3D printed porous biomass–derived SiCnw/SiC composite for structure–function integrated electromagnetic absorption

ABSTRACT

Ceramic-based absorbing composite is irreplaceable in high-temperature conditions. This work has fabricated a novel biomass–derived porous SiCnw/SiC composite for structure–function integrated electromagnetic wave (EMW) absorption through selective laser sintering (SLS) 3D printing and carbothermal reduction. SLS processed the biomass-derived wood precursor with unique porous microstructures. The structure–function properties were controlled by changing the SiO carbothermal reduction temperatures, which facilitated the growth of SiC nanowires for effective EMW absorption. The 3D printed porous SiCnw/SiC composite shows efficient EMW absorption abilities with a minimum reflection loss of −49.01 dB and an effective absorption bandwidth of 5.1 GHz. The bulk density and flexural strength of porous SiCnw/SiC composite are respectively 0.73 ± 0.001 g/cm³ and 6.21 ± 0.66 MPa. Despite a high open porosity of 75.58 ± 0.31%, the porous SiCnw/SiC composite demonstrates excellent thermal conductivities of 3.21∼4.99 W/(m·K) and superior fire-resistant ability. The 3D printed SiCnw/SiC composite integrates structure and functions, indicating wide applications in specific harsh environments.

KEYWORDS:

• Selective laser sintering
• 3D printing
• biomass-derived
• porous SiCnw/SiC composite
• electromagnetic wave absorption

1. Introduction

EMW absorption materials play a crucial part in civil and military fields owing to their vital function for ensuring human health and preventing information leakage (Yin et al. 2014; Liang et al. 2014; Wang, Cheng, and Zhang 2018). For better EMW absorbing ability, many studies have focused on controlling the impedance matching condition and attenuation characteristic to prepare efficient absorbing materials (Zhou et al. 2019; Cai et al. 2021; Javid et al. 2021; Li et al. 2021). Meanwhile, high-temperature structural absorbing materials, such as SiC matrix composites are irreplaceable in many harsh environments due to their high specific strength, temperature resistance, and chemical stability (Wang et al. 2018; Dong et al. 2019).

There are many traditional ways to apply SiC in EMW absorbing fields. Bio-inspired porous absorbers preparing methods have proven to be easily operated. Abundant microscopic pores facilitate the realisation of multiple reflections for EMW waves (Mei et al. 2021; Fan et al. 2021; Luo et al. 2019), beneficial to electromagnetic energy consumption. Liang et al. (Liang and Wang 2019) used the carbonised eggplants as templates to prepare lightweight and strong SiC aerogels by carbothermal reduction, obtaining the developed SiC aerogels with a RL_min of −43 dB. Cai et al. (Cai et al. 2021) prepared alternated ‘soft-and-stiff’ multilayered Si₃N₄/SiC aerogels, resulting in a ‘reflection−absorption−zigzag’ reflection absorbing pattern. The obtained N/C aerogel showed a RL_min of −45 dB. Another efficient way is the incorporation of one-dimensional SiC materials, including SiC fibre (SiC_f), SiC whisker (SiC_w), or SiC nanowire (SiCnw). SiO vapour carbothermal reduction is a desirable way to in-situ growing SiCnw (Zhu et al. 2016). Previous studies have demonstrated that SiCnw could serve as an efficient EMW absorbing agent attributed to its lattice defects, interfacial polarisation, and dielectric loss. Lan et al. (Lan and Wang 2020; Lan et al. 2020; Lan and Wang 2020) prepared binary porous SiC with pressed SiC_w as the skeleton and introduced the secondary SiCnw via carbothermal reduction, and they obtained a minimum reflection loss (RL_min) of −51 dB. Ren et al. (Ren et al. 2020) reported a novel carbon nanowires (CNWs)/SiCnw hybrid networks reinforced silicon carbonitride (SiCN) composite, realising a RL_min of −21.6 dB at 2.35 mm thickness.

However, micro-morphology regulation has many limitations in improving EMW absorbing performance, especially for broadband absorbing. Sun et al. (Sun et al. 2022) developed a macro-metamaterials design that can achieve super broadband absorbing of 2.4∼40 GHz. Nevertheless, the manufacturing process was complicated and could not meet the actual demand. Based on the layer accumulation principle (Chen et al. 2019; Dong et al. 2021; Feng et al. 2021; Yang et al. 2021), 3D printing shows the great potential of realising the structural complexity of absorbers. For example, Mei and co-workers (Mei et al. 2019) combined 3D printing and chemical vapour infiltration to adjust the micro–macro structures of Al₂O₃/SiC_w for the optimised wave absorbing properties. Further, they fabricated absorbers with unique twisted cross metamaterial structures and realised effective EMW absorbing of X band (Mei et al. 2021). Gong et al. (Gong et al. 2021) adopted selective laser sintering (SLS) 3D printing to fabricate carbon fibre/polyamide/carbonyl iron composite honeycomb structure with a RL_min of −47 dB at 16.2 GHz. However, as far as the authors’ knowledge, few works have been done on 3D printed biomass-derived SiC-based absorbing composite materials.

Wood is an abundant renewable biomass resource. The recycling and sustainability of wood waste have always been the focus of people's attention (Song et al. 2018; Guan et al. 2020). This work creatively pulverised wood waste into wood powder and prepared a highly adaptable phenolic resin coated cedarwood powder (PR/WP) by a solvent evaporation method. The PR/WP powder was used to build wood precursors by SLS. Then the wood precursor was treated by carbonisation and SiO vapour carbothermal reduction to obtain porous SiC with in-situ grown SiCnw. The growth of SiCnw could be adjusted by varying the carbothermal reduction temperatures. This work combined SLS 3D printing and carbothermal reduction processes to fabricate biomass-derived absorbing SiCnw/SiC composite, realising the structure–function integrated EMW absorption through a low-cost and environmental strategy.

2. Experimental section

2.1. Materials

The cedarwood waste blocks were obtained from a local wood factory. The phenolic resin was purchased from Henan Hengyuan New Material Co. Ltd. Ferrocene was purchased from Shanghai Aladdin Biochemical Technology Co. Ltd, Ltd. Ethyl alcohol and xylene were purchased from Sinopharm Chemical Reagent Co. Ltd., China. The PR/WP composite powder preparation for SLS is shown in Figure S1. Firstly, the cedarwood blocks with abundant micro holes were pulverised into wood powder below 100 mesh; then, the phenolic resin powder was entirely ultrasonic dissolved in the ethyl alcohol for 1 h; the wood powder was mechanically mixed in the solvent for 4 h at 900 rpm, the mass ratio of phenolic resin powder and wood powder was 1:1; after that, the mixture was vacuum dried for 24 h at 80°C; finally, the obtained PR/WP blocks were mechanically crushed and sieved to get the PR/WP powder below 160 mesh. The micro-morphologies of wood powder and PR/WP powder are illustrated in Figure S2(a, b). The holes of wood powder before and after coating are visible.

2.2. Manufacture of 3D printed biomass–derived SiCnw/SiC composite

The SLS 3D printing of wood biomass–derived green parts (SP) was conducted through the HK C250 with a CO₂ laser (Wuhan Huake 3D Technology Co. Ltd., China). The SLS processing parameters were settled as follows: laser powder 7 W, scanning speed 2000mm/s, scanning spacing 0.1 mm, layer thickness 0.13 mm, preheating temperature 40°C.

The carbonised preform (CP) was obtained by heating the SP at 800°C for 2 h with Ar protection in a tube furnace. Then the CP part was impregnated in a ferrocene catalyst solution under vacuum for 15 min and subsequently dried at 100°C for 2 h. The carbonisation process adopted a step temperature heating procedure, clearly described in Figure S3a.

In the carbothermal reduction process, the CP part was carefully placed on the mixed Si/SiO₂ powder in a corundum boat covered by a corundum lid. The molar ratio of Si and SiO₂ was 1:1, while the weight ratio of their mixing powder and CP part was 4.4:1. The carbothermal reducing process for CP parts was set at 1300°C, 1400°C, and 1500°C for 3 h with Ar protection. The obtained porous biomass-derived absorbing SiCnw/SiC parts were named SC1300, SC1400, and SC1500, respectively. The heating curve of carbothermal reduction is shown in Figure S3b.

2.3. Characterisation

The particle size distribution of the PR/WP powder was tested by a laser particle analyzer (Mastersizer 3000; Malvern Panalytical, UK). The chemical bonding information of materials was analyzed by a Fourier infrared Spectroscopy (FTIR) analyzer (Nicolet 6700; Thermo Scientific, USA). The weight retention of powder samples was obtained by a thermalgravimetric (TG) analyzer (Diamond; PerkinElmer, Germany) at 800°C with a heating rate of 20°C/min. The amorphous and in-situ formed SiC in the samples were analyzed by a Raman spectrometer (LabRAM HR800; Horiba Jobin Yvon, France). The phase composites of the samples were detected by X-ray diffraction (XRD-6100; Shimadzu, Japan). The apparent density, bulk density, and open porosity were tested and calculated using the Archimedes method. The dimensional accuracy (DA) of the samples at different processing stages was calculated as the following equation: (1) $\mathrm{DA}={\displaystyle \frac{D}{D_0}\times 100\text{\%}}$(1) where D is the measured value by a calipre with an accuracy of 0.01 mm, D₀ is the designed value of the CAD model. Rectangle blocks with a dimension of 30 mm (X) ×15 mm (Y) × 4 mm (Z) were used to evaluate the dimensional accuracy of samples. The flexural strength of specimens with a dimension of 22×6×4 mm³ was tested using an electronic dynamic and static fatigue testing machine (E1000; ITW, USA). The span spacing and crosshead speed were 18 mm and 0.5 mm/min, respectively. The microstructures of the samples were observed by a field emission scanning electron microscope (JSM-7600F; JEOL, Japan) and transmission electron microscopy (JEM2100; JEOL, Japan). The distribution and content of elements were analyzed by an electron probe microanalyzer (EPMA-8050G; Shimadzu, Japan) equipped with a wavelength dispersive spectrometer (WDS). The electromagnetic (EM) parameters within the 2∼18 GHz were measured by a vector network analyzer (E5071c; Agilent, USA) through a coaxial line method. The coaxial line samples were directly prepared by SLS, of which the inner diameter, external diameter, and thickness were respectively 3, 7, and 2.5 mm. The thermal conductivities of samples with a dimension of 10×10×2 mm³ were obtained by a laser thermal conductivity analyzer (LFA-427; NETZSCH, Germany) within a temperature range of 25∼900°C while the thermal conductivities at 25°C, 300°C, 600°C, 900°C and 1200°C were tested.

3. Results and discussion

Figure 1 illustrates a schematic of 3D printed porous biomass–derived SiCnw/SiC composite. Before SLS 3D printing, a highly SLS adaptable PR/WP powder is prepared by a solvent evaporation method with timber waste as feedstocks. The reuse of timber waste would contribute to a sustainable manner of wood biomass material and the added value of wood. The waste cedarwood blocks have anisotropic structural characteristics while abundant straight holes with a diameter of 15∼30 μm distributed on the cross-section and the wood fibre structures could be observed from the vertical section (Figure S1). The straight holes and fibre structures could be reserved in the PR/WP powder, significantly affecting the mechanical and EMW absorbing porous SiC ceramics. Moreover, lots of isotropic pores would also form during the SLS process due to the loose stacking state among powders as the spreading roller only applies a weak pressure on the powder bed (Chen et al. 2019; Chen et al. 2019). After carbonisation, amorphous porous carbon preform is formed due to the pyrolysis of organics. In the carbothermal reduction process, SiO vapour can react with carbon to generate porous SiC with in-situ grown SiCnw. 3D printing can endow the absorbing biomass–derived SiCnw/SiC composite with customised macro-shapes for wide structural and functional applications.

Figure 1. Schematic of 3D printed porous biomass–derived SiCnw/SiC composite.

The PR/WP powder morphology is regular with good dispersibility (Figure S1) and a median particle size of approximately 103 μm (Figure 2a), indicating that the solvent evaporation method is an excellent way to prepare feedstock adaptable for SLS 3D printing. The PF coated wood powder is easier to grow sintered neck among particles under the scanning laser. Hence a crosslinked wood network can be processed by SLS.

Figure 2. (a) Particle size distribution of PR/WP powder; (b) FTIR curves of wood powder, PF powder, PR/WP powder, and SP; (c) TG curves (in Ar) of wood powder, phenolic resin powder and PR/WP powder; (d) Raman, (e) XRD and (f) TG (in the air) curves of CP, SC1300, SC1400 and SC1500.

Figure 2b illustrates the FTIR absorption curves of wood powder, phenolic resin, PR/WP powder, and SP, in which absorption peaks at 3424, 2926, 1636, 1510 and 1267 (&1057) cm⁻¹ can be assigned to OH, CH₂, C = O, C = C and C–O-C for wood powder, respectively. The absorption peaks of phenolic resin at around 1636, 1510, and 1267 cm⁻¹ get sharp, which means the C = O, C = C, and C–O-C groups of phenolic resin increase. Compared to the FTIR curves, there are two combined characteristic peaks of wood powder and phenolic resin powder within the solvent evaporation prepared PR/WP powder and SLS processed SP. Therefore, in the processing procedures of material preparation and SLS processing processes, there is no disappearance of old groups or formation of new groups, indicating that only physical changes exist in the above two stages.

The TG curves of materials under an inert atmosphere are shown in Figure 2c. The weight retention of wood powder, phenolic resin, and PR/WP are 20.3%, 56.7%, and 40.0%, respectively. It is known that the conversion of wood powder into amorphous carbon is a complicated process involving water loss, hydrocarbon formation, aromatic polynuclear structures generation, carbon network shrinkage, thermal reaction termination, and carbon structure formation (Qian et al. 2003). Significantly, the carbon yield of phenolic resin is much higher than wood powder owing to lots of aromatic nuclei in the phenolic molecular. The wood powder acts as the main body of the SP skeleton, while the phenolic resin can help maintain the stability of the skeleton. As illustrated in the TG curve of PR/WP, the weight drops tempestuously during 300∼400°C and gradually keeps stable during 750∼800°C, determining the heating routine of the carbonisation process (Figure S3a).

As described in the 2.2 section, the SP samples would experience carbonisation and carbothermal reduction along with phase and microstructure evolutions. In Figure 2d, the Raman spectra of CP, SC1300, SC1400, and SC1500 are given to show the composition evolution with changed reaction conditions. There are two peaks at 785 and 947 cm⁻¹ representing the Si–C vibrations of the transversal optic (TO) and longitudinal optic (LO), while no Si–C vibration appears in the CP characteristic peaks (Xiao et al. 2020; Zhang et al. 2015). Moreover, the LO mode becomes sharper as the reduction temperature increases, attributed to the increased distance among stacking faults (SF) (Xiao et al. 2015). Besides, additional three peaks at around 1360, 1590, and 2700 cm⁻¹ are assigned to carbon characteristic peaks of D-band, G-band, and 2D-band. The intensity ratio of D peak and G peak (I_D/I_G) reveals the defects and degrees of crystal structure disorder in carbon material (Du et al. 2020). For CP samples without carbothermal reduction, the I_D/I_G ratio is 1.463 as vacancies dominate as primary defects. Further, the I_D/I_G ratio decreases with the carbothermal reduction temperature, indicating the increase of crystallization of carbon, and the sp³ defects mainly exist in the SC1500 samples with an I_D/I_G ratio of 0.993 (Lan, Li, and Wang 2020). Meanwhile, the slight fluctuation of 2D bands reflects phases with higher graphitisation presenting all samples (Ma et al. 2020; Xiao et al. 2020).

For a further qualitative explanation of phase evolutions, the XRD spectra of CP, SC1300, SC1400, and SC1500 are described in Figure 2e. The CP sample has two broad peaks at around 22.0° and 44.1°, corresponding to the (0 0 2) and (1 0 0) planes of amorphous carbon (Ye et al. 2019). After carbothermal reduction at different temperatures, the (0 0 2) peak becomes sharper but smaller as the temperature increases, which can be ascribed to the higher crystallinity, as the Raman results revealed. However, the (1 0 0) peak can hardly be detected in SC1300 and disappears in SC1400 and SC1500, suggesting more transformation from carbon to β-SiC through carbothermal reduction at higher temperatures. Clearly, in the SC samples, the peaks located at 35.7°, 41.2°, 60.1°, 71.8°, and 75.4° can be attributed to the (1 1 1), (2 0 0), (2 2 0), (3 1 1), and (2 2 2) planes of the β-SiC (JCPDS PDF#75-0254) (Liang and Wang 2019; Ye et al. 2019). In addition, another small peak at 33.8° is marked as SF in SiC crystals, as previous research reported (Jiang et al. 2018; Yan et al. 2017).

Figure 2f shows the weight retention of CP and SC samples under air with temperature changing from room temperature (25°C) to 800°C, from which it can be found that the weight retention ratio of CP, SC1300, SC1400, and SC1500 are respectively 2.5%, 25.3%, 87.3%, and 99.4%. The residue of CP is mainly metallic oxide impurity while the other SC samples left SiC and the rest carbon is entirely oxidised with temperature increase in air. Thus, we can conclude that 1500°C is more suitable for carbothermal reduction due to almost complete C-SiC transformation, and purer SiC would facilitate high-temperature applications of SLS fabricated components.

Figure 3 shows microstructure evolutions in wood biomimetic SiC ceramic processing. Figure 3a shows the surface morphology of SP. By the synergistic effect of both laser and powder spreading roller, the phenolic resin wraps evenly on the powder and is fully cured to connect the wood powder (Zhu et al. 2018) for constructing a 3D crosslinked biomimetic porous structure. After carbonisation, abundant micropores are observed in Figure 3b, including natural holes of wood powder and SLS acquired pores evenly distributed in the CP fracture surface. Furthermore, the SLS-developed pores are similar to the natural wood pores viewed from surface morphology shown in Figure 3c. During the carbonisation process, the wood powder would transform into soft amorphous carbon while phenolic resin tends to change into hard glassy carbon that seems like a ‘bridge’ to strengthen the bonding between wood powder, thus here forming a relatively solid biomimetic porous carbon network (Qian, Jin, and Wang 2004). All the pyrolytic carbon would serve as the carbon source to react with silicon sources to generate SiC bulks and SiC nanowires. The dense SiCnw arrays uniformly embed on the SiC bulks (Yan et al. 2017) viewed from the fracture surface of SC1300, SC1400, and SC1500 related with Figure 3e, h, and k, respectively. By further observation of the SC1300 (Figure 3f), SC1400 (Figure 3i), and SC1500 (Figure 3l) surfaces, the SiCnw becomes denser as the sintering temperature increases. More prominently seen from the local magnified micrographs, the diameter of SiCnw gets larger with increasing sintering temperatures. Purer SiC bulks aligned with denser and larger SiCnw can positively influence mechanical and EMW absorbing, which will be elucidated after this.

Figure 3. Illustrations of morphology evolutions in the processing of wood biomimetic SiC ceramic: (a) 3D printed wood precursor; (b) and (c) carbonised preform; fracture morphologies of (d) SC1300, (g) SC1400 and (j) SC1500; surface morphologies of (e) SC1300, (h) SC1400 and (k) SC1500; SiCnw morphologies of (f) SC1300, (i) SC1400 and (l) SC1500.

In order to verify the element compositions, contents, and distributions of the newly grown porous SiC, taken SC1500 as an EPMA testing object, there are four elements: C, Si, O, and Fe detected by a mapping scanning the surface. In Figure 4a, the morphology of the SiCnw seems like a tadpole with a globular crystal nucleus on the top. The element distributions display accumulation conditions due to different surface depths of the focus field. By comparing the distribution and content of C (Figure 4b) and Si (Figure 4c), it is clear that C and Si distribute roughly in the same area with a similar mass ratio to SiC molecular. Combined with the XRD characteristic peaks of SC1500, it can be preliminarily determined the generation of high purity SiC. In Figure 4d, a small amount of O shows a different distribution surrounding C and Si, which could be ascribed to SiO₂ in a silicon source. Fe element sporadically distributes in the place of the globular crystal nucleus as exhibited in Figure 4e, related to a vapour–liquid-solid (VLS) mechanism (Cao et al. 2017), which could be explained by the following chemical equations (Zhu et al. 2016; Zhu et al. 2020): (2) $\mathrm{Si}(\mathrm{s})+\mathrm{Si}{\mathrm{O}}_2(\mathrm{s})=2\mathrm{SiO}(\mathrm{g})$(2) (3) $\mathrm{SiO}(\mathrm{g})+2\mathrm{C}(\mathrm{s})=\mathrm{SiC}(\mathrm{s})+\mathrm{CO}(\mathrm{g})$(3) (4) $\mathrm{SiO}(\mathrm{g})+3\mathrm{CO}(\mathrm{g})=\mathrm{SiC}(\mathrm{s})+2\mathrm{C}{\mathrm{O}}_2(\mathrm{g})$(4) (5) $\mathrm{C}{\mathrm{O}}_2(\mathrm{g})+\mathrm{C}(\mathrm{s})=2\mathrm{CO}(\mathrm{g})$(5)

Figure 4. EPMA mapping results of (a) scanning area morphology, the distribution, and contents of (b) C, (c) Si, (d) O, and (e) Fe.

According to Equation 2, the mixed Si and SiO₂ react to generate SiO vapour in the initial stage. Then SiO vapour infiltrates the porous carbon preform and reacts to grow SiC bulks with a few SiC nanowires, considered a vapour–solid (VS) mechanism (Equation 3). However, the VS mechanism cannot produce a considerable number of nanowires. Thus, it is vital to employ ferrocene as a catalyst to form Fe molten droplets, which would act as nucleation sites at high temperatures (Lan and Wang 2020). As described in Equation 4, SiO and CO vapour can dissolve in liquid Fe to form a SiC nucleus. Once the sedimentary SiC approaches saturation point, the extra SiC crystals would precipitate and grow into desired nanowires (Mohammad 2008; Xiao et al. 2017). The rest of CO₂ would also consume carbon to provide CO for SiCnw growth according to Equation 5. As illustrated in Figure 5, this recyclable VLS mechanism plays an essential role in preparing porous biomass-derived SiCnw/SiC composite with superior absorbing ability.

Figure 5. VLS mechanism of 3D printed porous biomass–derived SiCnw/SiC composite.

Figure 6 exhibits the TEM images of SC1500 powder, which gives the exact information of newly formed SiC. As shown in Figure 6a, SiC nanowires have straight bamboo-like shapes with parallel stripes related to close-packed SF arrays (Zhu et al. 2016) according to the Raman and XRD analysis (also seen in Figure 5c). In Figure 6b, the marked SiC nucleus confirms the SiCnw's VLS growing mechanism. By HRTEM analysis for a local region of SiC bulks circled in Figure 6b, the interplanar spacing is 0.25 nm according to the distance among β-SiC (1 1 1) planes. The red dotted line limited region in the HRTEM can be identified as amorphous SiO₂ that grew due to side reaction during carbothermal reduction based on Equation 6 (Cao et al. 2017): (6) $3\mathrm{SiO}(\mathrm{g})+\mathrm{CO}(\mathrm{g})=\mathrm{SiC}(\mathrm{s})+2\mathrm{Si}{\mathrm{O}}_2(\mathrm{s})$(6)

Figure 6. Typical TEM pictures of (a, b) porous SiC (SC1500) and (c) the magnified SiCnw; HRTEM picture of (d) SiCnw.

In Figure 6d, the interplanar spacing of the SiCnw is also 0.25 nm, and a thin nano-silica layer enfolds the single SiC crystal. The formed SiO₂ layers in both SiC bulks, and SiC nanowires would form SiO₂/SiC interfaces and improve the impedance matching ability of newly formed SiC (Liang et al. 2016) due to the wave transparent performance of SiO₂ (Su et al. 2018). Accompanied by the selected area electron diffraction (SAED) in Figure 6d, a typical SiC crystalline structure with crystal defects is observed, in which the characteristic bright line can also suggest the crystalline growing direction along [1 1 1] direction due to the lowest surface energy in reduction processing (Ren et al. 2021; Zhu et al. 2020).

Accuracy and mechanical strength are fundamental indicators achieving the synergetic shape-performance control of 3D printed components, playing a decisive role in the practical application. Figure 7a shows the forming accuracy in the whole process of biomimetic porous SiC. In the SLS process, laser energy input is critical for SP's size and shape (Zhu et al. 2018). The laser selective scanned plane has a higher temperature than the surrounding loose powder. Therefore, thermal energy would transfer into uncured powder and cause secondary size growth (Yan et al. 2009). In general, the secondary size growth of the X-Y plane can be solved by a laser offset setting. However, for PR/WP powder, considerable size growth in Z-direction is hard to eliminate due to the laser's penetration and low curing point of phenolic resin (shown in Figure 7a). After carbonisation, the SP's volume shrinks owing to the pyrolysis of wood powder and phenolic resin. Only slight volume shrinkage occurs during the reaction for the SC samples, and the sizes decrease slightly as temperature increases. In order to fabricate parts with the required sizes, size amplification is usually necessary for model designing. In Figure 7b, the skeleton density, bulk density, and open porosity are calculated according to Archimedes’ test results. The skeleton density goes up as the processing steps proceed and is proportional to the reduction temperature. Significantly, the skeleton density of SC1500 is 3.01 g/cm³, close to the theoretical density of SiC (3.2 g/cm³; JCPDS PDF#75-0254) for the remaining unreacted carbon and inner formed closed pores. The bulk density also shows an upward trend in skeleton density and open pores. The open porosity decreases during carbonisation and increases modestly with increasing carbothermal reduction temperatures. It is worth noting that the flexural strength of samples increases in the whole process, though there are more open pores (Figure 7c), which can be ascribed to phase compositions and micromorphology evolutions elucidated in Figure 2 and Figure 4. Meanwhile, the stress–strain curves of samples are drawn in Figure 7d, and all the samples show typical brittleness characteristics. As cracks expand straightly in the porous samples, the stress-train curve decreases linearly when the stress value increases to the fracture load (Xia et al. 2019). The bulk density, open porosity, and flexural strength of SC1500 are 0.73 ± 0.001 g/cm³, 75.58 ± 0.31%, and 6.21 ± 0.66 MPa, respectively.

Figure 7. (a) Forming accuracy, (b) density and apparent porosity, (c) flexural strength, and (d) stress-strain curves of the samples at different fabrication stages.

It is well known that EMW absorbing abilities of absorption materials are mainly determined by the relative complex permittivity ϵ_r and permeability u_r as expressed by the following equations (Liu et al. 2016): (7) ${\epsilon }_{\mathrm{r}}={\epsilon }^{\mathrm{\prime }}-j{\epsilon }^{\mathrm{\prime }\mathrm{\prime }}$(7) (8) ${\mu }_{\mathrm{r}}={\mu }^{\mathrm{\prime }}-j{\mu }^{\mathrm{\prime }\mathrm{\prime }}$(8) where ϵ′ and ϵ′′ are respectively the real and imaginary part of relative complex permittivity; u′ and u′′ are respectively the real and imaginary part of permeability. As a typical dielectric absorbing material, the magnetic loss ability of SiC is close to 0, and thus we only focus on the relative complex permittivity ϵ_r, of which ϵ′ and ϵ′′ represent the storage and loss performance of EMW absorbing abilities (Wang et al. 2020). The required concentric bulk samples are directly printed by SLS 3D printing, as Figure 1 illustrated, so the performance can reflect 3D printing characteristics such as SLS acquired pores. Figure 8a shows the real part (ϵ′) values of CP, SC1300, SC1400, and SC1500 dependent on the frequency among 2–18 GHz, in which the ϵ′ values of CP and SC1300 show strong frequency-dependent phenomenon while SC1500 has the lowest ϵ′ values between 5.8∼11.3. As is well known to all, the real permittivity ϵ′ can be related to material's polarisation ability resulting from the dipolar and interface polarizability microwave frequency (Watts et al. 2003; Liu, Zhang, and Fan 2008; Yan et al. 2010). The carbon content of SC1500 is much lower than other samples, and thus fewer SiC/C interfaces are introduced, and weaker interface polarizability leads to a lower real permittivity. In Figure 8b, the ϵ′′ values of CP and SC1300 respectively display two prominent peaks at 9.5 and 5.5 GHz, as well as there is a small peak at 14.4 GHz of the SC1500’s ϵ′′ curve, the reason for these peaks can be ascribed to the interfacial relaxation (Yan et al. 2010; Mei et al. 2019). The Cole–Cole semicircle curves of CP, SC1300, SC1400, and SC1500 are shown in Figure S4a∼d. Each arrow-pointed semicircle represents a Debye dipolar relaxation process critical to the EMW absorption mechanism. The SC1500 samples have the most semicircles related to dielectric polarisation and relaxation behaviour (Yan et al. 2017). For SC1500 samples, there are abundant SiCnw within SiC networks, which favours the interfacial polarisation between different SiC surface junctions as they have distinguished microstructures. Meanwhile, many SF in SiCnw will generate heterojunctions, facilitating diploes polarisation when the material is exposed to an electromagnetic field (Liang and Wang 2019; Chen et al. 2021).

Figure 8. (a) and (d) Real permittivity ϵ′, (b) and (e) imaginary permittivity ϵ″, (c) and (f) dielectric loss angle tangent tan of CP, SC1300, SC1400, SC1500 bulks and SC1500 powder.

The ϵ′ and ϵ′′ values together determine the dielectric loss ability that is considered as the main attenuation characteristic of nonmagnetic SiC material, which is dielectric loss angle tangent (tanδ_ϵ= ϵ′/ϵ′′) (Yan et al. 2017). As pictured in Figure 8c, the CP samples possess the highest tanδ_ϵ, and the SC1300 samples possess the second place. According to Debye's theory, the heterogeneous carbon (glassy carbon and soft carbon) containing internal defects contributes to dipolar polarisation, while the multiple interfaces between SiC and C facilitate interfacial polarisation (Liang and Wang 2017). However, high dielectric loss ability usually means high EMW reflection owing to impedance mismatching between absorbing medium and air (ϵ_r  = 1−0 j). In a word, the EMW absorbing ability is not only affected by the dielectric loss ability but also the impedance matching. The reflection loss (RL) is calculated based on transmission line theory by Equations 9 and 10 (Mei et al. 2021; Chen et al. 2021): (9) $Z_{in}=Z_0\sqrt{\frac{{\mu }_r}{{ϵ}_r}}\tanh \left[j\left(\frac{2\pi fd}{c}\right)\sqrt{{\mu }_r{ϵ}_r}\right]$(9) (10) $RL(dB)=20log\left|{\displaystyle \frac{Z_{in}-Z_0}{Z_{in}+Z_0}}\right|$(10) Where $Z_{in}$ and $Z_0$ respectively represent the input impedance of the absorber and the impedance of air; $f$ is the EMW frequency; $d$ is the thickness of the absorber; $c$ is the speed of light.

Figure 9 exhibits the RL values varying with the EMW frequency and the thickness of absorbers computed by tested electromagnetic parameter. CP and SC1300 do not have effective absorption bandwidth (RL<−10 dB, EAB), and their absorbing abilities are much lower than the other two samples due to their impedance mismatching with air. However, they have high relative complex permittivity. For SC1400 and SC1500 samples, the dielectric loss ability tanδ_ϵ of SC1500 is close to the SC1400s, but a little higher (Figure 8c), and the SC1500 sample has the best EMW absorbing ability (Figure 9d&d’). As for SC1500, the minimum RL value is −49.01 dB at 9.8 GHz with a thickness of 2.8 mm, while its EAB value is 3.3 GHz (8.2∼11.5 GHz). The largest EAB value of SC1500 is 5.1 GHz of 11.1∼16.3 GHz when the thickness is 2.2 mm. As for SC1400, the minimum RL value and largest EAB are −24.7 dB at 4.9 GHz with a thickness of 4.0 mm and 2.8 GHz of 9.3∼12.1 GHz at a thickness of 2.0 mm, respectively. In order to study the influence of SLS formed pores on the EMW absorbing ability, the SC1500 powder (Figure 9e) is made into three paraffin bonded comparison groups with SC1500 powder content of 40 wt% (Figure 9e₁), 50 wt% (Figure 9e₂), and 60 wt% (Figure 9e₃), respectively. There are SiC nanowires and porous SiC bulks in the SC1500 powder as the initial feedstock for SLS is the PR/WP powder. By comparison, it can be concluded that SLS-formed pores have achieved a positive effect on the EMW absorbing ability, while the minimum RL value is −31.48 dB at 10.5 GHz, lower than SC1500. Indeed, the wood biomass-derived morphology is also reserved in SC1500 through the whole processing procedures, including SLS, carbonisation, and carbothermal reduction. As illustrated in Figure 8d (SC1500), the EMW absorbing peak corresponding frequency increases with the decrease in thickness, according to the law of quarter-wavelength attenuation as Equation 11(Han et al. 2017; Chen et al. 2021): (11) $d_m=n\lambda /4=nc/\left(4f_m\sqrt{u_r{\epsilon }_r}\right)(n=1,3,5,\cdot \cdot \cdot )$(11) where d_m is the thickness of the sample, λ is the wavelength of microwaves, c is the light velocity in air, f_m is the frequency of the RLmin value.

Figure 9. The reflection loss at 2∼18 GHz of (a) and (a’) CP, (b) and (b’) SC1300, (c) and (c’) SC1400, (d) and (d’) SC1500. The comparison group of (a) the morphology of SC1500 powder, the reflection loss of (e₁) SC40, (e₂) SC50 and (e₃) SC60.

Combining the above analysis, the increasing temperature between 1300∼1500°C contributes to C-SiC conversion. The suitable temperature of carbothermal reduction is 1500°C as the porous SiCnw/SiC composite can obtain the best EMW absorbing ability. Wood morphology is preserved throughout the process, including powder preparation, SLS 3D printing, carbonisation, and carbothermal reduction. Besides, SiC nanowires serve as effective EMW energy consumption agents associated with abundant SF. Moreover, Figure 10a exhibits the EMW absorption mechanism of SLS 3D printed porous biomass-derived SiCnw/SiC composite, as explained above. The SLS process facilitates porous structure for multiple reflection-absorption and can fabricate heteromorphic three-dimensional components as described in Figure 10b∼e. In order to make a direct comparison between our work and traditional methods, Table 1 shows the EMW absorption properties of relevant SiC-based materials reported in recent literature. In our work, the 3D printed porous SiCnw/SiC composite shows more outstanding absorption abilities, realising an RL_min of −49.01 dB at 2.8 mm thickness and an EAB of 5.1 GHz. Further, the timber waste is recycled and reused through our sustainable 3D printing strategy with the advantages of low cost and environmental protection.

Figure 10. (a) The EMW absorbing mechanism of 3D printed porous SiC with SiC nanowires; the pictures of SLS 3D printed components: (b) gear, (c) mirror blank, (d) honeycomb, and (e) carbothermal reduced honeycomb.

Table 1. The comparison of EMW absorption properties of relevant SiC-based materials.

Samples	Preparation process	Optimum thickness (mm)	Minimum RL values (dB)	EAB (GHz)	Reference
SiC aerogels	Freezing/pyrolyzing/carbothermal reduction	2.0	−43	4.0	(Liang and Wang 2019)
Porous SiC	Pressing/impregnation/carbothermal reduction	2.2	−29	3.2	(Lan, Li, and Wang 2020)
SiC whiskers/hollow carbon microspheres	Spray drying/thermal treatment/carbothermal reduction	2.6	−48.60	4.34	(Chen et al. 2021)
C/SiCnw	Electrospun/thermal curing/annealing	1.7	−27.8	4.7	(Wang et al. 2018)
SiCnw/C foam	Freezing/drying/carbothermal reduction	3.3	−31	4.2	(Xiao et al. 2017)
rGO/SiCnw foam	Freezing/carbothermal reduction	3	−19.6	4.2	(Han et al. 2017)
Graphene@SiC aerogels	Directional freeze-casting/freeze-drying/thermal annealing	3	−47.3	4.7	(Jiang et al. 2018)
Porous SiCnw/SiC	SLS 3D printing/carbonisation/carbonthermal reduction	2.8	−49.01	5.1	This work

In Figure 11a, the thermal conductivity of carbonised preform increases rapidly as the temperature increases. The thermal conductivity of the porous SiC parts flattens from 25°C to 1200°C with the increasing sintering temperature. Despite a high porosity of 75.58%, the SC1500 sample still demonstrates excellent thermal conductivity of 3.21∼4.99 W/(m·K) (listed in Table S1) since its skeleton is nearly dense crosslinked three-dimensional structure provides a pathway for heat transfer. With efficient EMW absorption performance, 3D printed porous biomass-derived SiC is a dual-functional material with heat conductive and EMW absorbing functions. In order to further compare the thermal management properties of samples, a simple fire resistance test was conducted to evaluate the fire resistance abilities of CP, SC1300, SC1400, and SC1500 samples (Figure 11b∼d). As the burning time prolongs, the CP and SC1300 samples have severe redness or even burning due to high carbon mass fraction. The SC1400 has slight redness while the burning time is 60 s. In stark contrast to the above, the SC1500 maintains a steady state during the burning process, indicating outstanding thermal stability.

Figure 11. (a) Thermal conductivities of CP, SC1300, SC1400, and SC1500 samples; Fire resistance abilities of (b) CP, (c) SC1300, (d) SC1400, and (e) SC1500 samples.

4. Conclusions

In summary, porous biomass-derived SiCnw/SiC composites were successfully fabricated using SLS 3D printing and carbothermal reduction. Originated from SLS 3D printing, the synthesised SiCnw/SiC composite integrates structure and functions consisting of mechanical, EMW absorption, and heat management properties. The bulk density, open porosity, and flexural strength of porous SiC ceramic are respectively 0.73 ± 0.001 g/cm³, 75.58 ± 0.31%, and 6.21 ± 0.66 MPa. As for EMW absorption performance, the porous SiCnw/SiC composite has a minimum reflection loss of −49.01 dB at 9.8 GHz with 2.8 mm thickness and an effective absorption bandwidth of 5.1 GHz (11.1∼16.3 GHz) with 2.2 mm thickness. Meanwhile, the porous SiC demonstrated excellent thermal conductivities of 3.21∼4.99 W/(m·K) and superior fire resistance ability. Complex structural absorbing SiC components would be 3D printed and applied in harsh environments. The dual-functions of heat conduction and absorbing would solve the heat conduction problem and restrain unnecessary electromagnetic energy coupling, resonance, and surface current generated by electromagnetic interference, suitable for applying in electronic components and equipment. In the future, the influence of macrostructure design will be further studied to explore the effect of structural change on the EMW absorption performance.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgment

We are grateful for Prof. Xian Wang's help in the electromagnetic parameter test for this paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by National Natural Science Foundation of China Aerospace Advanced Manufacturing Technology Research Joint Fund [grant number U2037203, 2020].

Notes on contributors

Changshun Wang

Changshun Wang received his BS in 2020 from Northeast Forestry University (Harbin, China). He is currently a Ph.D. candidate at the School of Materials Science and Engineering at Huazhong University of Science and Technology of Prof. Chunze Yan. His research direction is laser additive manufacturing of structure and function integrated SiC composites.

Siqi Wu

Siqi Wu received her BS in Engineering in 2018 from Wuhan University of Technology (Wuhan, China). She is currently a Ph.D. candidate at the School of Materials Science and Engineering at Huazhong University of Science and Technology under the supervision of Prof. Chunze Yan. Her research interests focus on structural design and mechanical properties of additive-manufactured silicon carbide ceramic structures.

Zhaoqing Li

Zhaoqing Li is an associate researcher of Huazhong University of Science and Technology mainly engaged in additive manufacturing and performance research of ceramics. Before working in HUST, he received his Ph.D. from Lanzhou University. His research interest is 3D printing of ceramics. Currently, as the first author and corresponding author, he has published 6 SCI papers.

Shuang Chen

Shuang Chen received his BS in 2017 from Huazhong University of Science and Technology (Wuhan, China). He is currently a Ph.D. candidate at the School of Materials Science and Engineering at Huazhong University of Science and Technology of Prof. Chunze Yan. His research interest is laser additive manufacturing of structural ceramic materials and their composites.

Annan Chen

Annan Chen is currently an Associate Researcher at the School of Materials Science and Engineering at Huazhong University of Science and Technology. He is mainly engaged in additive manufacturing and performance research of bioceramics. Currently, as the first author and corresponding author, he has published 18 SCI papers in J. Eur. Ceram. Soc., J. Am. Ceram. Soc. and other well-known journals in the field of materials, with a total of more than 380 citations. He is currently the subject editor of Materials journals and reviewers of journals such as Addit. Manuf., J. Eur. Ceram. Soc., J. Alloy. Compd., etc.

Chunze Yan

Chunze Yan is a professor in the School of Materials Science and Engineering at Huazhong University of Science and Technology, Wuhan China. He is currently the Changjiang Distinguished Professor of the Ministry of Education, China. Before joining HUST in 2013, he worked as a research fellow at the College of Engineering, Mathematics and Physical Sciences of the University of Exeter. His research interests focus on 3D/4D printing of polymer and its composite material, 4D printing of piezoelectric ceramic and 3D printing silicon carbide composite for structure-functional applications.

Yusheng Shi

Yusheng Shi Male, born in 1962, Ph.D., Scholar Leading Distinguished Professor of Huazhong University of Science and Technology. He is currently the Vice Director Member of the Special Materials Group of the National Additive Manufacturing Standardization Technical Committee, the Director of the Expert Committee of the Additive Manufacturing Process Technology Center of China Aerospace Science and Technology Corporation, the Director of the National and Local Joint Engineering Laboratory of Digital Material Processing Technology and Equipment (Hubei), the Member of the Expert Committee of China Additive Manufacturing Industry Alliance, the Vice Director Member of the Additive Manufacturing Branch of the Chinese Mechanical Engineering Society, and the Chairman of the Hubei 3D Printing Alliance. He has won one of China's Top Ten Scientific and Technological Advancements, one Second Prize of National Technological Invention, two Second Prizes of National Science and Technological Progress, and seven provincial and ministerial first prizes. He has won the Special Prize of China Invention and Entrepreneurship Award and Contemporary Inventor, Top Ten Outstanding Innovators of Chinese Science, Top Ten National Outstanding Scientific and Technological Worker Nomination Award, Special Government Allowance of the State Council, Individual Award for Major Contributions to Science and Technology in Wuhan City, and Wuyi Labor Medals in Hubei Province. His team was selected as the Innovation Team of Hubei Province and the Ministry of Education.

Haibo Zhang

Haibo Zhang is a professor in the School of Materials Science and Engineering at Huazhong University of Science and Technology, Wuhan China. Before joining HUST in 2011, he worked as a postdoctoral researcher at Kochi University Japan and Humboldt Research Fellow in Technique University of Darmstadt Germany. His research interests include lead-free piezoelectric ceramics and piezoelectric devices, dielectric materials for energy storage applications.

Pengyuan Fan

Pengyuan Fan is currently working in SMOORE International Holding Limited as a Senior Material Engineer. He received his Ph.D. degree in materials science from Huazhong University of Science and Technology in 2018. His research interests f