Introduction

The increasing use of large number of wireless gadgets in contemporary technological world and related electromagnetic (EM) radiation is becoming a serious problem that disturbs stable working conditions of electronic appliances and may harm to human being (Chung 2000). In civil and military aerospace vehicles, protection from EM radiations as well as regulation of thermal heating of electronic power systems is necessary to protect them from any form of damages. In particular, light weight shielding and structural materials are needed to mitigate EM interference (EMI) from electronic systems and to protect human from hazards of space radiation, e.g., aerospace vehicles, especially in 8.2–12.4 GHz frequency region (X-band) (Coleman et al. 2006; Du et al. 2006). The EMI shielding refers to the reflection and absorption of electromagnetic radiation by a material, which thereby acts as shield against the penetration of radiation through the shield. Electrical conductor such as metal and carbon materials mainly shield by reflection of radiation. On the other hand, magnetic materials mainly used for absorption of the radiation. The most materials used for shielding are chosen due to their electrical conductivity rather than their magnetic behavior. While in certain application such as sheath technology, shielding materials have mandatory to absorb maximum EM radiation. The magnetic absorbers depend on the magnetic hysteresis effect, which is attained in magnetic materials such as ferrites. But densities of the magnetic materials are generally high and absorbing bandwidths for magnetic absorbers are usually narrow. On the other hand, dielectric materials are light weight, but do not match up to the absorptivity of magnetic absorbers (Seo et al. 2004; Park et al. 2006; Oh et al. 2004). These two types of materials have different advantages and disadvantages when they are applied as absorbers. They can be used together as a composite, and the magnetic material is usually the base one, but high density of the material is still of great concern. Practically, an effective EMI shielding material being lightweight is an important requirement in aerospace transportation vehicles and space structures (Yang et al. 2005). The shielding effectiveness (SE) value should be >−50 dB in the X-band frequency (Wang et al. 2012). Among the different materials, carbon-based materials are comparatively light weight and structurally strong.

Carbon foam (CF) is a sponge-like rigid high-performance engineering materials in which carbon ligaments are interconnected to each other (Klett et al. 2000, 2004). Recently, it has attracted lot of attention owing to outstanding properties such as low density, large surface area with open cell wall structure, good thermal and mechanical stability (Gallego and Klett 2003; Inagaki et al. 2004).

There are different types of electromagnetic (EM) radiation shielding materials or RAM has been developed since after the advent of radar, but very few reports on the CF are available in literature. As discussed above, generally carbon is conducting materials in which SE is dominated by reflection losses that depend upon the processing temperature of CF and starting precursor.

Yang et al. (2004) reported the development of CF from mesophase pitch by foaming technique which is heated at temperature between 400 and 800 °C and studied its microwave (2–18 GHz) absorption characteristics. It is found that CF heat treated at 600 and 700 °C exhibits better microwave absorption (reflection loss 10 dB). Fang et al. (2006) reported the numerical prediction and experimental validation of CF as microwave (8.2–12.4 GHz) absorber. The CF is fabricated by traditional technique through the polymer foam replication method and it is heat treated at 700, 750 and 800 °C, and characterized them for microwave absorption. The reflection coefficient of 20 mm thick foam is in the order of 8–10 dB. Fang et al. (2007) have also studied the electromagnetic characteristics of CF having different pore size and electrical conductivities which are controlled by the carbonization temperature (700–760 °C). The electromagnetic parameters of these carbon foam and their corresponding pulverized powders are measured by a resonant cavity perturbation technique at a frequency of 2.45 GHz. The CF has dielectric loss several times larger than their corresponding pulverized powder. This suggests that low temperature heat treated foam EM-SE is dominated by absorption. Recently, Moglie et al. (2012) studied the EM-SE of CF (GRAFOAM FPA-20 and FRA-10) in the frequency band 1–4 GHz using the nested reverberation chamber method. It is reported that with increasing thickness of CF total SE increases. Micheli and Marcheti (2012) reported the effect of multi-walled carbon nanotubes epoxy resin mixture filled in the pores of CF and studied its role on EM radiation absorption properties of CF in frequency range 2–3 GHz. It is observed that absorption properties reaching peak of the reflection coefficient −45 dB for a normally incident plane wave. Blacker et al. (2008) reported that EM-SE greater than about 40 dB in the frequency range of 400 MHz–18 GHz of rigid porous electrically conductive CF is achieved by incorporating electrically conductive carbon nanofiber in a polymer matrix, which is used for the joining the CF enclosures. Blacker et al. (2011) reported that the development of electrically gradated CF material exhibits different electrical resistivity values at or near different surfaces of CF in which the electrical resistivity increases with increasing the thickness of foam and decreases with increasing processing temperature. These electrically gradated CF materials may be used as radar absorbers. But in the case of aerospace and military application, the shield material should have high SE which is mainly dominated by absorption losses as well thermal stability.

The carbon nanomaterials in combination with magnetic particles are used to improve EM radiation absorption and have become more intensive absorber as compared to carbon nanomaterials (Huang and Wu 2000; Liu et al. 2006; Che et al. 2004). Che et al. (2004, 2006) have reported that magnetic nanoparticles/carbon nanotubes (CNTs) exhibited improved microwave absorption properties because of their proper EM matching between the dielectric loss and the magnetic loss.

The purpose of this study is to improve the microwave absorption properties of carbon foam by the incorporation of magnetic nanoparticles. In the present work, CF is developed from a mixture of coal tar pitch and Ni nanoparticles by sacrificial template technique and then heat treated to 1000 °C. The results reveal that SE is dominated by EM radiation absorption mechanism rather than reflection mechanism in the Ni nanoparticle-embedded CF.

Experimental

Synthesis of nickel nanoparticles

The Ni nanoparticles were synthesized by the process reported literature (Saravanan et al. 2001). The 1 g of nickel chloride hexahydrate (NiCl2·6H2O) was dissolved in 100-ml ethylene glycol. The solution was heated to 70–80 °C. Later on, 4.5-ml hydrazine hydrate was added to above solution for complete reduction. After 10 min, 3.0 g of NaOH (1 M) solution was added slowly in continuation of stirring. The complete solution was kept under magnetic stirrer for 1 h at 60 °C. After 1 h, reaction was completed and black nickel nanoparticles were formed. These nanoparticles were collected, washed several times by ethanol and dried in the room temperature. The reduction reaction can be expressed as

$$2 {\text{NiCl}}_{ 2} + {\text{H}}_{ 4} {\text{N}}_{ 2} + 4 {\text{NaOH }} \to {\text{ 2Ni}}^{ + } + {\text{ N}}_{ 2} + {\text{ 4H}}_{ 2} {\text{O }} + {\text{ 4NaCl}}$$

It is seen that N2 gas is continuously bubbled during the reaction and as results it automatically creates N2 atmosphere throughout the reaction. Hence, no extra N2 gas is required for the synthesis of nickel nanoparticles.

Fabrication of nickel nanoparticles incorporated carbon foam

Initially, coal tar pitch was modified with the nickel nanoparticles 1.0 wt% and heat treated at 400 °C for 25 h in inert atmosphere. The carbon foam was prepared by sacrificial template technique in which the polyurethane (PU) foam (density 0.030 g/cc and average pore size 0.45 mm) was used as template. The Ni nanoparticle-based modified pitch was grounded by ball milling in a tungsten carbide jar for about 5–6 h to get the particle size of mixture less than 30μm, so that these particles can easily penetrate inside the cells of PU foam. The above mixture was converted into water-based slurry (with 3 % polyvinyl alcohol) and it was impregnated inside pore of PU template foam by vacuum infiltration technique (Kumar et al. 2013a, b, c; Yadav et al. 2011). The impregnated PU foam was converted into carbon foam by series of heat treatments in air as well as in an inert atmosphere up to 1,000 °C. Initially, the modified coal tar pitch and Ni nanoparticles mixture impregnated PU foams were heat treated @ 1 °C min−1 up to 275 °C in the nitrogen atmosphere for 1 h followed by oxidation and stabilization in air atmosphere at temperature 300 °C. The stabilized foam was carbonized in tubular high temperature furnace at 1,000 °C with heating rate 10 °C h−1 in inert atmosphere. The schematic of process for synthesis of nickel nanoparticle-based CF is shown in Fig. 1. The CF is designated as CF and Ni nanoparticle-embedded CF as CF-Ni.

Fig. 1
figure 1

Fabrication of Ni nanoparticle-embedded coal tar pitch-based carbon foam using sacrificial (PU foam) template technique

Full size image

Characterization

Raman spectra of the CF and Ni-CF samples were recorded using Renishaw in via Raman spectrometer, UK with laser as an excitation source at 514 nm. The crystal structure of CF samples was studied by X-ray diffraction (XRD, D–8 Advanced Bruker diffractometer) using CuKα radiation (λ = 1.5418 Å). The morphology of Ni nanoparticles was observed by transmission electron microscope (TEM, Tecnai, G20-Super-Twin, 300 kV). The TEM sample was prepared by Ni nanoparticle dispersed in solvent and put over the carbon-coated copper grid. The surface morphology of the CF samples was observed by scanning electron microscope (SEM, VP-EVO, MA-10, Carl-Zeiss, UK) operating at an accelerating potential of 10.0 kV. The electrical conductivity CF (size 60 × 20 × 4 mm) was measured as per standard ASTM C611-98. The dc four probe contact method was used in which a Keithley 224 programmable current source was used for providing current. The voltage drop was measured by Keithley 197A auto ranging digital microvoltmeter. The values reported in text are average of six readings of voltage drops at different places of the sample. The compressive strength of CF was measured by Instron Universal Testing Machine, model 1122 as per ASTM C695-91. EM-SE and EM attributes (complex permittivity and permeability) were measured by waveguide using vector network analyzer (VNA E8263BAgilent Technologies). The rectangular samples (26 × 13 mm) of thickness 2.75 mm were placed inside the cavity of sample holder which matches the internal dimensions of X-band (8.4–12.4 GHz) wave guide. The sample holder was placed between the flanges of the waveguide connected between the two ports of VNA. The magnetic property of the CF samples was measured by vibration sample magnetometer (VSM) model 7304, Lakeshore Cryotronics Inc., USA with a maximum magnetic field of 1.2 T, using Perspex holder vibrating horizontally at frequency of 76 Hz.

Results and discussion

In Fig. 2a, TEM image of Ni nanoparticles shows that the particles are of different shape and size, particularly somewhat star or starfruit shaped and it looks like agglomerated form. The average size of Ni nanoparticles is in between 50 and 100 nm. Figure 2b shows the powder XRD patterns of the Ni nanoparticles. The nanoparticles are single-phase with face-centered cubic (fcc) structure, and no phase of NiO or other impurity is observed. The peaks at 2θ = 44.45º, 51.73º and 76.84º correspond to diffraction plane of Ni (111), Ni (200) and Ni (220) reflections, respectively (The international centre for powder diffraction data, powder diffraction files 2001). This result is subsequently confirmed by the TEM examination of the annealed sample where larger metallic Ni nanoparticles are clearly observed.

Fig. 2
figure 2

a TEM micrograph and b XRD spectra of Ni nanoparticles

Full size image

The bulk density of CF is 0.50 g cc−1 and on addition of Ni nanoparticles its density increases to 0.56 g cc−1 even at 1 wt% of nickel nanoparticles. The Ni nanoparticles are in magnetic behavior, but in CF its acts also as an electrical conductivity enhances. The Ni has affinity toward carbon and it forms Ni–C. The CF is heat treated to 1,000 °C, at this temperature graphitic structure does not develop and as a result it is in amorphous nature and it possesses combination of sp2 and sp3 hybridized carbon. The Ni can readily react with sp3 bonded carbon and as a result sp3 contribution in CF-Ni decreases. This has positive effect on the electrical conductivity. The electrical conductivity of CF is 50 S cm−1 which is due to the delocalized π electron in the carbon network while electrical conductivity of CF-Ni increases to 65 S cm−1, this is due to the increases in sp2 carbon content in it as compared to CF.

The mechanical strength is one of the essential requirements of CF because compressive forces are often encountered during its application as structural component. Therefore, compressive strength of CF should be sufficient to avoid any form of structural damage. The compressive strength of CF depends mainly on two factors namely microstructure and bulk density. It is observed that the compressive strength of CF is 7.0 ± 1 MPa and that of CF-Ni is 8.8 ± 1.2 MPa. This enhancement in strength is related to increase in bulk density and decrease in stress concentration center.

Figure 3 depicts the SEM micrographs of CF and CF-Ni. In case of CF, cells of the different sizes and distributions of cells are not uniform (Fig. 3a). The cell walls (i.e., ligaments) are broken during machining of samples for SEM characterization because of the brittle and porous nature of material (Kumar et al. 2013a, b, c). Figure 3b, SEM image of CF-Ni, in which the brightness of the surface increases due to the increase in electrical conductivity of CF-Ni by catalytic carbonization. The Ni nanoparticles are embedded in bulk carbon as well as on the surface and pore walls. Figure 3c shows magnified view of the pore walls in which embedded Ni nanoparticles are visible because the melting point of Ni is well above 1,400 °C.

Fig. 3
figure 3

SEM micrographs of a CF, b CF-Ni, c Ni nanoparticles embedded with carbon on the surface

Full size image

Figure 4a shows the XRD spectra of CF and CF-Ni. In case of CF, there are two prominent peaks at 2θ = 24.85 and 43.15°, which correspond to the carbon of (002) and (101) diffraction plane. The XRD spectra of CF-Ni consist of peaks at 2θ = 25.55º, 42.76º, 44.45º, 51.77º and 76.35º corresponding to carbon (002) and (101), Ni crystal lattice (111), (200), and (220) planes. In case of CF-Ni, the 002 peak appears at higher diffraction angle. This suggests, Ni nanoparticle reacts with carbon, it acting as catalyst in improving the crystallite structure of carbon. The catalytic effect of carbon influences the structure in which sp2 hybridization content increases. This also confirmed by Raman spectroscopy investigation. Figure 4b shows the Raman spectra of CF and CF-Ni and illustrates common features in the Raman shift 1,000–3,000 cm−1 region, the G, D and 2D band which lies at around 1,560, 1,360 and 2,700 cm−1, respectively. The G band corresponds to the E 2 g phonon at the Brillion zone center. The D band is due to the breathing modes of sp2 atoms and requires a defect for its activation and it only gives knowledge to the amount of disorder in the given structure (Ferrari and Robertson 2000; Dhakate et al. 2011). Figure 3b shows the Raman spectra of CF and CF-Ni, consisting of two bands D and G appearing at Raman shift 1,350 and 1,598 cm−1, respectively. In case of CF, both bands almost appear at same intensity suggest the amorphous-type carbon while in case CF-Ni the G band intensity is higher as compared to D band. This suggests the structure of carbon modified with addition of Ni. The intensity ratio of D and G peak, i.e., (ID/IG) gives idea of defect level in the structure of CF, i.e., sp3 bonded carbon. The ID/IG ratio in CF is higher this may be due to the presence of lesser amount sp2 bonded carbon i.e., more defects. The CF has ID/IG ratio 0.9645 while in case of CF-Ni it decreases to 0.7065. This demonstrates that the Ni embedded in CF has positive effect on ID/IG ratio. The results of XRD and Raman spectroscopic are in agreement with each other and support the electrical conductivity data.

Fig. 4
figure 4

a XRD spectra and b Raman spectra of CF and CF-Ni

Full size image

Electromagnetic shielding can be explained by measuring SE in terms of reflection and absorption losses. The SE of any shield material is the capacity to attenuate EM radiation that can be expressed in terms of ratio of incoming (incident) and outgoing (transmitted) power (Yong et al. 2005). The EM attenuation offered by a shield may depend on the three mechanisms: reflection of the wave from the front face of shield, absorption of the wave as it passes through the shield and multiple reflections of the waves at various interfaces (Han and Wang 2007). Therefore, SE is endorsed to three types of losses, i.e., reflection loss (SER), absorption loss (SEA) and multiple reflection losses (SEM) and can be expressed (Eq. 1) as:

$${\text{SE}}_{\text{T}} \left( {\text{dB}} \right) \, = {\text{ SE}}_{\text{R}} + {\text{SE}}_{\text{A}} + {\text{SE}}_{\text{M}} = { 1}0{ \log }\left( {P_{\text{t}} /P_{i} } \right)$$
(1)

where P i and Pt are power of incident and transmitted EM waves, respectively. As, Pt is always less than P i , therefore, SE is a negative quantity such that a shift towards more negative value means increase in magnitude of SE. It is important to note that the losses associated with multiple reflections can be ignored (SEM ~0) in all practical cases when achieved SET is more than −10 dB. Therefore, SET can be expressed as (Eq. 2)

$${\text{SE}}_{\text{T}} \left( {\text{dB}} \right) \, = {\text{SE}}_{\text{R}} + {\text{SE}}_{\text{A}}$$
(2)

Figure 5a shows the EM-SE of CF and CF-Ni in the X-band (8.2–12.4 GHz) frequency region. It is observed that total shielding effectiveness (SET) varies with frequency in both the cases. The value of SET for CF is 25 ± 3 dB while that of CF-Ni is 61 ± 5 dB. It is interesting to note that in CF-Ni, SET is more than double to that of CF. In case of CF, the SET is shared in ratio of reflection:absorption losses::1:2, i.e., reflection losses are −8.5 ± 2 dB and absorption losses are −16.86 ± 1 dB. While in CF-Ni, SET is governed by absorption losses (−48.5 ± 4 dB) and partially shared by reflection losses (SER = −12.4 ± 1 dB).

Fig. 5
figure 5

a EMI SE SEA, SER, SET of CF and CF-Ni, b Real permittivity and permeability (ε′,μ′), c imaginary permittivity and permeability (ε″,µ″) and d room temperature magnetization plot of CF and CF-Ni

Full size image

The enhancement in the reflection components from −8.5 dB to 12.4 dB in case of CF-Ni is due to the increase in the electrical conductivity of CF-Ni as discussed in earlier section. While absorption losses enhancement from 16.86 to −48.5 dB in CF-Ni is due to the magnetic properties of material. Furthermore, in CF-Ni, Ni particles of size 200–300 nm on the surface of CF and embedded in CF provide higher surface and large interfacial area because of high aspect ratio of nanoparticles, which further enhances the SE due to the absorption.

To probe further, EM parameters which are responsible for EM radiation absorption, relative complex permittivity (ε* = ε′ − iε″) and relative complex permeability (µ* = µ′ − iµ″) of CF were measured in the frequency region 8.2–12.4 GHz to correlate with shielding properties. These complex parameters have been estimated from experimental scattering parameters (S11 and S21) by standard Nicholson–Ross and Weir theoretical calculation (Nicolson and Ross 1970; Singh et al. 2012). The estimated real part of the EM parameters (ε′, µ′) is directly associated with the amount of polarization occurring in the material which symbolizes the storage ability of the electric and magnetic energy, while imaginary part (ε″, µ″) signifies the dissipated electric and magnetic energy. From Fig. 5b and c, it is clearly demonstrated that both real and imaginary parts of the permittivity (ε′, ε″) vary with frequency. The real permittivity in CF is lower as compared to the CF-Ni and it decreases with increasing the frequency in CF-Ni.

The imaginary permittivity in CF-Ni increases with increasing the frequency (Fig. 5c). The decreasing permittivity in CF-Ni with frequency could be ascribed to the decreasing capability of the dipoles to sustain the in-phase movement with speedily pulsating electric vector of the incident radiation. According to EM theory (Colaneri and Shacklette 1992; Das et al. 2000), the ac conductivity (σac) and skin depth (δ) are related to the imaginary permittivity (ε″), frequency (ω) and real permeability (μ′) as σac = ωεοε″, (σ = σac + σdc) and \(\delta = \sqrt {2/\sigma_{\text{AC}} \omega \mu '}\) which gives reflection (SER) and absorption losses (SEA) as:

Reflection loss (SER) as

$${\text{SE}}_{\text{R}} ({\text{dB}}) = 10\log \{ \sigma_{\text{ac}} /16\omega \varepsilon_{0} \mu '\}$$
(3)

and absorption loss (SEA) as

$${\text{SE}}_{\text{A}} ({\text{dB}}) = 20\{ t/\delta \} \log e = 20d\sqrt {\mu \omega \sigma_{\text{ac}} /2} \log e = 8.68\{ t/\delta \}$$
(4)
$${\text{SE}}_{\text{A}} ({\text{dB}}) = 8.68t\sqrt {\sigma \omega \mu '/2}$$
(5)

From the Eq. (3), it deduces that SER is related to frequency (ω), conductivity (σac) and real permeability while the ac conductivity (σac = ωεοε″) depends upon the frequency and imaginary permittivity (ε″). As shown in Fig. 5b, real permeability increases with increasing the frequency and it is maximum in case of CF-Ni and minimum in CF. While imaginary permittivity (Fig. 5c) also varies with frequency and it is minimum in case of CF and maximum in CF-Ni. This clearly demonstrates that SER is minimum in cases of CF and maximum in CF-Ni due to the higher value of electrical conductivity and imaginary permittivity. The higher value of imaginary permittivity does not store energy, but stored energy will be converted into thermal energy inside the material, which attenuates the EM radiation instead of just storing it. The carbon can easily dissipate the thermal energy due to its high value thermal conductivity.

Equation (5) shows that SEA depends upon thickness, real permeability, frequency and conductivity of the shield material, from Fig. 5b, maximum real permeability is in case of CF-Ni as compared to CF where thickness is same in both the cases.

In case of CF-Ni, the existence of interfaces between Ni nanoparticles and carbon is responsible for interfacial polarization which further contributes to dielectric losses. Interfacial polarization occurs in heterogeneous materials due to the accumulation of charges at the interfaces and the formation of large dipoles. The magnetic nanoparticles act as tiny dipoles which get polarized in the presence of EM field and as a result better absorption.

The real permeability value of CF-Ni is more as compared to than that of CF (Fig. 5b). This is due to the improvement of magnetic properties along with the parallel reduction of eddy current losses since embedded Ni nanoparticles in magnetic behavior. In the X-band frequency region, the natural resonances can be attributed to the small size Ni particles in CF. Anisotropy energy of the small size materials, especially in the nanoscale, would be higher due to the surface anisotropic field effect of smaller material (Leslie-Pelecky and Rieke 1996). The higher anisotropy energy is also contributed in the enhancement of the EM radiation absorption.

The real and imaginary parts of complex permeability increase with increasing frequency in CF-Ni and maximum at frequency ~11 GHz. While in the CF, it is very small due to the non-magnetic behavior. The magnetic nanoparticles embedded in the CF lead to better matching of input impedance along with the reduction of skin depth. This is attributing further in the increase of absorption losses of CF-Ni. This fact also varied by measuring the saturation magnetization of the CF and CF-Ni. Figure 5d shows the room temperature magnetization plot of CF and CF-Ni. The data of magnetization reveal that CF does not show any magnetization though out the magnetic field because carbon is non-magnetic in nature. However, CF-Ni shows the magnetization that displays narrow hysteresis loop. The CF-Ni possesses saturation magnetization 2.5 emu g−1 at 4.9 kg. These results are in agreement with results of electromagnetic attributes. This clearly demonstrates that even small amount of magnetic nanoparticles embedded in conducting CF is very much effective for absorption of EM radiation due to its large surface area and its magnetic properties.

Conclusions

In this investigation, Ni nanoparticle-embedded CF is developed by sacrificial template technique from a mixture coal tar pitch and Ni nanoparticles. It reveals that addition of Ni nanoparticles in the CF influences electrical conductivity, complex permittivity and permeability, mechanical strength and EM-SE. The Ni nanoparticles are magnetic in nature, even though addition of nanoparticles improved the electrical conductivity of CF. This is due to catalytic effect of Ni nanoparticle in the carbon material which is revealed by XRD and Raman studies. The enhancements in electrical conductivity not significantly influence the EM-SE by reflection losses, but it is dominated by absorption of EM radiations due to the magnetic properties and large surface area of nanoparticles. Results of interfacial polarization occur in heterogeneous material (Ni nanoparticle-embedded CF). The shielding effectiveness of Ni nanoparticle-embedded CF increases by 144 % and absorption losses by 170 %. In addition to this, compressive strength of the CF increases from 7 to 8.8 MPa. This clearly demonstrates that the addition of small amount of magnetic nanoparticles in CF can be useful as EM radiation or radar absorbing material which can be used for stealth technology in civil and military applications.