/MgO content of 1.5 wt% (Figure 4f), severe agglomeration of CNTs occurs, despite no significant changes in CNT morphology. This result indicates that Ni possesses a high catalytic cracking efficiency for the carbon source. The graphitic carbon generated during the reaction rapidly encapsulates the Ni particles, causing the reaction to terminate [19]. Ultimately, the inherent properties of the metallic nanoparticles dictate their distinct growth mechanisms [20]. Specifically, Ni/MgO generates short, thin and thread-like CNTs with small diameter. While, the Fe/MgO produces robust and large-diameter CNTs with a network-like configuration.

3.5. Growth mechanism analysis of CNTs

3.5.3. Analysis of CNTs growth mode

To further elucidate the internal morphology and growth mechanisms of the synthesized CNTs, the TEM characterization is conducted on the samples prepared at 800°C. As shown in Figure 5(a), the 1.0 wt% Fe/MgO-CNTs exhibit a distinct hollow structure. The prominent encapsulation of variously shaped Fe nanoparticles at the tube tips unambiguously indicates a tip-growth mechanism Figure 5(b) [21]. Statistical analysis of 100 individual Fe-based CNTs (Figures S3a and S3b) reveals an average length of 8.31 μm and an outer diameter of 16.76 nm.,which is consistent with the Figure 5(c) analysis. Figure 5(d) demonstrates that the active growth tip contains a short, rod-like Fe nanoparticle (23.91 × 10.26 nm). The highly uniform graphitic interlayer spacing of d = 0.3304 nm is close to the ideal value (d = 0.3354 nm), confirming excellent crystallinity. Notably, the Fe/MgO-derived CNTs exhibit a unique bamboo-like stacked morphology. Although previous studies have attributed this bamboo-like structure to fluctuations in carbon precursor concentration [22]. In this system, the bamboo-like stacked structure is driven by the rapid thermal pyrolysis of the LDPE polymer (Figures S2c and S2d), which generates a highly dynamic and complex gaseous environment during the CNTs growth process.

Figure S3

Download

Close

Figure 5. TEM images of CNTs synthesized at 800°C with 1.0 wt% catalyst loading. (a-d) CNTs grown over the Fe/MgO catalyst and (e-h) CNTs grown over the Ni/MgO catalyst.

Export to PPT

Conversely, the CNTs derived from the Ni/MgO catalyst exhibit distinct structural differences. As shown in Figure 5(e), the Ni/MgO-CNTs display a helically stacked morphology Figure 5(f) shows that the CNTs are distributed relatively uniformly. The statistical analysis (Figures S3c and S3d) indicates that the Ni-based CNTs possess an average length of 3.74 μm and an outer diameter of 11.72 nm, which is consistent with the Figure 5(g) analysis. As shown in Figure 5(h), the graphitic interlayer spacing of d = 0.3304 nm is similarly close to the ideal value (0.3354 nm), further confirming excellent crystallinity. Furthermore, the metal nanoparticles are predominantly aggregated and anchored at the catalyst-substrate interface, suggesting a base-growth mechanism.

To further evaluate the activity of the catalytic metal particles during the catalytic pyrolysis system, a statistical analysis of 100 individual Fe- and Ni-based CNTs (Figures S3e and S3f) is conducted. The results reveal that the average sizes of the metallic Fe and Ni nanoparticles are 14.98 nm and 12.39 nm, respectively, which closely match the outer diameters of the as-prepared CNTs. This confirms that the Fe and Ni nanoparticles effectively serve as active templates to facilitate the diffusion and precipitation of carbon atoms during the catalytic pyrolysis process.

3.5.4. CNTs growth mechanism

According to established metal-catalyzed growth theories, the Fe₃C serves as a critical intermediate for CNTs growth [23]. The proposed CNTs growth mechanism, supported by previous literature, is illustrated in Figure 6(a). The Fe/MgO catalyst is utilized as a representative model to analyze this process. Initially, the reducing gases (e.g., H₂) generated from the pyrolysis of LDPE reduce the Fe₃⁺ specie to metallic Fe. Subsequently, carbon source gases (e.g., C₂H₄) diffuse into the metallic Fe nanoparticles, leading to the formation of metastable Fe₃C. Due to its thermodynamic instability, the Fe₃C decomposes, precipitating carbon in the form of graphitic layers. The continuous formation and decomposition of Fe₃C drive the progressive precipitation of multilayered graphitic sheets, thereby promoting the continuous growth of the CNTs. Furthermore, as shown in Figure 6(b), the resulting CNTs can be classified into distinct morphological categories, such as straight tubular and bamboo-like structures, which is consistent with the aforementioned microstructural analysis.

Close

Figure 6. (a) Growth mechanism of CNTs and (b) three growth types of CNTs.

Export to PPT

3.6. Analysis of electromagnetic shielding properties of CNTs

As shown in Figure S4, the total shielding effectiveness (SET) consists of the absorption shielding effectiveness (SEA) and the reflection shielding effectiveness (SER). The mechanism of SER involves the induction of surface currents when incident electromagnetic waves interact with conductive materials. These currents generate opposing electromagnetic waves that reflect the incident waves at the interface. Conversely, the mechanism of SEA originates from the material’s ability to convert incident electromagnetic energy into thermal energy through ohmic loss or polarization relaxation, a process known as energy dissipation. The three electromagnetic shielding effectiveness parameters can be precisely calculated using Equations 1-3. In the equations, P₀, P₁, and P_R represent the incident power, transmitted power, and reflected power, respectively.

3.6.1. Electromagnetic shielding properties of CNTs prepared from different catalysts

The EMI of the synthesized CNTs is evaluated using the coaxial line method by blending the CNTs with a paraffin matrix. Since paraffin is a non-polar, wave-transparent medium with a negligible complex permittivity in the microwave frequency range (2-18 GHz), the measured EMI shielding performance originates entirely from the conductive networks and interfacial polarization inherent to the embedded CNTs. Figure 7 presents the SET profiles of the CNTs synthesized at 800°C. As illustrated in Figure 7(a), at a filler loading of 50 wt%, the Fe/MgO-derived CNTs exhibit an average SET of 21.01 dB in the X-band (8-12 GHz), reaching a maximum peak of 26.3 dB. This significantly surpasses the commercial standard of 20 dB, demonstrating exceptional EMI shielding capabilities. As shown in Figure 7 (b-c), the SEA values of all samples increase with the increase in frequency and the SEA values are higher than the corresponding SER values (>8GHz).

Close

Figure 7. (a) SET, (b) SEA, and (c) SER curves of Fe/MgO-CNTs; (d) SET, (e) SEA, and (f) SER curves of Ni/MgO-CNTs at various filler loadings (20–50 wt%).

Export to PPT

Conversely, across the loading range of 20 wt% to 50 wt%, the highest average SET for the Ni/MgO-derived CNTs reaches only 14.73 dB, falling short of commercial requirements (Figure 7d). This inferior shielding performance is primarily attributed to their shorter, filamentous morphology, which hinders the construction of an efficient, interconnected conductive network [24]. CNTs attenuate electromagnetic waves primarily through absorption, as quantified by the shielding effectiveness due to absorption (SEA), which is highly dependent on their superior electrical conductivity and complex internal network architecture [25]. As shown in Figure 7 (e), the SEA values of all samples increase with the increase in frequency. The Figure 7(f) indicates that the SER values of the 800°C -CNTs50wt%-Fe/MgO increase at <6GHz and then decrease at>6GHz.

The AC conductivity (σ_AC) serves as a direct indicator of this network formation [26]. As shown in Figure S5(a) and (b), the σ_AC of both composites exhibits a strong positive correlation with the CNTs loading. Specifically, the Fe/MgO-CNTs system displays the most drastic performance enhancement when the filler content increases from 20 wt% to 35 wt%. The peak σ_AC surges from 11.14 S/m to 17.97 S/m (a 61.31% increase). Simultaneously, the SEA improves by 48.83%, rising from 8.54 dB to 12.71 dB. This remarkable leap in performance indicates a critical structural transition of the CNTs from isolated clusters into a continuous 3D percolation network [27,28]. The elevated filler loading significantly diminishes the inter-tube distance, thereby facilitating stable and robust electron transport via direct physical contact and quantum tunneling effects [29]. Consequently, the electromagnetic wave dissipation capacity of the macroscopic composite is substantially augmented.

Figure S5

Download

3.6.2. Effect of CNTs synthesis temperature on electromagnetic shielding performance

The pyrolysis temperature serves as a critical thermodynamic and kinetic determinant for the micromorphology and subsequent EMI shielding performance of the Fe/MgO-CNTs composites. As illustrated in the SEM images (Figures 8a-c), the CNTs synthesized at 700°C are relatively short and sparse. This inferior growth is attributed to the insufficient activation energy at lower temperatures, which limits the catalytic decomposition rate of hydrocarbon fragments and the subsequent diffusion of carbon atoms through the catalyst bulk [30]. Consequently, the resulting discrete CNTs fail to reach the percolation threshold, leading to an inefficient conductive pathway.

Close

Figure 8. EMI shielding curves of Fe/MgO-CNTs synthesized at (a) 700°C, (b) 800°C, and (c) 900°C; (d) SET, (e) SEA, and (f) SER curves of Fe/MgO-CNTs at a 50 wt% filler loading.

Export to PPT

In contrast, 800°C is identified as the optimal growth temperature, balancing the rates of carbon precursor decomposition and crystalline growth. This equilibrium yields thin, exceptionally long, and highly entangled CNTs with a high aspect ratio, which is essential for constructing a robust and long-range continuous 3D conductive network within the matrix. In contrast, 800°C is identified as the optimal growth temperature, balancing the rates of carbon precursor decomposition and crystalline growth. This equilibrium yields thin, exceptionally long, and highly entangled CNTs with a high aspect ratio, which is essential for constructing a robust and long-range continuous 3D conductive network within the matrix [31]. This morphological evolution directly dictates the macroscopic electrical properties, as a well-integrated network facilitates efficient electron hopping and tunneling.

Conversely, an excessively high temperature of 900°C leads to a decline in CNT quality. This deactivation mechanism involves the potential local sintering of Fe nanoparticles on the MgO support and severe secondary thermal cracking of hydrocarbons, which induces amorphous carbon deposition and disrupts network continuity. The AC conductivity measurements (Figure S6) corroborate this interpretation, showing that the 800°C sample exhibits the highest conductivity across the measured spectrum. Benefiting from this superior electron transport capability and the enhanced ohmic losses within the dense 3D network, the 800°C sample achieves the maximum SET (Figure 8(d)) [32]. This confirms that the synergistic effect of morphological integrity and network connectivity is the underlying driver for optimizing EMI shielding performance. As shown in Figure 8 (e-f), the SEA values of all samples increase with the increase in frequency and the SEA values are higher than the corresponding SER values (>6GHz). The SEA values are much larger than the SER values strongly proves that the CNTs prepared under the condition of 1.0 wt% - Fe/MgO belong to a typical absorption-dominated electromagnetic shielding material.

Figure S6

Download