dielectric properties of isotactic polypropylene/nitrile rubber blends: effects of blend ratio,...

Dielectric Properties of Isotactic Polypropylene/NitrileRubber Blends: Effects of Blend Ratio, Filler Addition,and Dynamic Vulcanization

SNOOPPY GEORGE,1 K. T. VARUGHESE,2 SABU THOMAS1

1 School of Chemical Sciences, Mahatma Gandhi University, Priyadarshini Hills P.O., Kottayam 686 560, Kerala, India

2 Central Power Research Institute, Bangalore 560 094

Received 13 January 1998; accepted 3 November 1998

ABSTRACT: The dielectric properties of isotactic polypropylene/acrylonitrile–butadienerubber blends have been investigated as a function of frequency with special referenceto the effect of blend ratio. The dielectric properties measured were volume resistivity,dielectric constant (e9), dissipation factor (tan d), and loss factor (e0). At high frequen-cies, a transition in relaxation behavior was observed whereby the dielectric constant ofthe blends decreased with frequency, whereas the loss tangent and loss factor increasedon reaching a maximum. The variation of the dielectric properties with blend compo-sition was correlated with blend morphology, and relationships were established withreference to blend composition. Experimental e9 values were compared with theoreticalpredictions. The effect of the addition of fillers on the dielectric properties was alsoinvestigated for different fillers and filler loadings. It was found that silica fillerincreases the dissipation factor, whereas carbon black and cork gave a reverse trend.The variation in dielectric properties upon dynamic vulcanization of the rubber phaseusing different vulcanizing agents (such as sulfur, peroxide, and mixed systems) wasalso investigated. © 1999 John Wiley & Sons, Inc. J Appl Polym Sci 73: 255–270, 1999

Key words: polypropylene; nitrile rubber; blends; dielectric properties

INTRODUCTION

During the last few decades, thermoplastic elas-tomers (TPEs) replace usual thermoset rubbers inmany areas like the automotive and electricalsector. This replacement is due to the fact thatTPEs possess the excellent processing character-istics of thermoplastics with the properties of vul-canized rubbers and that the products from TPEscan be prepared at a lesser cost than that re-quired for vulcanized rubbers.1–3

Many of the TPEs are used in various electricalapplications (e.g., wire and cable), and as insula-

tion and jacketing materials due to their uniquecombination of properties (such as low tempera-ture flexibility, excellent insulating characteris-tics, and resistance to moisture absorption).4 Byblending suitably selected plastics with elas-tomers, materials with desirable final propertiescan be prepared. Electrical properties of variouspolymer blends have been investigated by differ-ent researchers.5–11 In their publications, it hasbeen shown that the dielectric properties of poly-mers and blends in general depend on structure,crystallinity, morphology, and the presence of fill-ers or other additives. The dielectric constant ofthe blends is found to increase with an increase inthe effective dipole moment. In styrene–buta-diene rubber/acrylonitrile–butadiene rubber(NBR) blends, the permittivity e9 was found to

Correspondence to: S. Thomas.Journal of Applied Polymer Science, Vol. 73, 255–270 (1999)© 1999 John Wiley & Sons, Inc. CCC 0021-8995/99/020255-16

255

increase with an increase in concentration ofC'N dipoles.11 The incorporation of polar compo-nents into polyethylene increased the dielectricconstant and dielectric loss of the blends.12 Themeasurement of dielectric properties as a func-tion of temperature was used as a way to studythe miscibility of different polymer blend sys-tems.5 Dielectric monitoring of thermosets andtheir blends have been used for the monitoring ofcuring.7 In heterogeneous polymer blends, the di-electric constants of the polymers are influencedby interfacial effects (i.e., due to the polarizationarising from difference in conductivities of the twophases).12 The volume resistivity of various poly-mer blends on incorporation of filler was alsostudied.13–17 The addition of conductive black intodifferent polymer blends to make conductingpolymers was reported.

Polypropylene (PP) has excellent electricalproperties like low dielectric constant, low lossfactor, high volume resistivity, and high surfaceresistivity, and is used as high-frequency insula-tor.18,19 NBRs have moderate insulating proper-ties and good oil resistance. Therefore, by blend-ing PP with NBR material with easy processabil-ity, good flexibility and excellent oil resistancecan be achieved. Earlier, we investigated the me-chanical, dynamic mechanical, and rheologicalproperties of these blends.20–22 In the presentarticle, we report on the effect of blend ratio anddynamic vulcanization on dielectric constant, dis-sipation factor, loss factor, and volume resistivityof PP/NBR blends. The effect of different fillers onthe dielectric properties is also investigated.

EXPERIMENTAL

Isotactic PP having a melt flow index (MFI) of 3g/10 min was supplied by IPCL Baroda. Acryloni-

trile-co-butadiene rubber, having an acrylonitrilecontent of 34%, was obtained from Synthetics andChemicals (Bareli, UP). The fillers used werehigh abrasion furnace black (HAF) (N 330), si-lane, treated silica, and cork powder. The silicawas prepared by the treatment of silica (100 g)with a Union Carbide A174 silane coupling agent(5 g).

Blends of isotactic PP/NBR was prepared indifferent blend ratios by melt mixing PP and NBRin a Brabender plasticorder PLE-330 at a temper-ature of 180°C. The rotor speed was 60 rpm, andtime of mixing was 6 min. The dynamically vul-canized blends were prepared by using the formu-lation given in Table I. The binary blends weredenoted P100, P70, P50, P30, and P0, where thesubscripts indicate the weight percentage of PP inthe blend. In the case of filled blends, the fillerloading was varied from 10 to 30 phr and is des-ignated as PxTsiy, PxKy, and PxCy, respectively,for silane-treated silica, cork, and carbon blackloading, where x denotes the weight percentage ofPP in the blends and y denotes the weight per-centage of filler with respect to the rubber phasein the blends.

The samples for electrical property measure-ments were prepared by compression molding thesamples into 2-mm-thick sheets in a hydraulicpress at a temperature of 200°C. Samples forvolume resistivity measurements were preparedby compression molding the samples into 0.3-mm-thick films.

The volume resistance of the samples weremeasured with a Twenty Million Megohmmetermodel 29A at room temperature. The volume re-sistivity (rv) was calculated using

rv 5 R z A/t V cm (1)

Table I Formulation of Dynamic-Vulcanized PP/NBR Blends

Ingredients Sulfur System (PS) DCP System (PC) Mixed System (PM)

PP 70.0 70.0 70.0NBR 30.0 30.0 30.0ZnO 5.0 — 5.0Stearic acid 2.0 — 2.0CBS 2.0 — 2.0TMTD 2.5 — 2.5S 0.2 — 0.1DCP — 2.0 1.0

ZnO–Zinc oxide; CBS, N-cyclohexyl benzothiazyl sulfenamide; TMTD, tetramethyl thiuram disulfide; S, sulfur.

256 GEORGE, VARUGHESE, AND THOMAS

where R is the volume resistance, A is the area,and t is the thickness of the sample.

The capacitance and tan d of the blends weremeasured at frequencies ranging from 31.6 Hz–10MHz at room temperature using a 4192 Imped-ance Analyzer of Hewlett–Packard. The dielectricconstant was obtained from the capacitance using

e9 5 C z t/e0 z A (2)

where C is capacitance, t is thickness, A is area,and e0 is 8.85 3 10212 F/m and

Loss factor: e0 5 e9 z tan d. (3)

The weight percentage crystallinity of PP wasdetermined using differential scanning calorime-try from the ratio of the heat of fusion of the blendto that of the 100% crystalline PP (DHpp 5 138 Jg21).23 Samples for scanning electron microscopicstudies were prepared by cryogenically fracturingin liquid nitrogen. From the fractured samples,the NBR phase was preferentially extracted withchloroform. The morphology of PP/NBR blendswas examined using a JEOL scanning electronmicroscope.

RESULTS AND DISCUSSION

Volume Resistivity

Resistivity studies are important for insulatingmaterials, because the most desirable character-istic of an insulator is its ability to resist theleakage of electrical current. Figure 1 shows thevariation of volume resistivity (rv) of PP/NBRblends as a function of NBR concentration. PP isa good insulator, with a volume resistivity in theorder 1016 V cm, whereas the volume resistivity ofnitrile rubber is in the order of 1010 V cm. In thecase of blends the rv has intermediate values inbetween that of PP and NBR. As the concentra-tion of NBR increases, the rv values decrease.This decrease is due to the introduction of lowresistivity NBR material. The curve shows asharp change after 50 wt % NBR. This change canbe correlated with the morphology of the blends.In P70 and P50 NBR is dispersed as domains in thecontinuous PP matrix, whereas in P30, NBR alsoforms a continuous phase resulting in a co-contin-uous morphology (Figure 2). The change of NBRfrom the dispersed phase to the continuous phase

should cause a sharp change in the electricalproperties.

The effect of dynamic vulcanization on rv val-ues of PP/NBR blends is given in Table II. Ondynamic vulcanization of the blends using sulfur,the resistivity of the blends decreases for differentblend compositions. Among the various vulca-nized systems, the dicumyl peroxide (DCP) sys-tem shows the highest resistivity value and sulfurthe lowest. The mixed system shows an interme-diate value. The difference in conductivity amongthe different vulcanizing systems may arise be-cause of the difference in the type of crosslinksformed during vulcanization. In the sulfur-vulca-nized system, SOS linkages are formed; whereasin the DCP crosslinked system, COC linkages areformed (Figure 3). In a mixed vulcanized system,a combination of both COC and SOS linkages areformed. Compared with COC linkages, SOS link-ages have ionic character, which may further in-crease the conductivity of the system. On vulca-nization of blends, some ionic impurities are in-corporated into the blend system as vulcanizingingredients; thus, conductivity of the final systemincreases upon vulcanization. In the case of per-oxide, vulcanization free radicals are present,which arise due the degradation of PP that mayalso increase the conductivity.

The effect of incorporation of fillers on the re-sistivity values is shown in Figure 4. It is seenthat the incorporation of 10 wt % carbon black,

Figure 1 Variation of volume resistivity of PP/NBRblends with NBR content.

DIELECTRIC PROPERTIES OF ISOTACTIC PP AND NBR BLENDS 257

silane-treated silica, and cork into P50 blend de-creases the resistivity values. In the case of car-bon black and silica, a further increase in theloading decreases the resistivity; whereas, in thecase of cork, as the loading increases, the resis-tivity rather increases. The decrease in resistivityfor silica and carbon black may arise from thepresence of ionic groups present in silica and car-bon black, which facilitates the conducting pro-cess. The nonconducting nature of the cork filler

accounts for the increase in resistivity of the PP/NBR blend at higher loadings of cork. By theincorporation of the conducting filler into poly-mers and blends, the resistivity generally de-creases and there is a critical loading of filler (i.e.,percolation threshold at which the polymer com-posites change from insulating to conducting ma-terial). Hashem and colleagues15 have reportedthat, in semi-reinforcing furnace (SRF) black-loaded butyl rubber, the percolation threshold is

Figure 2 Scanning electron micrographs of the morphology of PP/NBR blends. (a)70/30, (b) 50/50, and (c) 30/70 PP/NBR blends.


at 30 wt % carbon black. In PP/NBR blends, thechange in resistivity on increasing the loadingup to 30 wt % is less, and no percolation thresh-old is observed up to this loading.

Dielectric Constant, Loss Factor,and Dissipation Factor

Effect of Blend Ratio

The dielectric constant and loss tangent are im-portant parameters in the selection of an insulat-ing material. The variation of dielectric constantsof pure PP and the blends of PP/NBR as a func-tion of frequency is shown in Figure 5. The dielec-tric constant values of pure PP, NBR, and theirblends were decreased with increase in frequency.Generally the dielectric constant of a materialarises due to polarization of molecules, and thedielectric constant increases with increase in po-larizability. The different types of polarizationspossible in a material are the polarizations aris-ing from: (1) electronic polarization, (2) atomicpolarization, and (3) orientation polarization dueto the orientation of dipoles parallel to the appliedfield.24 For heterogeneous materials, there is alsothe possibility for interfacial polarization thatarises due to the differences in conductivities ofthe two phases.25 The time required for each typeof polarization to reach the equilibrium level var-ies with the nature of the polarization.

The orientation polarization requires muchmore time, compared with electronic and atomicpolarization to reach static field value. Therefore,the orientation polarization decreases with in-crease in frequency, at a lower frequency region,compared with electronic and atomic polariza-tions. Interfacial polarization generally occurs atstill lower frequencies. In the dielectric constantvs. frequency plot of PP/NBR blends, it is seenthat the reduction in e9 occurs in three stages. Inthe first stage, the high values of dielectric con-stant can be attributed to the interfacial polariza-

tion effects. The high value of e9 for PP indicatesthe presence of impurities in the PP. It may alsodue to the fact that the impurities in such poly-mers may cause some oxidation at structural de-fects that lead to the formation of oxidized centersthat then causes interfacial polarization.4

In the case of blends, due to the presence of twophases of NBR and PP with different conductivi-ties, interfacial polarization occurs leading to anincrease in dielectric constant. Because interfa-cial polarization decreases with an increase infrequency, as the frequency increases to 100 Hz,

Table II Volume Resistivity Values ofUnvulcanized and Dynamic-Vulcanized P70

Blends

Sample Volume Resistivity 3 10213 V cm

P70 24.00PS70 3.52PC70 8.65PM70 5.82

Figure 3 Schematic representation of variouscrosslinks formed during vulcanization using: (a) sul-fur, (b) peroxide, and (c) mixed system.


the dielectric constant decreases considerably. Inthe region 3.16 3 102–105 Hz frequency, the di-electric constant has contribution from orienta-tion and atomic and electronic polarization. Here,the dispersion region spreads over a wide range offrequencies. Above 3.16 3 105 Hz frequency the e9further reduces, which may be due to the drop inorientation polarization. At this frequency, thedrop in e9 increases with the increase in the rub-ber content. It is due to the increase in dipoles inthe materials at a high rubber concentration,which leads to a decrease in orientation polariza-tion at a higher rate, compared with the blendswith a lesser number of dipoles.

The variation of dielectric constant with weightpercentage of NBR at three different frequenciesis shown in Figure 6. PP shows the lowest e9value, similar to a nonpolar material, and nitrilerubber shows the highest value. Generally, for anonpolar material, the dielectric constant equalsthe square of refractive index (n2). The refractiveindex of PP is 1.47 and hence e9 5 2.16. Theexperimental value is 2.805. The higher valuesmay be due to the presence of interfacial polar-ization due to the presence of impurities. For po-lar materials, the e9 values are generally greaterthan n2 and e9 increases with an increase in po-larity. In the case of blends, the dielectric con-stant increases as the rubber content increases.This increase in e9 with the incorporation of rub-

ber is due to an increase in C'N dipoles, whichincrease the orientation polarization and also dueto the presence of interfacial polarization. Again,the orientation of dipoles depends on the crystal-linity of the medium. Upon the introduction ofrubber, the crystallinity of the system decreases(Table III). As the crystallinity decreases, the di-

Figure 4 Variation of volume resistivity of 50/50 PP/NBR blends with wt % of filler for different fillers.C-black, carbon black.

Figure 5 Variation of dielectric constant of PP/NBRblends with frequency.

Figure 6 Variation of dielectric constant of PP/NBRblends with NBR content.


poles can orient more easily. Such an increase indielectric constant on incorporation of polar poly-mers was reported.12 The dielectric constant de-pends on resistivity by the equation26

log R10~298 K! 5 23 2 2 z e9~298 K! (4)

(i.e., the electrical resistance of polymers de-creases exponentially with increasing dielectricconstant). In PP/NBR blends, as the dielectricconstant increases with an increase in rubbercontent, the volume resistivity decreases (Figure1). The increase in e9 is more pronounced above 50wt % NBR (Figure 6). This can be correlated withblend morphology. In PP/NBR blends at 30 and50 wt % NBR, NBR forms the dispersed phase;while in blend with 70 wt % NBR, both NBR andPP forms continuous phases leading to a co-con-tinuous morphology (Figure 2). The continuousnature of the NBR phase leads to a better orien-tation of dipoles and results in a high dielectricconstant.

Experimental data can be compared withmodel calculations. The dielectric constant of acomposite containing two components can be ex-pressed in the general form:

e9c 5 V1e91 1 e92~1 2 V1!, (Model I) (5)

where e9 and e92 are dielectric constants of compo-nents 1 and 2, and V1 and V2 5 1 2 V1 are thevolume fractions of components 1 and 2, respec-tively. The logarithmic variation of dielectric con-stant was also expressed by the equation:

log e9c 5 V1log e91 1 ~1 2 V1!log e92 (Model II) (6)

The dielectric constant of two phase mixturesbased on spherical particle, which consider all thepossible interaction, was also calculated for PP/NBR in which NBR is dispersed as spherical do-mains in the continuous PP matrix using27

e9c 5 1/4~H 1 ~H2 1 8e91e92!1/2! (Model III) (7)

where

H 5 ~3V1 2 1!e91 1 ~3V2 2 1!e92. (8)

The Maxwell–Wagner–Sillars equation was alsoused to predict the e9 values that are given as28

e9c 5 e922e92 1 e91 1 2V1~e91 2 e92!

2e92 1 e91 2 V1~e91 2 e92!(Model IV) (9)

Figure 7 gives the experimental and the theoret-ical variations of dielectric constants with blendcomposition. The experimental values are lowerthan that of the parallel model and is close to loga-rithmic mixing rule. The Maxwell–Wagner–Sillarsequation gives a better correlation of dielectric con-stants with experimental results.

The measurement of dissipation factor (tan d)and loss factor (e0) of insulating material is im-portant because the loss tangent is a measure ofthe alternating current electrical energy that isconverted to heat. This heat raises the tempera-ture and accelerates deterioration. The variationof dissipation factor and loss factor (e0) with fre-quency of PP and PP/NBR blends is shown inFigure 8. Frequency increases the dissipation fac-tor and increases the loss factor for pure PP and

Figure 7 Experimental (at 31.6 KHz) and theoreticalvariations of e9 values with wt % of NBR.

Table III Crystallinity of PP/NBR Blends

Composition Crystallinity (%) from DSC

P100 55.3P70 33.9P50 20.9P30 13.7

DSC, differential scanning calorimetry.


the blends. A relaxation region is observed in thefrequency range of 3.16 3 105–3.16 3 106 Hz,which may be due to a lag in dipole orientationbehind the alternating electric field. Also, it isseen from Figure 8 that the dissipation factor andmagnitude of relaxation increase with an increasein rubber content. In the case of PP, which isnonpolar, relaxation is not due to dipole orienta-tion. But, on the incorporation of nitrile rubber,dipoles are introduced into the system; this leadsto a lag in orientation of dipoles on the applicationof the electric field. Hence, the dissipation factorincreases with an increase in rubber content.Blends of NBR with PP show the presence of arelaxation peak maximum at a frequency 106 Hz.As the concentration of NBR increases, the relax-ation peak maximum shifts to higher frequency(i.e., to 3.16 3 106 Hz). This suggests that therelaxation time decreases with increase in rubbercontent.

In polar polymer, it was reported that the in-crease in crystallinity reduces the tan d valuesand also leads to an increase in relaxation time(i.e., their peaks corresponding to a and b relax-ation process shifted to lower frequency), becausethe relaxation time t is given by

t 5 1/2pfm (10)

where fm is the frequency corresponding to themaximum in relaxation peak. In PP/NBR, as the

concentration of NBR increases, the crystallinityin PP/NBR blends decreases. Hence, the observedshift in relaxation frequency or reduction in re-laxation time is due to the reduction in crystallin-ity. As the crystallinity of the system decreases,the rotatory motion of the dipoles becomes moreeasy, which leads to higher values of tan d.

The sharp increase in tan d beyond 50 wt %NBR is due to a change in morphology of theblend (i.e., the phase inversion of NBR from thedispersed to the continuous phase). It is seen thatthe incorporation of rubber into PP did not in-crease the value of tan d by . 3%, which is fre-quently desired to avoid problems leading to fail-ure of the insulator.

Effect of Fillers

Fillers are generally incorporated into TPEs tomodify physical and electrical properties. The ef-fect of the addition of carbon black on e9 values in50/50 PP/NBR blend is shown in Figure 9. Theincorporation of carbon black into the P50 blendincreases the dielectric constant. This increase isdue to the conducting nature of the black that isassociated with a conjugated bonding structurepresent in the crystalline regions and also due tothe presence of polar groups in carbon black. It isagain seen that, as the loading of carbon blackincreases, the dielectric constant increases, and

Figure 9 Variation of dielectric constant of carbonblack-loaded P50 blends with frequency.

Figure 8 Variation of dissipation factor tan d and lossfactor e0 of PP/NBR blends with frequency.


this increase is more pronounced for 30% carbonblack. Figure 10 depicts the variation of the di-electric constant of PP/NBR blends at 30 wt %loading of carbon black. Upon the incorporation ofcarbon black, the dielectric constants of all theblends are increased, and this increase becomesmore pronounced with an increase in rubber con-tent. It is seen from the figure that, at 70 wt %NBR, the dielectric constant shows a sharp in-crease in size, compared to P70 and P50. This canbe explained in terms of the morphology of thesystem. The morphology of filled blends can beschematically represented as shown in Figure 11.In P70 and P50, NBR forms a dispersed phase in acontinuous PP phase, whereas in P30, NBR alsoforms a continuous phase resulting in a co-contin-uous morphology. Again, at this black concentra-tion (30%), the carbon black forms a continuousnetwork in the continuous nitrile rubber phase.This leads to an increase in conductivity of systemand an increase in dielectric constant. Carbonblack is predominantly dispersed in the NBRphase due to its high polarity. Generally, in ther-moplastic elastomers, the fillers locate in the rub-ber phase.4

The variation of dielectric constant with theincorporation of silane-treated silica is shown inFigure 12. In these samples, also the e9 valuesdecrease with frequency in three steps, due to therelaxation in interfacial and dipole polarizations.

It is seen from Figure 12 that, upon incorporationsilica, the dielectric constant increases up to 20 wt% loading. However, at 30 wt % of loading, the e9value decrease. The increase in e9 on the additionof silica may be due to the presence of polargroups present in the filler. As the filler loadingincreases, the density of the system is also in-creased and the extent of orientation of dipoles isretarded; thus the e9 shows a decrease at thisconcentration. Figure 13 shows the variation of e9with frequency for the P70, P50, and P30 blendscontaining 30 wt % treated silica. As the concen-tration of rubber increases, the e9 increases and itis seen that the incorporation of silica increasesthe dielectric constant. The increase is more inthe case of the P70 blend, whereas in P50 and P30,the silica only slightly affects the e9 values.

Figure 11 Schematic representation of the morphol-ogy of carbon black-loaded PP/NBR blends.

Figure 10 Variation of dielectric constant of 30 wt %carbon black-loaded PP/NBR blends with frequency.


Figure 14 depicts the variation of e9 with fre-quency for the P50 blend containing a varyingconcentration of cork. Up to 10 wt % cork, the e9increases while further loading decreases the e9values. The increase in e9 on the addition of corkis due to interfacial polarization. The variation of

dielectric constant with 30 wt % cork-loaded P70,P50, and P30 blends are shown in Figure 15. Thenature of relaxation spectra is not affected by thepresence of cork. Here, too, the dielectric constantincreases with an increase in rubber content, butthe difference is less than those with the other

Figure 12 Variation of dielectric constant of silane-treated, silica-loaded P50 blends with frequency.

Figure 13 Variation of dielectric constant of 30 wt %silane-treated, silica-loaded PP/NBR blends with fre-quency.

Figure 14 Variation of dielectric constant of cork-loaded P50 blends with frequency.

Figure 15 Variation of dielectric constant of 30 wt %cork-loaded PP/NBR blends with frequency.


two fillers, which may be due to the nonconduct-ing character of cork.

The variation of loss factor with frequency forcarbon black, cork, and silane-treated silica-filled50/50 PP/NBR blends is shown in Figures 16–18,respectively, for different filler loadings. Figure16 shows the variation of loss factor, e0, with

frequency for carbon black-loaded samples. Theloss factor shows an increase with filler loading.The addition of carbon black did not alter therelaxation process; however, the peak height cor-responding to the relaxation of C'N dipole in-creases. At 30 phr loading, the peak maximumshifts to a lower frequency of 106 Hz, comparedwith 3.16 3 106 Hz for the other loadings. Theincorporation of filler into the blend system in-creases the dielectric loss, and this may be due tothe presence of some oxidized groups in carbonblack and also due to the increase in conductanceloss. The shift in relaxation frequency to low val-ues (i.e., reduced relaxation time) indicates thatthe resistance for relaxation of C'N dipoles in-creases at 30 phr loading. This indicates that thefiller is mainly located in the rubber phase.

Figure 17 shows the change in loss factor withfrequency for different loadings of cork filler inthe P50 system. The incorporation of cork into theblend system slightly decreases the loss factor. Asthe loading of cork increases, the loss factor alsodecreases. The lowest value is obtained for theblend with 30 phr filler loading. The reduction inloss factor for the cork filler is due to the hin-drance for the orientation of C'N dipoles in thepresence of filler. As the loading of filler increases,the orientation of dipoles is retarded by the pres-ence of the cork. The width of the relaxation peakalso decreases with increase in filler loading.

Figure 16 Variation of loss factor of carbon black-loaded P50 blends with frequency.

Figure 17 Variation of loss factor of cork-loaded P50

blends with frequency.

Figure 18 Variation of loss factor of silane-treated,silica-loaded P50 blends with frequency.


In the case of treated silica filler, the variationof loss factor with frequency is depicted in Figure18. Here, the loss factor increases up to 20 phrloading for 30 phr loading of silica filler, the lossfactor shows a decrease. The silica filler containspolar groups; this also contributes to the loss fac-tor due to the increase in effective dipole moment.This, in fact, leads to an increase of e0 values onincorporation of treated silica filler to the P50blend. However at 30 phr loading, although thereis an increase in the effective dipole moment, therelaxation process is retarded due to an increasein viscosity. In effect, the loss factor shows a de-crease at this loading.

The variations in loss factor with NBR contentfor different fillers are shown in Figures 19–21.As the rubber content in the blend increases, theloss factor (e0) increases in all cases. The increasein loss factor (e0) with rubber content is low up to50 wt % and after that the loss factor (e0) in-creases sharply. This sharp increase in e0 above50 wt % NBR is due to the change in morphology(i.e., phase inversion of NBR from the dispersedphase to the continuous phase . 50 wt % NBRleads to a better orientation of dipoles). Amongthe filled blends, in carbon black-loaded samples,the relaxation peak is observed at a frequency of106 Hz in all composition, whereas in silane-treated silica-filled samples, the peak is observedat 3.16 3 106 Hz. In cork-loaded samples, thepeak is at 106 Hz for 70/30 composition; and, at

higher NBR content, the peak is shifted to higherfrequency.

Effect of Dynamic Vulcanization

Vulcanization of the rubber phase during mixing,known as dynamic vulcanization, is reported to bean efficient technology to improve the properties

Figure 19 Variation of loss factor of 30 wt % carbonblack-loaded PP/NBR blends with frequency.

Figure 20 Variation of loss factor of 30 wt % cork-loaded PP/NBR blends with frequency.

Figure 21 Variation of loss factor of 30 wt % silane-treated, silica-loaded PP/NBR blends with frequency.


of thermoplastic elastomers.29,30 Dynamic vulca-nization leads to a uniform distribution of rubberparticles in the thermoplastic matrix. The mor-phology is also stabilized. This is schematicallyrepresented in Figure 22. Figure 23 depicts thevariation of dielectric constant e9 with frequencyfor unvulcanized, sulfur, peroxide, and mixed-sys-tem vulcanized P70 blends. It is seen that vulca-nization of the rubber phase leads to an increasein dielectric constant. The variation in e9 is mar-ginal in the case of the sulfur crosslinked system,compared with the other systems. Among the dif-ferent systems used, the peroxide system showsthe highest e9 values, whereas the sulfur systemshows the lowest value. The mixed (S 1 peroxide)system shows e9 values in between. The highervalues of peroxide and mixed system may be dueto the increased interfacial polarization, but alsodue to the degradation of the PP phase in thepresence of DCP. The degradation of the PP phasemay decrease the crystallinity, which leads tobetter dipole orientation. Because the rubber par-ticles are finely dispersed upon dynamic vulcani-zation, the interfacial area increases and thisleads to an increase in interfacial polarization.Therefore, the dielectric constant increases. Thephase morphologies of the 70/30 blend vulcanizedwith sulfur, peroxide, and mixed system areshown in Figure 24. It is seen that, in the DCPvulcanized system, the rubber is more finely anduniformly distributed, compared with mixed andsulfur vulcanized systems. The size of NBR do-mains increase in the order sulfur . mixed . per-oxide systems. Although the DCP-vulcanized sys-tem has the highest crosslink density, the effect isnot observed in the properties due to the degra-dation of PP.

Figure 25 shows the variation of dielectric con-stant with frequency for P70, P50, and P30 blendsvulcanized with sulfur. As the concentration ofrubber increases, the dielectric constant of the

crosslinked blends are found to be slightly higherthan that of the uncrosslinked ones at low fre-quencies; whereas, at high frequencies, the trendis reversed. The increase in dielectric constant forvulcanized blends at low frequencies may be dueto the increase in interfacial polarization, whichreduces with an increase in frequency; at highfrequencies, the effect from orientation polariza-tion predominates. On vulcanization of the rubberphase, the relaxation of C'N dipoles is retardeddue to the presence of crosslinks between thechains. This leads to a low value of dielectricconstant that arises from orientation polariza-tion.

The variations of dissipation factor (tan d) andloss factor with frequency are shown in Figures26 and 27. It is seen that the dissipation factor isnot affected much by vulcanization. But, at thehigh-frequency region of 107 Hz, the sulfur sys-tem shows the highest tan d value, and the per-oxide system showed the lowest. The mixed vul-canized system takes an intermediate position.On dynamic vulcanization, the magnitude of e0increases slightly. However, as the concentrationof rubber increases to 50 and 70 wt % NBR, therelaxation peak frequency is shifted to a lowervalue (i.e., the relaxation time increases). Theincrease in relaxation time arises due to the in-

Figure 23 Variation of dielectric constant of unvul-canized and dynamic-vulcanized 70/30 PP/NBR blendswith frequency.

Figure 22 Schematic representation of the morphol-ogy of dynamic-vulcanized thermoplastic elastomers.


crease in viscosity that leads to more constraintsto the relaxation of C'N dipoles. The relaxationspectra of 70/30 blends vulcanized with differentvulcanizing systems are shown in Figure 27. Theperoxide and mixed system vulcanized blendshave higher peak heights, compared with unvul-canized and sulfur-vulcanized systems. In thepresence of DCP, the PP phase is degraded, andthis may lead to the presence of some oxidizedcenters. The dielectric constant is also affected by

the decrease in crystallinity. This will increasethe dielectric relaxation and hence the peroxideand mixed-vulcanized systems show higher val-ues of e0. It is seen that the relaxation peak widthincreases upon dynamic vulcanization.

CONCLUSIONS

The dielectric properties (such as volume resistiv-ity, dielectric constant and tan d of PP/NBR

Figure 24 Scanning electron micrographs of the morphology of dynamic-vulcanized70/30 PP/NBR blends. (a) Sulfur, (b) peroxide, and (c) mixed system.


blends) were investigated over a wide range offrequencies. The dielectric constant of the purecomponents and blends decreases with an in-crease in frequency. Variation in dielectric con-stant is in three stages due to the relaxation ininterfacial and orientation polarization. Upon theincrease of NBR content, the dielectric constant

increased and showed a sharp change after 50 wt% loading. This sharp change was correlated withthe morphology of the blend. The experimentaldielectric constant values were compared withtheoretical models. It was found that experimen-tal data fit well with data obtained from the Max-well–Wagner–Sillars equation. The effects of var-ious fillers (such as HAF black, silane-treatedsilica, and cork) on dielectric properties were alsoinvestigated. The fillers affect the dielectric prop-erties. The dielectric constant increases on theaddition of carbon black, and the increase is morepronounced in the P30 blend. The change is corre-lated with the formation of a network of carbonblack particles in the continuous NBR phase at 70wt % NBR. In silane-treated silica-filled blends,the dielectric constant increased up to 20 wt %loading, whereas at 30 wt % loading, the e9 valuedecreased; in cork-filled blends, the e9 value in-creased by the addition of 10 wt % cork, whereasfurther loading decreased the e9 values. The in-crease in e9 value upon the addition of fillers wascorrelated with the presence of polar groups inthe filler and interfacial polarization. The lossfactor increased upon the addition of carbon blackand silica, whereas the addition of cork decreasedthe e9 values. Dynamic vulcanization leads to anincrease in e9 values. Among the different vulca-nizing systems used, the sulfur system showedthe lowest value, and the DCP-vulcanized system

Figure 25 Variation of dielectric constant of dynam-ic-vulcanized PP/NBR blends with frequency.

Figure 26 Variation of dissipation factor and lossfactor of unvulcanized and dynamic-vulcanized 70/30PP/NBR blends with frequency.

Figure 27 Variation of dissipation factor and lossfactor of dynamic-vulcanized PP/NBR blends with fre-quency.


showed the highest e9 value. The mixed-vulca-nized system showed an intermediate value. Thedissipation factor and loss factor values also in-creased due to vulcanization.

One of the authors (S.G.) is thankful to the Council ofScientific and Industrial Research, New Delhi, for fi-nancial support. We also thank Dr. K. G. K. Warrier,Head, Ceramic Division, Regional Research Labora-tory, Thiruvananthapuram, and authorities of the Cen-tral Power Research Institute, Bangalore, for providingtesting facilities.

REFERENCES

1. Legge, N. R.; Holden, G.; Schoeder, H. E., Eds.Thermoplastic Elastomers—A Comprehensive Re-view; Hanser Publishers: Munich, 1987.

2. Wheelan, A.; Lee, K. S., Eds. In Developments inRubber Technology; Applied Science Publishers:New York, 1988.

3. De, S. K.; Bhowmick, A. K., Eds. ThermoplasticElastomers from Rubber-Plastic Blends, Ellis Hor-wood: New York, 1990.

4. Walker, B. M.; Rader, C. P., Eds. Handbook ofThermoplastic Elastomers; Van Nostrand Rein-hold: New York, 1979.

5. Malik, T. M.; Prud’homme, R. E. Polym Eng Sci1984, 24, 144.

6. Pathmanathan, K.; Cavaille, J. Y.; Johari, G. P.Polymer 1988, 29, 311.

7. Maistros, G. M.; Block, H.; Bucknall, C. B.;Patridge, I. K. Polymer 1992, 33, 4470.

8. Pillai, P. K. C.; Narula, G. R.; Tripathy, A. K. PolymJ 1984, 16, 575.

9. Radhakrishnan, S.; Saini, D. R. J Appl Polym Sci1994, 52, 1577.

10. Gustafsson, A.; Salot, R.; Gedde, U. W. PolymComp 1993, 14, 5421.

11. Mansour, A. A.; Sabagh, S. EL.; Yehia, A. A. JElastomers and Plastics 1994, 26, 367.

12. Schifani, R.; Spadaro, G.; Cassata, F.; Valenza, A.Eur Polym J 1995, 31, 841.

13. Soares, B. G.; Gubbels, F.; Jerome, R.; Vanlathem,E.; Deltour, R. Rubber Chem Technol 1997, 70, 60.

14. Sumita, M.; Sakata, K.; Asai, S.; Miyasaka, K.;Nakagawa, H. Polym Bull 1991, 25, 265.

15. Abo-Hashem, A.; Saad, H. M.; Ashor, A. H. PlastRubb Comp Proc Applic 1994, 21, 125.

16. Sircar, A. K. Rubber Chem Technol 1981, 54, 820.17. Gubbels, F.; Jerome, R.; Teyssie, Ph.; Vanlathem,

E.; Deltour, R.; Calderone, A.; Parente, V.; Bredas,J. L. Macromolecules 1994, 27, 1972.

18. MacCrum, N. G.; Read, B. E.; Williams, G. Anelas-tic and Dielectric Effects in Polymeric Solids; JohnWiley: New York, 1967.

19. Brydson, J. A. Plastic Materials; Butterworths:London, 1989; p. 105.

20. George, S.; Joseph, R.; Varughese, K. T.; Thomas,S. Polymer 1995, 36, 4405.

21. George, S.; Prasannakumari, K. L.; Koshy, P.; Va-rughese, K. T.; Thomas, S. Mater Lett 1996, 26, 51.

22. George, S.; Neelakantan, N. R.; Varughese, K. T.;Thomas, S. J Polym Sci Polym Phys (Edn) 1997, 35,2309.

23. Boder, G. Structural Investigations of Polymers;Ellis Horwood: New York, 1991.

24. Mathes, K. N. In Encyclopedia of Polymer Scienceand Engineering, Vol. 5; Kroschwitz, J. I., Ed.;John Wiley and Sons: New York, 1986; p. 507.

25. Ku, C. C.; Liepins, R. Electrical Properties of Poly-mers, Chemical Principles; Hanser Publishers:Munich, 1987.

26. Bicerano, J. Prediction of Polymer Properties; Mar-cel Dekker, Inc.: New York, 1993; Chap. 9.

27. Reynolds, J. A.; Hough, J. M. Proc Phys Soc LondonSect B 1957, 70, 769.

28. Debnath, S.; De, P. P.; Khastgir, D. Rubber ChemTechnol 1987, 61, 555.

29. Liao, F. S.; Su, A. C.; Hsu, T. C. J. Polymer 1994,35, 2579.

30. Kuriakose, B.; De, S. K.; Bhagawan, S. S.; Si-varama Krishnan, R.; Athithan, S. K. J Appl PolymSci 1986, 32, 5509.


dielectric properties of isotactic polypropylene/nitrile rubber blends: effects of blend ratio,...

Documents