Electromagnetic interference shielding effectiveness of MWCNT filled poly(ether sulfone) and poly(ether imide) nanocomposites.
The electromagnetic wave interference (EMI) is an undesired common phenomenon that has come with use of electromagnetic waves in equipment and devices employed in sectors like commerce, health care, defence and aerospace [1-6]. The effect of EMI can range from degradation to interception and obstruction of the performance of electronic or electrical equipment. The disturbances across communication channels, automation and process control may lead to loss of valuable time, energy, resources, money or even precious human life. Besides, some diseases affect human health such as leukemia, miscarriages, breast cancer on continuous exposure to EM fields and pulses , Therefore, in order to alleviate these troubles, the development of EMI shielding materials for microwave and millimeter waves is receiving increasing attention.
EMI shielding refers to the reflection and/or adsorption of electromagnetic radiation by a material, which thereby acts as a shield against the penetration of the radiation through it. The shielding effectiveness (SE) is a number that quantifies the amount of attenuation typical of a particular material. If [P.sub.inc] is the incident power density at a measuring point before shield is in place, [P.sub.trans] is the transmitted power density at the same measuring point after shield is in place and [P.sub.ref] is the reflected power density at same measuring point, then
[SE.sub.total] = -10 log ([P.sub.trans]/[P.sub.inc]) = -10 log T (1)
return loss, RL = -10 log ([P.sub.ref]/[P.sub.inc]) = -10 log R (2)
where R and T represent reflectivity and transmittivity. It is evident that for a lossless material,
[P.sub.inc] = [P.sub.ref] + [P.sub.trans] (3)
A can be quantified as the absorbed power inside the dielectric, and
A = 1 - ([P.sub.ref])/([P.sub.inc]) - ([P.sub.trans])/([P.sub.inc]) = 1 - R - T (4)
where A is the power absorbed of a lossy dielectric . However, EMI is the consequence of reflection loss, transmission or absorption loss and internal reflection loss at exiting interfaces of the incident electromagnetic waves in the sample. These three losses are inter-related [2, 8] by
[SE.sub.total] = (R + A + B) (5)
where [SE.sub.total] is the insertion loss that represents the reduction (expressed in dB) of the level of an electromagnetic field at a point in space after a conductive barrier is inserted between that point and the source. R is the sum of initial reflection losses from both surfaces of the shield exclusive of additional reflection losses, A is the absorption or penetration loss within the barrier itself and B is the internal reflection loss at the exiting interfaces. This term may be either positive or negative and is negligible for A [greater than or equal to] 15 dB.
Depending on the shielding mechanism, a shielding material should have at least one the following properties such as moderate conductivity and/or electric or magnetic permeability , Since a polymer matrix is usually electrically insulating and does not contribute directly to the resultant shielding effect, different conductive fillers such as metal particles [9, 10], carbon fiber , carbon black , graphite , and carbon nanotubes (CNTs) [14-20] are commonly used to prepare conductive polymer composite. Compared with conventional metal based EMI shielding materials, electrically conductive polymer composites have gained popularity recently because of their light weight, resistance to corrosion, flexibility, and processing advantages [2-5], For any filler, the EMI shielding effectiveness increases with increasing filler concentration in the composites, but the maximum filler loading is limited by the reduced mechanical properties of most composites at higher filler loadings resulting from poor filler-matrix bonding . For material and process cost saving and good mechanical properties, the attainment of a high shielding effectiveness at a low filler loading is desirable. The EMI shielding efficiency (SE) of a composite material depends on many factors, including the filler's intrinsic conductivity, dielectric constant, and aspect ratio [2, 6], Thus, in the field of conductive fillers, the overall properties of CNTs such as high conductivity, small diameter, high aspect ratio, along with high thermo-oxidative stability and mechanical strength make them an excellent option to create a continuous conductive network in composites for high-performance EMI shielding materials at low filler concentration (at low percolation threshold). In recent years, extensive research in CNT composites has been undertaken with different polymer matrices such as PC , PANI [21, 22], PPY , PS [23, 24], PMMA [25, 26], epoxy [8, 27], PU [28-30], LCP  for their end-use application as effective light weight EMI shielding material. In most cases, the main contribution to overall EMI shielding is due to reflection [8, 23, 24, 28], It has been also found that when catalyst particle like Fe are encapsulated in CNTs (before or after the growing stage), the main contribution to overall EMI shielding is absorption rather than reflection [25, 32]. This is due to the presence of magnetic dipoles in the Fe particles which interact with the electromagnetic field of radiation. Also, the influence of the aspect ratio [8, 31], wall defects , and alignment of CNTs  in the composite has been studied. As stated above, though, the polymer matrix is insulating and does not directly contribute to total EMI shielding, it can influence the connectivity among the conducting fillers and thereby, indirectly enhance the EMI shielding effect. Dispersion and distribution of any particular filler always depends next to processing conditions on the type of polymer matrix having different properties like polymer surface tension, crystallinity, polarity and molecular weight . A lower filler-polymer interfacial tension leads to better wettability, dispersion, and distribution of the filler in the polymer matrix. On the other hand, higher filler-polymer interfacial tension results in more pronounced agglomeration of fillers. Similarly, the polarity of polymer and filler plays an important role in improving filler dispersion. For graphite filler, better conductivity and thus good EMI shielding has been reportedly achieved in an amorphous polymer (polyetherimide) compared to a semi crystalline polymer (polypropylene), due to better interactions between the graphite filler and polyetherimide . Fillers are ejected out from the crystalline regions during crystallization and accumulates in the amorphous regions giving rise to an uneven distribution at the nanomicroscale, despite good or bad dispersion. In addition, the nature of the polymer matrix decides about the way of composite processing and the applicable processing conditions.
Polyimides (PI), poly(ether imides) (PEI), and poly (ether sulfone) (PES) are known to be high performance polymers based on their outstanding set of properties . In this article, PEI and PES are chosen as polymer matrices because of their good thermo-oxidative stability, good chemical resistance, high mechanical strength, and most importantly very good processability for the preparation of polymer/MWCNT composites for EMI shielding application. Though, extensive articles have been published for PI/MWCNT composites [14-19], to the best of our knowledge, their applicability as EMI-shielding material is rarely investigated, except a recent patent  which has reported a shielding effectiveness of 40 to 45 dB at 1 to 3 GHz. While the claimed shielding effectiveness is high enough from application point of view, the very high CNT loading (30 wt%) in the composite may have adverse effect on the mechanical properties of the composite which is not properly highlighted in the invention. Similarly, metal particles based PES composites [36, 37] have been studied, but no studies are found on PES/ MWCNT for EMI shielding application. So, in the present investigation the influence of different weight percentages of MWCNTs on the EMI shielding has been studied in two different polymer matrices (PEI and PES) for a frequency range (8-12 GHz) in the X-band region. The electrical, thermomechanical and the morphological properties of the polymer composites prepared by solution mixing technique have also been investigated.
Bischloro monomer 4,4'-dichlorodiphenyl sulfone (DCDPS) and 1,3-phenylenediamine (mPD) (99%) were purchased from Aldrich (USA) and were used as received; 4,4'-Isopropylidenediphenoxy) bis(phthalic anhydride) (BPADA) (97%) was purchased from Aldrich, USA and oven dried overnight at 120[degrees]C before use. Bisphenol namely 4,4'-isopropylidenediphenol (BPA) was purchased from Loba Chemical Company (India) and was recrystallized from toluene. [K.sub.2]C[O.sub.3] and dichloromethane (DCM) (E. Merck, India) were used as received. A-methyl-2-pyrrolidone (NMP) (E. Merck, India) was purified from NaOH and distilled from [P.sub.2][O.sub.5] before use. Toluene (Merck, India) was refluxed over Na metal to remove water and freshly distilled before use. 1,2-Dichlorobenzene and methanol were purchased from (E. Merck, India) and used as received. Multiwalled carbon nanotubes (NC 7000) were purchased from Nanocyl S.A., Sambreville, Belgium. According to the provided data sheet from the manufacturer, the average diameter, average length, carbon purity, metal oxide impurity, and surface area are 9.5 nm, 1.5 [micro]m, 90 wt%, 10 wt%, and 250 to 300 [m.sup.2]/g, respectively.
Preparation of Poly(ether imide) (PEI). The polymerization reaction was conducted by the reaction of equimolar amounts of BPADA as dianhydride and mPD as diamine. The reaction was conducted under constant flow of nitrogen. A standard protocol of solution polymerization technique was adopted as reported elsewhere , In a 500 mL, three necked round-bottomed flask equipped with nitrogen inlet, a magnetic stirrer and Dean-Stark trap fitted with a condenser was charged with 41.34 g (79.537 mmol) of BPADA and 8.60 g (79.537 mmol) of mPD and 180 mL of 1,2-dichorobenzene. The reaction temperature was raised slowly from room temperature to 180[degrees]C. During the reaction, the solution was observed to turn viscous. The reaction was continued for 6 h at 180[degrees]C under nitrogen. The resulting viscous polymeric solution was then cooled to room temperature and was precipitated in large excess of methanol. The fibrous product obtained was dried and dissolved in 500 mL dichloromethane and re-precipitated from methanol. The [T.sub.g] of the polyimide recorded by DSC was 215[degrees]C. The polymer did not exhibit crystallization or melt transition in the DSC measurement. GPC results indicated the formation of a high molar mass product ([M.sub.w] ~ 75,000 g/mol) with a polydispersity of 2.3. The structure of the repeating unit of the synthesized PEI is presented in Fig. 1a.
Preparation of Poly(ether sulfone) (PES). The polymerization reaction was conducted by the reaction of equimolar amount of DCDPS as bischloro monomer and BPA as bisphenol. The reaction was conducted under constant flow of nitrogen. Polymerization reactions were carried out in a 2 L, three-necked round-bottomed flask equipped with a nitrogen inlet, a stir bar and a Dean-Stark trap fitted with condenser. The reactions were conducted under constant flow of nitrogen. A representative poly merization procedure is as follows. The flask was charged with equimolar amounts of 4,4'-dichlorodiphenyl sulfone (27.86 g, 97.025 mmol), BPA (22.15 g, 97.025 mmol), [K.sub.2]C[O.sub.3] (32.18 g, 232.86 mmol), NMP (375 mL), and toluene (1000 mL). The mixture was then heated to reflux (140-150[degrees]C, oil bath temperature) for 2 to 3 h to remove the water azeotropically with toluene. After removal of the toluene from the Dean-Stark trap, the reaction temperature was increased to 180[degrees]C and maintained for another 6 h. After cooling to room temperature, the polymer was precipitated from large excess of methanol containing 125 mL of HC1. Fibrous solids were isolated. These products were washed several times in boiling water to remove any inorganic impurities and were dried in a vacuum at 120[degrees]C for overnight. The [T.sub.g] of the polyethersulfone recorded by DSC was 196[degrees]C. The polymer did not exhibit crystallization or melt transition in the DSC measurement. GPC results indicated the formation of high molar mass product ([M.sub.w] ~ 135,000 g/mol) with a polydispersity of 2.4. The structure of the repeating unit of the synthesized PES is presented in Fig. 1b.
Preparation of PEI/MWCNT and PES/MWCNT Nanocomposite Films. A particular amount of the dried polymer sample (PEI or PES) (about 2 g) was dissolved in DCM solvent (around 15-20 mL) in a beaker. Different weight percentages (0.5, 0.75, 1.0, 1.5, 2.0, 3.0, 4.0 and 5.0 wt%) of MWCNTs were taken separately in ~15 mL DCM in another beaker and subjected to ultrasonication (Ultrasonic processor Sonapros PR-250, Horn-type) for 40 min. Then, the polymer solution was added to the CNT dispersion and the ultrasonication was continued for another 10 min. The entire ultrasonication process was done with 1 Amp current, 5 watt power and frequency 20 [+ or -] 3 KHz. The entire MWCNT polymer uniform dispersion was then poured on flat bottomed Petri dishes and the solvent was allowed to evaporate at a controlled rate at 30[degrees]C overnight to obtain composite films. The Petri dish was then kept in a vacuum oven and the temperature of the oven was slowly raised to 120[degrees]C and kept under continuous vacuum for 5 to 6 h to remove any trace of solvent. The obtained samples had a thickness in the range of ~0.155 mm. The nomenclature and preparation technique of composite specimens along with their different experimental measurements are shown in Table 1.
Fabrication of Compression Molded Specimens. Three specimens were prepared for each of carbon nanotube composites having different weight percentages such as 3, 4, 5 wt% for both PEI and PES matrices. For the preparation of composites, the PEI/MWCNT and PES/MWCNT nanocomposite fdms were cut into small pieces and were filled in the X-band waveguide flanges for compression molding under a pressure of about 20 MPa at 320[degrees]C (for PEI/MWCNT) or at 260[degrees]C (for PES/ MWCNT) respectively for 10 min. The resulting compression molded specimens were allowed to cool to room temperature under the same pressure at the rate of 2 K/min. Samples (thickness ~5 mm) with compact and smooth surface were obtained in all the cases.
Electron Microscopy. Scanning electron microscopic (SEM) analysis was done using a JEOL JSM-5800 SEM on cryo-fractured gold plated surfaces of the compression molded specimens in order to observe dispersion of CNT in the composites. Observation was done under vacuum ([10.sup.-4] to [10.sup.-6] mm Hg).
The dispersion of the MWCNT in the compression molded specimens was further studied by a transmission electron microscope (TEM, JEM-2100, JEOL, Japan) operating at an accelerating voltage of 200 kV. The compression molded specimens were ultramicrotomed under cryogenic condition with a thickness of 100 nm. Since the MWCNTs have much higher electron density than the polymers, they appeared dark in the TEM images.
Dynamic Mechanical Analysis. Dynamic mechanical analysis (DMA) was performed with an Eplexor 2000N (Gabo Qualimeter, Ahlden, Germany) using a constant frequency of 10 Hz in a temperature range from 20 to 230[degrees]C under nitrogen atmosphere. The heating rate was 2 K/min. The measurements were done in tension mode with solution cast membranes of thickness ~0.155 mm. For the measurement of complex modulus, G' a static load at 1% pre-strain was applied and then the sample was oscillated to a dynamic load at 0.5% strain. The glass transition temperature, [T.sub.g] was determined from the maximum in the tan delta versus temperature plots.
AC Electrical Conductivity. The dielectric properties were measured with a broadband dielectric spectrometer BDS 40 system (Novocontrol GmbH, Germany) on solution cast membranes with a thickness of ~0.155 mm and a diameter of 20 mm. To ensure good contact, gold plated brass electrodes were used. The samples were sputtered with a thin layer of gold on both sides. The frequency range was 0.1 Hz to 10 MHz and the real part of the conductivity was directly obtained from the instrument.
EMI Shielding Measurement. The EMI SE was measured with a test chamber set-up consisting of a sweep oscillator (model HP 8350B, Hewlett Packard), a power splitter, a detector and a scalar network analyzer (model HP 8757 C/E, Hewlett Packard) (Fig. 2). The SE was measured using the co-axial cable line method. The SE in the frequency range 8 to 12 GHz was measured using X-band wave-guide as sample holder. Compression molded specimens as prepared inside the waveguide flanges having a thickness of ~5.0 mm were used in the measurement. EMI shielding measurement was carried out for each sample by continuously sweeping in the frequency range of 8 to 12 GHz. When SE was found to be higher than 40 dB, some noise was encountered during the measurement; otherwise, the result showed smooth curve of SE against frequency.
RESULT AND DISCUSSION
Morphology of the Composites
In order to characterize the state of dispersion of the MWCNTs within the compression molded samples, which influences the EMI behavior, SEM investigations of cryofractured surfaces of compression molded samples with 5 wt% filler loading were performed, as shown in Fig. 3. The bright spots and lines in the images are attributed to the MWCNTs. At 5 wt% of MWCNT, the images give the impression of a percolative morphology due to overlapping of the tubes. Though it is difficult to comment on the dispersion of MWCNT in the PEI matrix from these images with high percentage of loading, clearly a good dispersion and random distribution of the tubes could be found (Fig. 3c) for the PES matrix. Compared to our previous publication  where melt mixing technique was used, solution mixing technique provided better dispersion of CNTs in PES matrix. It seems as a whole that for both composites, strong van der Waals forces between the polymer and the nanotubes exceed the existing inter-tube forces within the primary MWCNT agglomerates, thus allowing a good dispersion during sonication and preventing tube reagglomeration.
The bulk morphology of the composites was further investigated by transmission electron microscopy (TEM). Figure 4 shows TEM images of thin cut samples prepared from the compression molded PEI/MWCNT and PES/MWCNT composites with different MWCNT loadings (3, 4, and 5 wt%) at two different magnifications. TEM micrographs revealed that MWCNTs were randomly well dispersed in both PEI and PES matrix. At lower loading (Fig. 4a and d) the MWCNTs are dispersed as single tubes. As the nanotube loading increased, the randomly distributed carbon nanotubes came closer (in other words, decrease in path length for charge carriers between nanotubes occurred) and even in the thin cut samples representing only a two-dimensional view of the three-dimensional structure, a percolative network structure became quite evident (Fig. 4c and 0 .
Dynamic Mechanical Properties
DMA in temperature sweeps was used to determine the glass transition temperature ([T.sub.g]) of the polymers and their composites. As shown on the storage modulus (E') curves in Fig. 5a and b, an increase in [T.sub.g] of PEI (at 217[degrees]C) by about 10 to 15 K was observed after incorporation of MWCNT in the matrix, however, such an increase in [T.sub.g] was not observed in case of PES as matrix ([T.sub.g] = 203[degrees]C in all cases) after incorporation of MWCNT. Whereas, a rubbery plateau-like behavior in the temperature range above the [T.sub.g] was found in the storage modulus vs. temperature plot of PES (Fig. 5b). In case of PEI and its nanocomposites, the E' values fall dramatically above [T.sub.g] and the samples flow out of the clamps. From the tan [delta] versus temperature curves, surprisingly a prominent extra peak just below the [T.sub.g] of pure PES was observed for all composites comprised with MWCNT. Obviously, this extra peak hints at an own separate relaxation of the nanotubes themselves. Here, the nanotube used was MWCNT and the entangled nature of these nanotubes is well known in literature. The reason of this peak could be the co-operative relaxation process accompanying the transition from stiff (glassy like) semiflexible tube conformations to a more flexible additional network of the tubes. Figure 5b indicates a corresponding tube network plateau modulus of order 1 MPa. Notably, this (viscoelastic) behavior of tubes in polymer environment is different with respect to the behavior of networks consisting of pure carbon nanotubes. In the latter case, the nature of carbon nanotubes as thermal statistical objects exploring their conformational space like polymer chains still remained unclear [41, 42], Obviously, single-wall carbon nanotubes display entropic behavior akin to a polymer network, whereas multi-wall carbon nanotubes displays granular-like viscoelasticity, i.e. the mechanical viscoelastic response is independent of temperature. Another explanation of the extra tan [delta] peak, may be due to the existence of secondary glassy like phase of the polymer fragments being close to the tubes surface, which can be neglected due to the reason that the more tight or semi-flexible chains of the polymer should give rise to an additional peak at higher temperature, but not at lower temperature, than the corresponding bulk polymer [T.sub.g]. In the case of PEI, such an effect was not found. Most probably, the dispersion of the tubes was good enough (Fig. 4a) such that the tubes did not form any entangled network. Here, it should be noted that with the increase in the tube content in PES matrix the storage modulus at temperatures above [T.sub.g] increases and reflects a rubbery plateau which is not found in pure PES. This also suggests that during the DMA run (cyclic deformation with static load at 1% and dynamic load was at 0.5% strain), the pure PES polymer being un-crosslinked, can no longer sustain the stress. However, the entangled tube network behaved as a (physically) cross-linked polymer. The following argument will support the conclusions. We assume that the bonding between two neighboring tubes of a tube network junction occurs due to the van der Waals interaction across the diameter of a tube. Furthermore, we assume--similar as has been found for single wall carbon nanotube suspensions  that the tube network exhibits rigidity percolation in that the elasticity appears to be due to the bonding (rather than stretching or bending) of the semiflexible tubes. Hence, the tubes form an elastic network held together by freely jointed bonds of interaction energy [E.sub.int] ~ A [k.sub.B]T where A = 40 for SWCNT suspensions , In our case, the prefactor A may be different (obviously larger), however this is of less importance for our crude and qualitative discussion here. According to Fig. 5, the tube network plateau shear modulus is roughly estimated of order E'3 ~ G' ~ 0.1... 1 MPa in the regime of temperatures T ~ 500 K in which [E.sub.int] ~ 1.25 eV can be calculated with A = 40. If we assume that the elastically stored energy of the tubes network comes from the interaction energy per a characteristic mesh wide ([xi]) of the tube network, i.e. [E.sub.int] [[xi].sup.3] ~ G', we estimate the corresponding length scale as [xi]~6.13nm, i.e. of order 10 nm which is in agreement with Fig. 4 by simple eye inspection. Note, the estimated length scale [xi] is one order of magnitude smaller as the corresponding persistence length of carbon nanotubes.
To get a clear understanding about the dynamic mechanical behavior of entangled nanotubes a more detailed investigation with CNT composites with other high [T.sub.g] polymers like PES might be a future task of us.
AC Electrical Conductivity
As EMI shielding is strongly connected to the electrical conductivity of the materials, we also studied AC conductivity. For PEI and PES, the conductivity increased significantly with frequency (Fig. 6a and b). After incorporation of MWCNTs in the PEI matrix, the composites behaved like a conductor, independent of frequency and much higher conductivity values even at loadings as low as 0.5 wt%. Owing to the presence of MWCNT, the formed three-dimensional CNT network carry charge by hopping and/or tunneling mechanisms. With increase in MWCNT content, the conductivity of the composites increased and reached value of ~0.003 S/cm at 2 wt% MWCNT loading. At this loading, the PES based composites showed a conductivity of 0.03 S/cm which is 10 times higher than that of the PEI composite. All the compositions of PES based composites showed higher values than those based on PEI. There might be three main reasons for the higher conductivity of the PES/2 wt% MWCNT composite compared to PEI/2 wt% MWCNT composite. Firstly, there might be better dispersion in the PES based composites, as it was indicated in the microscopic study. Secondly, the wrapping of polymer around the MWCNTs resulting in a thin insulting layer has to be overcome by electron hopping/tunneling. Most probably the MWCNTs were better wrapped by PEI polymer chains and the direct contacts between the tubes were reduced resulting in lower conductivity of PEI composites compared to the PES composites. As a third reason, the lower viscosity of the PES-MWCNT solution (during the solution processing) might have induced a lower CNT shortening during the sample preparation including sonication. Sonication is known to shorten CNTs and the CNT aspect ratio is indirectly proportional to the percolation threshold and also influences the achievable conductivity. The percolation threshold for both the composites was found to be lower than 0.5 wt% and cannot be determined exactly as lower loadings were not investigated in this study. At low MWCNT loading, still a slight frequency dependency of the conductivity at higher frequency was observed.
EMI Shielding Effectiveness
The results of the measurements on EMI shielding effectiveness over the frequency range of 8 to 12 GHz for PEI and PES and their composites with various MWCNT loadings are shown in Fig. 7a and b. In general, a huge difference in SE was observed between the pure polymer and their nanocomposites and the SE of nanocomposites increased with MWCNT loading at all frequencies of the incident radiation for both PEI and PES. All composites exhibited a nearly frequency independent EMI SE performance. An appreciable increase in SE was found when the frequency was changed from the lower to higher frequency regime; for example, the nanocomposites of PES at 3 wt% of MWCNT exhibited a shielding level at about 28 dB over 8-12 GHz, while for the nanocomposites with 4% of MWCNT, the shielding values measured were between 35 and 43 dB. This increment of the EMI shielding effectiveness is ascribed mainly due to increase in the formation of conductive interconnected nanotube networks in the polymer matrix with increase in MWCNT loading. The conductive network formed due to the dispersion of MWCNT behaves like a conductive mesh which intercepts electromagnetic radiation . The size of the mesh generally determines the effectiveness of shielding. A smaller mesh size generally increases the shielding effectiveness. However, the size of the mesh is also related to frequency of the incident radiation which will be intercepted. The SE depends on the thickness of the material, its electrical characteristics and the nature of the incident radiation . The openings in the conductive network due to the orientation of the conducting filler in the composite affect the coupling of the electromagnetic wave. If the openings are electrically small then the coupling is mainly due to the tangential magnetic field and the electric field component normal to the specimen plane. The highest shielding effect we observed was for the composite with 5 wt% MWCNT loading, that is, SE = 46.8 dB for the PEI nanocomposite and SE = 46.5 dB for the PES nanocomposite at 12 GHz. It is interesting to note that at 5 wt% loading, almost similar shielding effectiveness has been achieved for both PEI/ MWCNT and PES/MWCNT composite despite the lower conductivity values achieved for the PEI/MWCNT composites at 2 wt% loading. This fact is quite obviously related to the attainment of similar order of conductivity for the high MWCNT loading (5 wt%) which is ultimately related to microstructural features and network structures of the MWCNT in the two different matrices.
According to EMI theory, filler with a higher aspect ratio and intrinsic conductivity should have a better EMI shielding performance in the same polymer matrix. This fact is obviously proved when we compare the EMI shielding of our previously reported PEI/graphite composite (SE = 40 dB for 40 wt% loading of graphite and beyond) with that of the PEI/MWCNT composites (SE = 46.8 dB for 5 wt% loading of MWCNTs). These results demonstrate that the aspect ratio of raw carbon fillers (~ 150 for MWCNT; 1 for graphite powder) is essential for electrical conductivity and shielding effectiveness of composite materials, since the high aspect ratio of raw fillers in composites assists the construction of conductive network. Besides, the intrinsic electrical conductivity of the carbon filler is also an important factor which is quite low for graphite compared with MWCNT. Since the conductive direction of graphite is in planes not on c-axis, it is difficult to make a three-dimensional conductive network.
Figure 8 shows the variation of SE with MWCNT loading at fixed frequencies of 8 and 12 GHz, respectively. The shielding effectiveness increased linearly with the MWCNT loading. With increasing frequency from 8 GHz to 12 GHz, the shielding effectiveness also increased. The SE values observed for 5 wt% of MWCNT loading in PEI and PES was in the range of 42.6 to 46.8 dB over the frequency range of 8 to 12 GHz. This shielding value of 5 wt% MWCNT is higher than the reported values, for examples 20 wt% SWCNT in polyurethane  (16-17 dB), 15 wt% SWCNT in epoxy  (23-28 dB), and 7 wt% MWCNT in PS foam 3 18-19 dB). Although this comparison is made without taking into account the thickness of the specimens which also plays an important role in the shielding effectiveness. The target value of the EMI shielding effectiveness for commercial applications is around 20 dB. Since composites from PEI and PES (3-5 wt% MWCNT) exhibit SE value higher than 20 dB according to the Fig. 8, these composite materials can meet easily the EMI commercial application requirement.
The variation of return loss (RL) over the frequency range of 8-12 GHz for different wt% of MWCNT loading is shown in Fig. 9. For low filler loadings such as 3 wt%, a reverse trend was observed (Fig. 7), representing that composites having high SE values showed low RL values. For example, the PEI based nanocomposite with 4 wt% MWCNT loading showed an SE of 40.91 dB whereas, the RL of this composite was 0.26 dB at 12 GHz. However, it is surprising to note that composites from PEI and PES at 5 wt% loading did not show the lowest return loss. This might be due to formation of a number of finer conductive meshes in the composite morphology which resulted in greater reflectivity of electromagnetic radiation. Also, the result of the RL measurements shows that RL is frequency dependent. The uneven variation of RL against frequency is due to the random distribution of the conducting MWCNT inside the specimen leading to the formation of voids of different sizes in the conducting mesh. These voids also affect RL due to their effect on the external reflection. In addition, RL and absorption loss also contribute to the total SE.
The change in absorption loss (AL) for PEI and PES and their nanocomposites with different MWCNT loadings over the frequency 8 to 12 GHz is shown in Fig. 10. When SE> 15 dB the contribution due to multiple internal reflections was in general negligible. In that case, Eq. 5 can be assumed as [SE.sub.total] = R + A. Then transmittivity, T, reflectivity, R, and absorptivity, A can be calculated using Eqs. 1, 2, and 4, respectively. It is seen that at higher frequency (12 GHz), comparing 3 and 4 wt% MWCNT in PEI or PES, both A and R increases with the loading. Interestingly, when the loading was higher (5 wt%), an increase in A and decrease in R was observed. For example, for 3 wt% MWCNT loading, the R value was as high as 93% and A was 6%. When the loading was increased to 4 wt%, the R and A values changed slightly to 94% and 5%, respectively. When the loading was further increased to 5 wt%, a shift in contribution behavior was noticed: the R value decreased from 94% to 83% and the A value increased to 17% which was almost three times to what observed at 3 wt% loading. Similarly, a general trend was found for PES/MWCNT composite with the increase in MWCNT loading at the same frequency i.e., R decreased and A increased at the same frequency. For example, at 12 GHz, the reflectivity decreased from 89.0% to 85.8% and then to 79.7% as the MWCNT loading increased from 3 wt% to 4 wt% and then to 5 wt%. So, overall, for the entire PE1/MWCNT and PES/MWCNT composites, the major contribution to total SE comes from reflection rather than from absorption. This is consistent with the EMI SE findings reported earlier for polyurethane/SWCNT , epoxy/SWCNT , and PS foam/MWCNT  composites. However, this trend is in contrast to the previous reports where Fe hybridized MWCNTs were used as conducting filler for PMMA  and epoxy , the contribution from absorption being much larger than that from reflectance.
In this study, two different sets of composites from polyetherimide (PEI) and polyethersulfone (PES) with multiwalled carbon nanotubes (MWCNTs) were prepared with variation in the weight fraction of MWCNT. As expected, the electrical conductivity and EMI SE of PEI/MWCNT and PES/MWCNT nanocomposites at room temperature increased with increasing weight fraction of MWCNT. The MWCNT composites of PEI and PES showed very high electrical conductivity already at very low loading of MWCNTs. Percolation of MWCNT in the composites was achieved below 0.5 wt%. Furthermore, small differences in the morphological microstructure features and network structures were detected which might influence the conductivity and EMI SE. The EMI SE of both PEI/MWCNT and PES/MWCNT nanocomposites were found to be independent of any major fluctuation due to frequency in the X-band region. The nanocomposites with 5 wt% MWCNTs exhibited EMI SE between 42 and 45 dB at 8 GHz.
For all the nanocomposites, the main contribution to total SE comes from reflection. However, particularly at high frequency, there is a gradual shift from the factors contributing to EMI from reflection to absorption with increasing MWCNT loading which is similar to the findings in the literature , This can be attributed to an increase in dielectric loss values with increasing MWCNT fraction. Because of high EMI SE (in the range 29-47 dB) along with excellent physical properties of PEI and PES, these composites are promising light weight shielding materials for use in both civil and military applications.
The authors are thankful to German Institute of Rubber Technology (DIK, Hannover) for the AC conductivity measurements of the samples.
[1.] H.W. Markstein, Electron Packag. Prod., 35, 4 (1995).
[2.] D.D.L. Chung, Carbon, 39, 279 (2001).
[3.] N.C. Das, T.K. Chaki, D. Khastgir, and A. Chakraborty, J. Appl. Polym. Sci., 80, 1601 (2001).
[4.] K.-S. Chou, K.-C. Huang, and Z.-H. Shih, J. Appl. Polym. Sci., 97, 128 (2005).
[5.] S. Yang, K. Lozano, A. Lomeli, H. D. Foltz, and R. Iones, Compos. A, 36, 691 (2005).
[6.] I. Joo, and C.Y. Lee, J. Appl. Phys., 88, 513 (2000).
[7.] B. Hornbostel, U. Leute, P. Potschke, I. Kotz, D. Kornfeld, P.W. Chiu, and S. Roth, Phys. E 40, 2425 (2008).
[8.] N. Li, Y. Huang, F. Du, X. He, X. Lin, H. Gao, Y. Ma, F. Li, Y. Chen, and P.C. Eklund, Nano Lett., 6, 1141 (2006).
[9.] L. Li, and D.D.L. Chung, Composites 25, 215 (1994).
[10.] X.P. Shui, and D.D.L. Chung, J. Mater. Sci., 35, 1773 (2000).
[11.] N.C. Das, T.K. Chaki, and D. Khastgir, Adv. Polym. Technol., 20, 226 (2001).
[12.] Q.I. Krueger, and I.A. King, Adv. Polym. Technol., 22, 96 (2003).
[13.] P. Sawai, and S. Banerjee, J. Appl. Polym. Sci., 109, 2054 (2008).
[14.] E.J. Siochi, D.C. Working, C. Park, P.T. Lillehei, J.H. Rouse, C. C. Topping, A.R. Bhattacharyya, and S. Kumar, Compos. B, 35, 439 (2004).
[15.] B.K. Zhu, S.H. Xie, Z.K. Xu, and Y.Y. Xu, Compos. Sci. Techno!., 66, 548 (2006).
[16.] H. H. So, J.W. Cho, and N.G. Sahoo, Eur. Polym. J. 43, 3750 (2007).
[17.] S.M. Yuen, C.C.M. Ma, Y.Y. Lin, and H.C. Kuan, Compos. Sci. Teclmol., 67, 2564 (2007).
[18.] S. Kumar, B. Li, S. Caceres, R.G. Maguire, and W.H. Zhong, Nanotechnology, 20, 465708 (2009).
[19.] R. Srivastava, S. Banerjee, D. Jehnichen, B. Voit, and F. Bohme, Macromol. Mater. Eng., 294, 96 (2009).
[20.] B.K. Singh, P. Kar, N.K. Shrivastava, S. Banerjee, and B.B. Khatua, J. Appl. Polym. Sci., 124, 3165 (2012).
[21.] Y. Wang, and X. Jing, Polym. Adv. Teclmol., 16, 344 (2005).
[22.] P. Saini, V. Choudhary, B.P. Singh, R.B. Mathur, and S.K. Dhawan, Mater. Chem. Pltys., 113, 919 (2009).
[23.] Y.L. Yang, and M.C. Gupta, Nano Lett., 5, 2131 (2005).
[24.] Y.L. Yang, M.C. Gupta, K.L. Dudley, and R.W. Lawrence, J. Nanosci. Nanotechnol., 5, 927 (2005).
[25.] H.M. Kim, K. Kim, S.J. Lee, J. Joo, H.S. Yoon, S.J. Cho, S.C. Lyu, and C. Lee, J. Carr Appl Phys, 4, 577 (2004).
[26.] H.M. Kim, K. Kim, C.Y. Lee, J. Joo, S.J. Cho, H.S. Yoon, D. A. Pejakovic, J.W. Yoo, and A.J. Epstein, Appl. Phys. Lett., 84, 589 (2004).
[27.] Y. Huang, N. Li, Y. Ma, F. Du, F. Li, X. He, X. Lin, H. Gao, Y. Ma, F. Li, Y. Chen, and P.C. Eklund, Carbon, 45, 1614 (2007).
[28.] Z. Liu, G. Bai, Y. Huang, Y. Ma, F. Du, F. Li, T. Guo, and Y. Chen, Carbon, 45, 821 (2007).
[29.] C.C.M. Ma, Y.L. Huang, H.C. Kuan, and Y.S. Chiu, J. Polym. Sci. Part B: Polym. Phys., 43, 345 (2005).
[30.] H.L. Wu, C.C.M. Ma, Y.T. Yang, H.C. Kuan, C.C. Yang, and C.L. Chiang,7. Polym. Sci. PartB: Polym. Pltys., 44, 1096 (2006).
[31.] W.S. Jou, H.Z. Cheng, and C.F. Hsu, J. Alloys Compd., 434, 641 (2007).
[32.] R.C. Che, L.M. Peng, X.F. Duan, Q. Chen, and X.L. Liang, Adv. Mater., 15, 401 (2004).
[33.] M.H. Al-Saleh, and U. Sundararaj, Carbon, 47, 2 (2009).
[34.] N.G. Sahoo, S. Rana, J.W. Cho, L. Li, and S.W. Chan, Prog. Polym. Sci., 35, 837 (2010).
[35.] J.J. Lin, W.H. Cheng, Y.F. Lan, J.C. Chiu, J.W. Lin, and C.M. Chang, U.S. Patent 000,396,5A1 (2011).
[36.] L. Li, and D.D.L. Chung, Composites, 25, 215 (1994).
[37.] X.P. Shui, and D.D.L. Chung, J. Mater. Sci., 35, 1773 (2000).
[38.] W.S. Jou, J. Electron. Mater. 33. 162 (2004).
[39.] S. Chakraborty, J. Pionteck, B. Krause, S. Banerjee, and B. Voit, Compos. Sci. Teclmol., 72, (1933) (2012).
[40.] S. Barrau, P. Demont, A. Peigney, C. Laurent, and C. Lacabanne, Macromolecules, 36, 5187 (2003).
[41.] S.V. Ahir, and E.M. Terentjev, Phys. Rev. B, 76, 165437 (2007).
[42.] Y.Y. Huang, S.V. Ahir, and E.M. Terentjev, Phys. Rev. B, 73, 125422 (2006).
[43.] L.A. Hough, M.F. Islam, P.A. Janmey, and A.G. Yodg, Phys. Rev. Lett A 93, 168102 (2004).
Aruna Kumar Mohanty, (1) Anindita Ghosh, (1,2) Pravin Sawai, (1) Kapil Pareek, (1) Susanta Banerjee, (1) Amit Das, (3) Petra Potschke, (3) Gert Heinrich, (3) Brigitte Voit (3)
(1) Materials Science Centre, Indian Institute of Technology, Kharagpur 721302, India
(2) Department of Applied Science, Symbiosis International University, Lavale, Pune 412115, India
(3) Leibniz Institute of Polymer Research Dresden, D-01069 Dresden, Germany
Correspondence to: Susanta Banerjee; e-mail: firstname.lastname@example.org and Amit Das; e-mail: email@example.com
Published online in Wiley Online Library (wileyonlinelibrary.com).
TABLE 1. Sample nomenclature, preparation technique and measurements. Sample MWCNT Polymer (a) Sample preparation code (wt%) (wt%) PEI-0.50 0.50 99.50 Solution cast PEI-0.75 0.75 99.25 Solution cast PEI-LOO 1.00 99.00 Solution cast PEI-1.50 1.50 98.50 Solution cast PEI-2.00 2.00 98.00 Solution cast PEI-3.00 3.00 97.00 Solution cast followed by compression molding PEI-4.00 4.00 96.00 Solution cast followed by compression molding PEI-5.00 5.00 95.00 Solution cast followed by compression molding PES-0.50 0.50 99.50 Solution cast PES-0.75 0.75 99.25 Solution cast PES-1.00 1.00 99.00 Solution cast PES-1.50 1.50 98.50 Solution cast PES-2.00 2.00 98.00 Solution cast PES-3.00 3.00 97.00 Solution cast followed by compression molding PES-4.00 4.00 96.00 Solution cast followed by compression molding PES-5.00 5.00 95.00 Solution cast followed by compression molding Sample Measurements code PEI-0.50 Conductivity, DMA PEI-0.75 Conductivity, DMA PEI-1.00 Conductivity, DMA PEI-1.50 Conductivity, DMA PEI-2.00 Conductivity, DMA PEI-3.00 SEM, TEM, SE PEI-4.00 SEM, TEM, SE PEI-5.00 SEM, TEM, SE PES-0.50 Conductivity, DMA PES-0.75 Conductivity, DMA PES-1.00 Conductivity, DMA PES-1.50 Conductivity, DMA PES-2.00 Conductivity, DMA PES-3.00 SEM, TEM, SE PES-4.00 SEM, TEM, SE PES-5.00 SEM, TEM, SE (a) Polymer: poly(ether imide) (PEI), poly(ether sulfone) (PES).
|Printer friendly Cite/link Email Feedback|
|Author:||Mohanty, Aruna Kumar; Ghosh, Anindita; Sawai, Pravin; Pareek, Kapil; Banerjee, Susanta; Das, Amit; P|
|Publication:||Polymer Engineering and Science|
|Date:||Nov 1, 2014|
|Previous Article:||Numerical simulation studies on heat and mass transfer using vacuum membrane distillation.|
|Next Article:||In vitro degradation of PLLA/nHA composite scaffolds.|