Preparation and Characterization of Highly Elastic Foams with Enhanced Electromagnetic Wave Absorption Based On Ethylene-Propylene-Diene-Monomer Rubber Filled with Barium Titanate/Multiwall Carbon Nanotube Hybrid

Hybrid ethylene-propylene-diene-monomer (EPDM) nanocomposite foams were produced via compression molding with enhanced electromagnetic wave absorption efficiency. The hybrid filler, consisting of 20 phr ferroelectric barium titanate (BT) and various loading fractions of multi-wall carbon nanotubes (MWCNTs), synergistically increased the electromagnetic (EM) wave absorption characteristics of the EPDM foam. Accordingly, while the EPDM foam filled with 20 phr BT was transparent to the EM wave within the frequency range of 8.2–12.4 GHz (X-band), the hybrid EPDM nanocomposite foam loaded with 20 phr BT and 10 phr MWCNTs presented a total shielding effectiveness (SE) of ~22.3 dB compared to ~16.0 dB of the MWCNTs (10 phr). This synergistic effect is suggested to be due to the segregation of MWCNT networks within the cellular structure of EPDM, resulting in enhanced electrical conductivity, and also high dielectric permittivity of the foam imparted by the BT particles. Moreover, the total SE of the BT/MWCNTs loaded foam samples remained almost unchanged when subjected to repeated bending due to the elastic recovery behavior of the crosslinked EPDM foamed nanocomposites.


Introduction
In recent years, the high influx of electronic devices has increased the number of electromagnetic (EM) contaminants around the world, which is called electromagnetic interference (EMI) [1]. EMI can disturb the performance of electronic devices, and stop them from functioning. Metals, which have traditionally been used to counteract the negative effects of EM waves, have disadvantages such as high cost and weight, low corrosion resistance, and most importantly, their inability to absorb electromagnetic waves [2]. Thus, conductive polymer nanocomposites have been studied as potential EM absorbers due to their lightness, high corrosion and chemical resistance, ease of processability, and low production cost. Additionally, these materials are able to absorb and dissipate EM wave energy to heat [3,4]. Characterization: The crystal structure of BT was characterized by Raman spectroscopy using a Renishaw InVia Confocal Raman microscope (Renishaw plc, Wotton-under-Edge, UK). The analyses were performed at 514 nm excitation wavelength and resolution of 0.02 cm −1 .
The porous structure of the foamed BT/MWCNTs, BT, and MWCNTs nanocomposites was examined using a scanning electron microscope (Philips ESEM XL30, Amsterdam, Netherlands) at 25 kV. The prepared foamed samples were cryo-fractured in liquid nitrogen, and the fractured surfaces were Au/Pd sputter-coated. The images were processed using Digimizer software to analyze the cells diameter and distribution as well as cells density. The cell density was calculated by using the following relationship: N = (nM 2 /A) 1.5 (1) where n, M, and A are denoted as the number of cells, magnification factor, and area, respectively. AC electrical conductivity of the cured BT/MWCNTs, BT, and MWCNTs nanocomposite foams was measured using an ALPHA high-resolution Novocontrol broadband dielectric spectrometer (Novocontrol Technologies GmbH, Hundsangen, Germany) at room temperature over a frequency range region of 10 −1 Hz-10 MHz. Cylindrical samples, 22 ± 0.1 mm in diameter and 1 ± 0.1 mm thickness, were placed in the dielectric cell between two parallel gold-plated electrodes, 20 and 30 mm in diameter for the top and bottom electrodes, respectively, and applying 1 V as the amplitude of the AC electrical current. To eliminate contamination, the electrodes and specimen surfaces were cleaned with ethanol before testing.
The EMI SE properties were obtained by measuring the real and imaginary parts of the dielectric permittivity (ε and ε") and magnetic permeability (µ and µ") within the X-band frequency range employing a vector network analyzer (Agilent E8364B). The test was conducted on 3 mm thick samples and the results are the average of five samples for each foam. The measurement errors were, on average, less than 3%, for all the samples over the X-band frequency region. To investigate the effect of repeated bending on EMI SE, the hybrid nanocomposite foam, containing 10 phr MWCNTs, was subjected to 1000 bending with a radius of 2.0 mm and the EMI SE was obtained.
The thermal conductivity of samples was measured in accordance with ASTM E 1530 using FOX 50 thermal conductivity analyzer on cylindrical samples, 51 ± 0.1 mm diameter and 12.5 ± 0.1 mm thickness. The experimental error of this measurement is less than 5% of the absolute value.

Results and Discussion
Among the different crystal structures of BT, the tetragonal type presents higher polarization ability and microwave absorbing properties when subjected to an external EM wave [22,23]. As the Curie temperature of BT (120-130 • C) is relatively low, we first examined the possible effects of the compounding process upon its crystalline structure. For this purpose, raman spectroscopy, which has been known as a powerful tool to analyze the crystalline structure of BT [24], was performed on the neat BT powder and corresponding BT filled EPDM foam. Figure 1 shows the spectra of BT powder and EPDM foam filled with 20 phr BT (B20C0) with no significant difference between the absorption compounding process upon its crystalline structure. For this purpose, raman spectroscopy, which has been known as a powerful tool to analyze the crystalline structure of BT [24], was performed on the neat BT powder and corresponding BT filled EPDM foam. Figure 1 shows the spectra of BT powder and EPDM foam filled with 20 phr BT (B20C0) with no significant difference between the absorption peaks at 305 and 718 cm −1 , which are assigned to the tetragonal phase of BT [25]. This evidences that the compounding and foaming processes have not affected the tetragonal structure of BT crystals and, hence, its piezoelectric behavior is expected to remain unchanged.  The images were further analyzed to determine the foam morphology including cell size, distribution, and cell density ( Figure 3). As can be observed, the cells sizes decrease from 311 ± 84 μm to 67 ± 35μm for the foamed samples containing 20 phr BT and 20 phr BT/10 phr MWCNTs, respectively. Moreover, as observed in Figure 3a, the presence of MWCNTs resulted in smaller cell sizes and a narrower cell size distribution, implying a more homogeneous cell morphology. These results together with the increase in the cells density (Figure 3c) are indicative of the nucleating effect of the CNT particles during the foaming process [26]. Similar results have also been reported for foam nanocomposites composed of various types of nanofillers fabricated using both chemical and physical (supercritical carbon dioxide) blowing agents [26][27][28]. The influence of the MWCNTs on the density and porosity of the foams is also presented in Figure 3d, where an opposite trend is observed. All these results, morphology and density, are ascribed to the reduction in curing scorch time, and hence the onset of the crosslinking reaction by the presence of MWCNTs particles, which limits the time for the foam expansion within the mold. This has been reported and discussed in our recently published paper [29]. Meanwhile, the apparent higher density of the samples composed of BT is due to the higher density of the BT (6.02 gr/cm 3 ). Moreover, these results lead to the conclusion that BT does not play any significant role in the development of the cellular structure of the EPDM foams compared with MWCNTs.  Figure 2 compares the SEM micrographs of the EPDM foamed samples comprising different amounts of BT/MWCNTs hybrid. The images were further analyzed to determine the foam morphology including cell size, distribution, and cell density ( Figure 3). As can be observed, the cells sizes decrease from 311 ± 84 µm to 67 ± 35µm for the foamed samples containing 20 phr BT and 20 phr BT/10 phr MWCNTs, respectively. Moreover, as observed in Figure 3a, the presence of MWCNTs resulted in smaller cell sizes and a narrower cell size distribution, implying a more homogeneous cell morphology. These results together with the increase in the cells density (Figure 3c) are indicative of the nucleating effect of the CNT particles during the foaming process [26]. Similar results have also been reported for foam nanocomposites composed of various types of nanofillers fabricated using both chemical and physical (supercritical carbon dioxide) blowing agents [26][27][28]. The influence of the MWCNTs on the density and porosity of the foams is also presented in Figure 3d, where an opposite trend is observed. All these results, morphology and density, are ascribed to the reduction in curing scorch time, and hence the onset of the crosslinking reaction by the presence of MWCNTs particles, which limits the time for the foam expansion within the mold. This has been reported and discussed in our recently published paper [29]. Meanwhile, the apparent higher density of the samples composed of BT is due to the higher density of the BT (6.02 gr/cm 3 ). Moreover, these results lead to the conclusion that BT does not play any significant role in the development of the cellular structure of the EPDM foams compared with MWCNTs.
Polymers 2020, 12, x FOR PEER REVIEW 4 of 15 compounding process upon its crystalline structure. For this purpose, raman spectroscopy, which has been known as a powerful tool to analyze the crystalline structure of BT [24], was performed on the neat BT powder and corresponding BT filled EPDM foam. Figure 1 shows the spectra of BT powder and EPDM foam filled with 20 phr BT (B20C0) with no significant difference between the absorption peaks at 305 and 718 cm −1 , which are assigned to the tetragonal phase of BT [25]. This evidences that the compounding and foaming processes have not affected the tetragonal structure of BT crystals and, hence, its piezoelectric behavior is expected to remain unchanged.  The images were further analyzed to determine the foam morphology including cell size, distribution, and cell density ( Figure 3). As can be observed, the cells sizes decrease from 311 ± 84 μm to 67 ± 35μm for the foamed samples containing 20 phr BT and 20 phr BT/10 phr MWCNTs, respectively. Moreover, as observed in Figure 3a, the presence of MWCNTs resulted in smaller cell sizes and a narrower cell size distribution, implying a more homogeneous cell morphology. These results together with the increase in the cells density (Figure 3c) are indicative of the nucleating effect of the CNT particles during the foaming process [26]. Similar results have also been reported for foam nanocomposites composed of various types of nanofillers fabricated using both chemical and physical (supercritical carbon dioxide) blowing agents [26][27][28]. The influence of the MWCNTs on the density and porosity of the foams is also presented in Figure 3d, where an opposite trend is observed. All these results, morphology and density, are ascribed to the reduction in curing scorch time, and hence the onset of the crosslinking reaction by the presence of MWCNTs particles, which limits the time for the foam expansion within the mold. This has been reported and discussed in our recently published paper [29]. Meanwhile, the apparent higher density of the samples composed of BT is due to the higher density of the BT (6.02 gr/cm 3 ). Moreover, these results lead to the conclusion that BT does not play any significant role in the development of the cellular structure of the EPDM foams compared with MWCNTs.  Electrical conductivity is a vital index to evaluate the potential EMI absorption and shielding performance of conductive nanocomposites. The AC conductivity (AC) of a dielectric material presents the extent of its frequency-dependent behavior. Figure 4a displays the variation of AC conductivity as a function of frequency for the prepared foamed samples. As can be observed, above 2 phr MWCNTs, the foams exhibit frequency-independent behavior (insulation-conduction transition) implying the formation of interconnected conductive paths [30]. This behavior is commonly described in terms of the percolation of CNT particles throughout the polymer matrix [6]. Additionally, the DC electrical conductivity (DC) of the samples can be extracted by the following relationship where ω = 2πf, f is frequency, A is a constant, and s is a power-law exponent (s = 1 for polymers). Thus, DC value can be obtained by extrapolating the AC to low frequencies (Figure 4b). The insulating nature of samples below the percolation threshold of the MWCNTs (≤2 phr) is clearly observed with DC conductivities below 10 −12 S/cm. As the percolation threshold is passed, the DC electrical conductivity increases several orders of magnitude to 10 −5 S/cm, 4  10 −4 S/cm, and 9  10 −3 for the foamed samples filled with 4 phr, 10 phr MWCNTs and 20 phr BT / 10 phr MWCNTs, respectively, which are suitable for EM wave absorbing applications [31,32]. Therefore, the DC Electrical conductivity is a vital index to evaluate the potential EMI absorption and shielding performance of conductive nanocomposites. The AC conductivity (σ AC ) of a dielectric material presents the extent of its frequency-dependent behavior. Figure 4a displays the variation of AC conductivity as a function of frequency for the prepared foamed samples. As can be observed, above 2 phr MWCNTs, the foams exhibit frequency-independent behavior (insulation-conduction transition) implying the formation of interconnected conductive paths [30]. This behavior is commonly described in terms of the percolation of CNT particles throughout the polymer matrix [6]. Additionally, the DC electrical conductivity (σ DC ) of the samples can be extracted by the following relationship where ω = 2πf, f is frequency, A is a constant, and s is a power-law exponent (s = 1 for polymers). Thus, σ DC value can be obtained by extrapolating the σ AC to low frequencies ( Figure 4b). The insulating nature of samples below the percolation threshold of the MWCNTs (≤2 phr) is clearly observed with DC conductivities below 10 −12 S/cm. As the percolation threshold is passed, the DC electrical conductivity increases several orders of magnitude to 10 −5 S/cm, 4 × 10 −4 S/cm, and 9 × 10 −3 for the foamed samples filled with 4 phr, 10 phr MWCNTs and 20 phr BT/10 phr MWCNTs, respectively, which are suitable for EM wave absorbing applications [31,32]. Therefore, the DC conductivity of the hybrid sample containing 10 phr of MWCNTs is one order of magnitude higher than the sample without BT. Furthermore, the sample based on B20C6 hybrid system shows similar AC values to that of the sample composed of only 10 phr MWCNTs (B0C10). These results imply the synergistic effect of BT and MWCNTs upon the conductivity of the foams. This could be ascribed to a dual excluded volume effect, originated from the formation of foam microstructure as well as the presence of the micro-size BT particles. In other words, the evolution of the cellular structure and the presence of the BT particles restricts the available volume for the MWCNTs and segregates them towards the cell walls and edges [5,33]. The segregation of conductive fillers during the foaming process has previously been observed as a factor to reduce the percolation threshold compared to their counterpart solid samples [5,16,29,34], and has also been recently reported for polypropylene (PP) foams containing hybrids of carbon black (CB) and MWCNTs [35]. Moreover, considering the excluded volume theory [36,37], the presence of a secondary particulate filler, such as BT, would further reduce the available volume for the CNT particles. Similar results have been reported for PP/BT/MWCNTs solid nanocomposites [38] and PP/MWCNTs/CB foam samples [35]. In these studies, this type of behavior has been ascribed to the improved electrical conductivity as a result of enhanced connectivity between the one-dimensional MWCNTs in conjunction with the zero-dimensional particles. This improved connectivity has been modeled based on a Voronoi geometry, resulting from Swiss cheese model, combined with the percolation theory [39]. Based on this model the improved conductivity for the hybrid filler systems depends on the secondary filler size and was favored with micro-size particulate fillers, such as the BT particles used in the present work.
Electrical conductivity is a vital index to evaluate the potential EMI absorption and shielding performance of conductive nanocomposites. The AC conductivity (AC) of a dielectric material presents the extent of its frequency-dependent behavior. Figure 4a displays the variation of AC conductivity as a function of frequency for the prepared foamed samples. As can be observed, above 2 phr MWCNTs, the foams exhibit frequency-independent behavior (insulation-conduction transition) implying the formation of interconnected conductive paths [30]. This behavior is commonly described in terms of the percolation of CNT particles throughout the polymer matrix [6]. Additionally, the DC electrical conductivity (DC) of the samples can be extracted by the following relationship where ω = 2πf, f is frequency, A is a constant, and s is a power-law exponent (s = 1 for polymers). Thus, DC value can be obtained by extrapolating the AC to low frequencies (Figure 4b). The insulating nature of samples below the percolation threshold of the MWCNTs (≤2 phr) is clearly observed with DC conductivities below 10 −12 S/cm. As the percolation threshold is passed, the DC electrical conductivity increases several orders of magnitude to 10 −5 S/cm, 4  10 −4 S/cm, and 9  10 −3 for the foamed samples filled with 4 phr, 10 phr MWCNTs and 20 phr BT / 10 phr MWCNTs, respectively, which are suitable for EM wave absorbing applications [31,32]. Therefore, the DC The material ability to save and dissipate EM energy is related to both its dielectric permittivity and magnetic permeability. Therefore, both parameters were measured to analyze the EMI shielding efficiency (SE) for our prepared foamed samples. The real and imaginary parts of the relative permeability were found to be 1 and 0, respectively, within the X-band frequencies, due to the absence of magnetic filler. Figure 5 demonstrates the variation of both real (ε r ) and imaginary (ε r ) components of the dielectric permittivity over the X-band. In our recently published paper [29], the values of ε r and ε r for the unfilled EPDM foam showed to be 1.6 and 0.001, respectively, which were slightly lower than the B20C0 foam (ε r = 1.9 and ε r = 0.0045). Meanwhile, the incorporation of MWCNTs into the composition of the EPDM/BT foams led to an increase in ε r , ε r , and dielectric loss factor (tan delta). Such increases are indicative of an enhanced degree of polarization, which is ascribed to two simultaneous polarization processes: a rotational polarization, from the dipoles attached to the surface of BT and MWCNTs, and interfacial space charge polarization, from the interface between the two fillers. The increase in ε r is indicative of the higher extent of polarization of the samples due to the presence of both ferroelectric BT particles and MWCNTs. In other words, both MWCNTs and BT particles can behave as carriers of polarizable dipoles within the polymer matrix, whereas the increase in the ε r would be the result of both polarization [15], and conductance loss [40]. Therefore, the increase in permittivity components (ε r and ε r ) is also reflected in the loss tangent (tan δ = ε r /ε r ) (Figure 5e), which is a reasonable criterion of the intrinsic ability of a material to dissipate EM energy. Figure 5b,d,f represent the variation of the dielectric parameters (ε r , ε r , and tan δ), with MWCNTs content at 10 GHz, respectively. As can be observed, the nanocomposite originated from the BT/MWCNTs hybrid exhibits higher values for e , e , and tan delta than counterpart samples composed of only MWCNTs, which evidences the synergistic effect of these two fillers on dielectric properties of the foamed EPDM composites.  , d, and f).
The EMI shielding effectiveness is the result of three mechanisms, including reflection of the incident wave from the surface of the shield as a result of the impedance mismatch between the air and the material, and absorption in which the energy of EM wave is dissipated and transformed into thermal energy while passing through the shield. The third is the multiple reflections of the waves at various surfaces or interfaces within the shield, i.e., air cells and nanoparticles. Therefore, the total SE (SET) can be expressed as follows [41,42]: The EMI shielding effectiveness is the result of three mechanisms, including reflection of the incident wave from the surface of the shield as a result of the impedance mismatch between the air and the material, and absorption in which the energy of EM wave is dissipated and transformed into thermal energy while passing through the shield. The third is the multiple reflections of the waves at Polymers 2020, 12, 2278 9 of 15 various surfaces or interfaces within the shield, i.e., air cells and nanoparticles. Therefore, the total SE (SET) can be expressed as follows [41,42]: where SE A , SE R , and SE M are the shielding effectivenesses by absorption, reflection, and multiple reflections, respectively, which are defined as [42]: where Pi, Pt, σ, f, µ, and δ are the incident and transmitted power, conductivity, frequency, permeability, and skin depth, respectively. Since the shield thickness of the developed foams is greater than skin depth (defined as the distance from the surface of the shield where the EM energy has dropped to 1/e of its initial value at the surface) [42], the effect of multiple reflections can be overlooked [43,44]. The three SE parameters were calculated ( Figure 6) for our tested samples with the thickness of 9 mm over the X-band frequencies [41]. The EPDM foamed sample filled with only 20 phr BT was transparent to EM waves and no EMI SE was measured for this sample, due to the low dielectric loss of BT. Meanwhile, the incorporation of MWCNTs in combination with BT into the foams compositions led to the improved shielding efficiencies via both absorption and reflection mechanisms. The presence of free electrons in the structure of MWCNTs results in closed induced electrons current inside the shield by the incident EM wave which enhances the depletion of the wave energy via absorption mechanism [45]. Nevertheless, the foamed EPDM composites composed of BT/MWCNTs exhibited higher values of SE A and SE R than the counterpart samples filled only with MWCNTS or BT. The hybrid EPDM nanocomposite foam loaded with 20 phr BT and 10 phr MWCNTs exhibited a total SE of~22.3 dB compared to~16 dB of the 10 phr MWCNTs at 10 GHz (Figure 7a). The improved EM wave absorption and efficiency imparted by BT/MWCNTs hybrid are mainly attributed to the enhanced total polarization as a result of contribution of the space charge polarization, in parallel to the orientational polarization of the dipoles attached to the surfaces of the BT and MWCNTs particles. Moreover, as was discussed, the composite based on 20 phr BT and 10 phr MWCNTs showed higher electrical conductivity than their counterpart (B0C10), which resulted in higher dissipation of the EM wave via conductance loss mechanism. As can be observed in Figure 7a, the SE A shows higher values than SE R over the studied range of frequencies, implying that absorption is the dominant mechanism in the shielding of the EM waves. Although most EMI shielding research is based on polymer matrices, some metal foams have also been studied as a strategy to solve one of the main drawbacks of metals, their weight. Aluminium metal foams, made from melt foaming method [46] and PU sacrifial templates [47], exhibited EMI shielding performance in the X-band of 75 dB and 45 dB, respectively. However, the SE of foams made from melt foaming was not constant throughout the entire frequency range and, hence, part of the EM was transmitted. Additionally, the dominant mechanism in these systems is reflection and they can not be used in applications that require high EM absorbing properties [47,48]. We normalized the total EMI SE values by the density, which is called specific EMI SE (Figure 7b). The specific SE T of hybrid foam containing 10 phr MWCNTs was 25% higher compared with the EPDM/MWCNTs counterpart, in spite of the high density of BT. As the EMI shielding efficiency of materials is governed by various parameters such as conductive particle filler concentration and intrinsic electrical conductivity, type of matrix, thickness, and fabrication process, it is therefore difficult to directly compare the results of different investigations without considering these factors. Hence, we calculated the extent of improvement reported on foamed nanocomposites filled with hybrid fillers as a way to establish a comparative analysis (Table 2). It can be observed that our synthesized EPDM foam filled with BT/MWCNTs (20/10 phr) shows the highest percentage of improvement for SE T (40%), compared to the other polymeric foams reported by others [7,13,34,44,49,50].      For long-life applications of an EM wave shield, high stability of the performance, especially when subjected to repeated bending deformations, is required. Hence, the EMI SE of the prepared EPDM foam composed of BT/MWCNTs (20/10 phr) was measured after 1000 times repeated bending (with a radius of 2.0 mm) (Figure 7c). Interestingly, the EMI SE showed an insignificant decrease over the entire X-band frequency range. The small reduction could be attributed to the high elastic shape recovery resulting from the entropic features of cross-linked EPDM networks assisted by the air molecules trapped within the cells. Only one study has been found reporting a decrease in SET after repeated bending for polyurethane/reduced graphene oxide nanocomposite foams at 10 GHz [50].
The thermal conductivity of shielding materials is also regarded as an important factor for EMI shielding to minimize the temperature impact on electronic devices. The thermal conductivity of the hybrid EPDM foam showed higher value [0.217 W/(m K)] compared to that measured [0.077 W/(m K)] for the sample composed of only 20 phr BT (Figure 8). This improvement is likely the result of improved thermal conduction through the solid and gas phases of the foam. Both convection and radiation mechanisms can be considered negligible, since the cell size is below 4 mm [51], and the relative density is above 0.2 [52], as reported elsewhere [29]. It is worth mentioning that the thermal  For long-life applications of an EM wave shield, high stability of the performance, especially when subjected to repeated bending deformations, is required. Hence, the EMI SE of the prepared EPDM foam composed of BT/MWCNTs (20/10 phr) was measured after 1000 times repeated bending (with a radius of 2.0 mm) (Figure 7c). Interestingly, the EMI SE showed an insignificant decrease over the entire X-band frequency range. The small reduction could be attributed to the high elastic shape recovery resulting from the entropic features of cross-linked EPDM networks assisted by the air molecules trapped within the cells. Only one study has been found reporting a decrease in SE T after repeated bending for polyurethane/reduced graphene oxide nanocomposite foams at 10 GHz [50].
The thermal conductivity of shielding materials is also regarded as an important factor for EMI shielding to minimize the temperature impact on electronic devices. The thermal conductivity of the hybrid EPDM foam showed higher value [0.217 W/(m K)] compared to that measured [0.077 W/(m K)] for the sample composed of only 20 phr BT (Figure 8). This improvement is likely the result of improved thermal conduction through the solid and gas phases of the foam. Both convection and radiation mechanisms can be considered negligible, since the cell size is below 4 mm [51], and the relative density is above 0.2 [52], as reported elsewhere [29]. It is worth mentioning that the thermal conductivity of the MWCNTs foams is similar to their BT/MWCNTs counterparts due to the low volumetric percentage of BT (as the maximum BT volume percentage is about 2 vol.%). Previous studies have shown that a large loading fractions (above 20 vol.%) of BT is needed to obtain a significant increase in the thermal conductivity of polymer composites [53,54].
conductivity of the MWCNTs foams is similar to their BT/MWCNTs counterparts due to the low volumetric percentage of BT (as the maximum BT volume percentage is about 2 vol.%). Previous studies have shown that a large loading fractions (above 20 vol.%) of BT is needed to obtain a significant increase in the thermal conductivity of polymer composites [53,54].

Conclusion
In the present study, EMI shielding and absorption have synergistically been improved through the use of a hybrid of conductive and ferroelectric ceramic fillers in the composition of crosslinked EPDM foam. Consequently, while the EPDM foam filled by only 20 phr BT was shown to be transparent to the EM wave within the X-band frequency, the hybrid EPDM nanocomposite foam loaded with 20 phr BT and 10 phr MWCNTs presented a total shielding effectiveness (SE) of ~22.3 dB compared to ~16.0 dB measured for EPDM foam filled with 10 phr MWCNTs. Additionally, the total SE of the BT/MWCNTs foam remained almost constant when subjected to repeated bending due to the high rubber elasticity as well as the elastic memory induced when the sample was deformed. Finally, the BT/MWCNTs-filled EPDM foam showed improved thermal conductivity compared to that of the foamed samples composed of only BT-filled foam. Since other properties, such as resistance to chemicals, moisture, and ozone, are inherent for EPDM rubber, these foams can be regarded as a high potential material for the absorption of high-frequency EM wave in various applications, such as aeronautical, civil, and military.

Conclusion
In the present study, EMI shielding and absorption have synergistically been improved through the use of a hybrid of conductive and ferroelectric ceramic fillers in the composition of crosslinked EPDM foam. Consequently, while the EPDM foam filled by only 20 phr BT was shown to be transparent to the EM wave within the X-band frequency, the hybrid EPDM nanocomposite foam loaded with 20 phr BT and 10 phr MWCNTs presented a total shielding effectiveness (SE) of~22.3 dB compared to~16.0 dB measured for EPDM foam filled with 10 phr MWCNTs. Additionally, the total SE of the BT/MWCNTs foam remained almost constant when subjected to repeated bending due to the high rubber elasticity as well as the elastic memory induced when the sample was deformed. Finally, the BT/MWCNTs-filled EPDM foam showed improved thermal conductivity compared to that of the foamed samples composed of only BT-filled foam. Since other properties, such as resistance to chemicals, moisture, and ozone, are inherent for EPDM rubber, these foams can be regarded as a high potential material for the absorption of high-frequency EM wave in various applications, such as aeronautical, civil, and military. Funding: Financial support from MICIU is acknowledged through the project MAT2016-81138-R.