The effect of poly[styrene- b -(ethylene- co -butylene)- b -styrene] on dielectric, thermal, and morphological characteristics of polypropylene/silica nanocomposites

The Effect of Poly[styrene-b-(ethylene-co-butylene)-b-styrene] on Dielectric, Thermal, and MorphologicalCharacteristics of Polypropylene/Silica Nanocomposites

Denis Mihaela Panaitescu,1 Zina Vuluga,1 Petru V. Notingher,2 Cristian Nicolae1

1 Polymer Department, National Institute for Research and Development in Chemistry and Petrochemistry,060021 Bucharest, Romania

2 Faculty of Electrical Engineering, ELMAT Laboratory, University POLITEHNICA of Bucharest,060042 Bucharest, Romania

The effect of poly[styrene-b-(ethylene-co-butylene)-b-styrene] (SEBS) copolymer on the thermal and dielec-tric properties of polypropylene (PP)—nanosilica (NS)composites in relation with morphological aspectsrevealed by atomic force microscopy (AFM) was inves-tigated in this article. SEBS hindered the crystallizationprocess of PP in PP/NS composites, leading to asmaller degree of crystallinity and lower perfection ofcrystalline structure. Broader lamellar thickness distri-bution was obtained in nanocomposites containingSEBS. Almost two times higher dielectric loss as com-pared to PP reference and two relaxation processeswere detected in e00r (f) curves of nanocomposites. Thefirst peak, in the same frequency domain as for thereferences, was assigned to a-relaxation of polymercomponents together with interfacial polarization. Therelaxation time follows the Arrhenius law with an acti-vation energy of 80–90 kJ/mol. For the second process,the temperature dependence of the relaxation timesobeyed the VFT equation. The dielectric changes fol-lowing the incorporation of SEBS support its tendencyto hinder the motional processes in PP, in accordancewith DSC results. A smooth transition from a phaserich in SEBS to one containing mainly PP was detectedin the AFM image of the composite with the largeramount of SEBS, emphasizing the good compatibilityat the PP/SEBS interface. POLYM. ENG. SCI., 53:2081–2092, 2013. ª 2013 Society of Plastics Engineers

INTRODUCTION

The incorporation of nanosilica (NS) in polypropylene

(PP) is a promising method to obtain materials with high

stiffness and strength. Polypropylene/NS composites

combine the excellent processability, thermal stability,

recyclability, and low cost of PP with the high stiffness

of NS [1, 2]. Nevertheless, compatibilizing agents like

maleic anhydride-modified PP (MA-PP) must be added to

ensure a good dispersion of the nanofiller and a good PP/

NS interface [3–8]. To grow PP competitiveness in engi-

neering applications, a simultaneously increase in stiffness

and toughness is necessary. Toughness of PP/NS compo-

sites can be considerably enhanced by the incorporation

of a dispersed elastomer phase [9]. In recent studies,

poly(styrene-b-ethylene-co-butylene-b-styrene) copolymer

(SEBS) was preferred to conventional elastomers for

improving the toughness and compatibility in PP compo-

sites because it determines the increase of ductility at low

contents and an acceptable decrease in stiffness, leading

to better mechanical performance [10–14]. Therefore, a

good stiffness–toughness balance in PP composites may

be achieved by the melt-mixing of PP with NS and SEBS.

Only a few studies emphasize the influence of SEBS on

the mechanical and morphological properties of NS rein-

forced PP [15]. In the case of PP/SEBS (70/30 vol.%)

blend containing 6 vol.% NS, Mae et al. observed that the

elastic modulus increased when NS were located outside

SEBS domains and the strain energy up to failure

decrease was prevented when nanoparticles were located

inside rubber particles leading to ductile fracture [15]. To

our knowledge, no reports have documented the dielectric

properties of these materials and no correlation with

microstructural features have been done, although these

materials are extremely interesting for electrotechnic and

automotive industries.

Dielectric spectroscopy can provide quantified insights

into the molecular dynamics of polymeric materials and

to give valuable information on the interaction between

components, the reinforcing effect of nanofillers, or the

toughening effect of the elastomer phases in polymer

blends [16, 17]. Several dielectric studies on the molecu-

lar relaxation behavior of PP are available [18–20].

Although oxidation and chain scission introduce polar

Correspondence to: D.M. Panaitescu; e-mail: [email protected]

DOI 10.1002/pen.23475

Published online in Wiley Online Library (wileyonlinelibrary.com).

VVC 2013 Society of Plastics Engineers

POLYMER ENGINEERING AND SCIENCE—-2013

groups in PP, dielectric spectroscopy of this nonpolar

polymer is less commonly used as a characterization

tool. Nevertheless, the addition of low concentration of

dielectric probe (polar rigid-rod-like chromophores)

allowed the dielectric study of glass transition dynamics,

crystallization, and melting of PP [18, 19]. Recently, a

complete relaxation map of gamma-radiated PP, obtained

by dielectric measurements as a function of temperature,

has been reported [20]. Two relaxation processes, a-

relaxation at high temperature and b-relaxation related to

glass-rubber transition (Tg), were observed in irradiated

samples. Although the basic relaxation processes in PP

are clear, the molecular origin and the morphological

assignment of each molecular relaxation are still contro-

versial issues. The most important difference in maleated

polypropylene (MA-PP) as compared to PP is that the

maleic anhydride (MAH) modifies the polarity of the PP

chains and provides structural units with a permanent

dipole moment [21, 22]. Bohning et al. [21] have

reported that pure PP shows no dielectrically active

relaxation process in comparison with MA-PP, which

shows dielectric loss at ambient temperature and low fre-

quency (10 Hz). This relaxation follows an Arrhenius

law for temperatures below 208C (activation energy of

47.3 kJ/mol) and was attributed to the local molecular

fluctuations. Information regarding the frequency range

of the main relaxation processes in PP and PP modified

with MA-PP in the published literature is even scarce.

Pandis et al. [23] have shown that PP without additives

exhibits no detectable dielectric relaxation mechanisms

and e0 remains constant from 1022 to 106 Hz, but Suljov-

rujic [20] has detected a b relaxation process, related to

the glass transition, close to the room temperature in the

isochronal dielectric spectrum of isotactic PP at 1 MHz.

Few works related to dielectric properties of PP blends

and nanocomposites have been reported [21, 24–27].

Two relaxation processes were detected in MA-PP–clay

nanocomposites by Bo†hning et al. [21], which considered

that the first relaxation process, also present in neat MA-

PP, could be assigned to localized fluctuations of the po-

lar MAH groups (a-relaxation), and the second, the high-

temperature process, could be due to Maxwell–Wagner–

Sillars (MWS) polarization.

In most cases, exploitation temperatures coincide with

the a-relaxation zone and information about dielectric

behavior becomes extremely important for engineering

applications. There is no information, to the best of our

knowledge, on dielectric spectroscopy of PP/SEBS blends

and PP/NS composites containing SEBS. Therefore, in the

present article, the dielectric spectra of PP/NS composites

containing different amount of SEBS have been discussed

and correlated with morphological aspects revealed by

atomic force microscopy (AFM) and differential scanning

calorimetry (DSC). MA-PP was used to improve the

compatibility between NS and the matrix and to induce

preferential location of the nanofiller in PP to obtain

balanced stiffness-toughness composites [15].

EXPERIMENTAL

Materials

Polypropylene homopolymer Moplen HP400R (PP)

produced by Bassel Polyolefins (Italy) with a MFI of 25

g/10 min (2308C/2.16 kg) and a density of 0.90 g/cm3

was used for blends and nanocomposites preparation. The

impact modifier was a Kraton 1652 G (SEBS) from

Kraton Polymers, a linear poly [styrene-(ethylene-co-bu-

tylene)-styrene] block copolymer with a styrene content

of 29%, Mn ¼79,100, density of 0.91 g/cm3, and MFI ¼5.00 g/10 min (2308C/5 kg). Maleic anhydride grafted PP

(MA-PP) from Aldrich, with a density of 0.91 g/cm3, a

melting point of 1578C, and containing 1 wt% grafted

MA was used as a compatibilizing agent. Silica nanopow-

der (NS), amorphous, with a content of 99.5% SiO2, the

average particle diameter of 15 nm, density 2.2 g/cm3,

bulk density 0.011 g/cm3, and the specific surface area of

180 m2/g was supplied by Aldrich.

Preparation of PP Nanocomposites

PP/NS nanocomposite was prepared by a two-step pro-

cess. 1:1 NS/MA-PP masterbatch was prepared by direct

mixing MA-PP with NS in a mixing chamber of

Brabender Plasticorder LabStation at a temperature of

165–1708C. SiO2 nanopowder was slowly added in the

polymer melt over a period of 5 min and then mixed for

another 10 min. PP was mixed with NS masterbatch using

the same mixing chamber at a temperature of 1758C for

10 min, the speed of the rotors being set at 100 rpm.

Three references were processed in the same conditions

specified above: neat PP, PP with 5 wt% MA-PP desig-

nated as PPM, and neat SEBS. For the preparation of

nanocomposites containing SEBS, the block copolymer

was added in the mixing chamber immediately after PP

melted, before the addition of NS masterbatch. PP/NS

composite without SEBS was denoted as C1, and PP/NS

composites containing 5 and 10 wt% SEBS were denoted

as C2 and C3, respectively (Table 1). The samples used

for thermal and electrical characterization, square plates

100 3 100 3 0.5 mm, were prepared by hot pressing in

an electrically heated press (Dr. Collin) at 1708C for 5

min with a force of 50 kN. After compression molding,

the samples were cooled to room temperature under

pressure in a cooling cassette.

TABLE 1. Formulations of all prepared and characterized samples.

Sample PP (wt%) MA-PP (wt%) NS (wt%) SEBS (wt%)

PP 100 — — —

PPM 95 5 — —

SEBS — — — 100

C1 90 5 5 —

C2 85 5 5 5

C3 80 5 5 10

2082 POLYMER ENGINEERING AND SCIENCE—-2013 DOI 10.1002/pen

Thermal Characterization

Differential scanning calorimetry (DSC) was performed

using a DSC Q2000 from TA instruments calibrated with

sapphire standard, under nitrogen flow (100 mL/min). The

samples weighing between 12 and 14 mg were cut from

compression-molded plates, packed in aluminum pans and

placed in the DSC cell. They were cooled from ambient

temperature to 2508C and maintained at this temperature

for 2 min, and then, they were heated to 2008C at a heat-

ing rate of 108C/min. The degree of crystallinity XC was

calculated from DSC curves as follows [28]:

XC ½%� ¼DH

DH0 wPP

� 100 (1)

where DH is the heat of fusion for the composite, DH0 is the

heat of fusion for 100% crystalline isotactic PP (207 J/g [29,

30]), and wPP is the weight fraction of the polymer matrix.

In the conditions mentioned above, unambiguous Tg

values could not be obtained for references and nanocom-

posites because PP is a very fast crystallizing material

and has a week glass transition. A method proposed for

Tg evaluation in syndiotactic PP consists in quenching the

sample in liquid nitrogen, quickly transferring it in the

DSC cell and heating from 2408C at a rate of 208C/min

[30]. To avoid the differences that can appear when sev-

eral samples must be transferred in the DSC cell, we used

different conditions: the samples were heated from room

temperature to 2008C with 508C/min, held for 5 min to

erase any thermal history, and quenched to 2608C in the

same device. After equilibration, the samples were heated

at a rate of 208C/min to determine Tg. Two trials were

performed for each of the samples.

Dielectric Characterization

Dielectric spectroscopy measurements were performed

isothermally, at temperatures between 303 K and 353 K

using a Novocontrol Alpha-A Analyzer in combination with

an active sample cell ZGS. The system was allowed to reach

thermal equilibrium for 30 min at each selected temperature.

The real and imaginary parts of the relative complex permit-

tivity were determined over the frequency range from 1022

Hz to 106 Hz. The percentage error in the measurements

was found to be \2%. For best electrical contact between

samples and electrode, silver coating was used. Two parallel

measurements were performed for each of the samples.

Atomic Force Microscopy

AFM images were captured in ScanAsyst mode by a

MultiMode 8 atomic force microscope equipped with a

Nanoscope V converter from Bruker. ScanAsyst mode auto-

matically optimizes imaging parameters, including set-point,

feedback gains, and scan rate to get an optimized image.

Real-time scanning was performed in air at room tempera-

ture with the scan rates of 0.7 Hz and scan angle 908 using a

silicon tip (nominal radius 2 nm, from Bruker) with a canti-

lever length of 115 lm and a resonant frequency of �70

kHz. The images (256 3 256) were recorded and analyzed

using the AFM software NanoScope version 1.20.

Measurements by Torque Rheometer

Data concerning the melting behavior of PP nanocom-

posites as compared to that of neat PP and SEBS were

collected directly from the Brabender Lab Station mixer

using the WINMIX software. All data were collected in

the same processing conditions: temperature 1758C, shear

rate 100 rpm, and filling degree 0.75.

RESULTS AND DISCUSSION

Torque Measurements

The matrix of our composites is a polymer blend, with

PP the major component and SEBS the minor component,

and the importance of viscosity ratio on the morphological

properties of polymer blends has already been emphasized

[31, 32]. If the minor component viscosity is substantially

higher than that of the major component, high mean aver-

age particle size of the dispersed phase (SEBS) in the PP

matrix would be expected [32]. The torque–time curves

recorded during melt processing of PP, SEBS, and nano-

composites are shown in Fig. 1. All the curves, except that

of SEBS, had similar shapes: one or more loading peaks

[33] are followed by a very gentle decrease of the torque

till the end of the test, indicating little change in the melt

viscosity. A rapid drop in the torque–time curve was

recorded for SEBS. This could be caused by the degrada-

tion/oxidation processes, but the equilibration of the curve

at the end of the tested period indicated another type of

influence. We believe that this behavior is related to the

fusion process of SEBS, which was hindered by the anti-

stick coating, used for SEBS pellets. The analysis of X-ray

FIG. 1. Torque versus time curves obtained with Brabender Lab Station

for PP, SEBS, and nanocomposites.

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—-2013 2083

diffractogram of SEBS (not shown here) indicated talc as

the anti-stick agent used in this case.

Different values of the equilibrium torque measured af-

ter 10 min (T10) were obtained for nanocomposites and

references. The addition of 5 wt% NS in PP determined

an increase with 20% of T10, from 9.25 Nm for PP to

11.13 Nm for C1. SEBS showed a double T10 value than

PP, 18.83 instead of 9.25 Nm, and this difference could

result in high mean size of SEBS domains in the PP ma-

trix. However, only a slight increase of T10 with the

increase of SEBS concentration in nanocomposites was

detected: from 11.13 Nm for C1 to 11.83 Nm for C2 and

11.89 Nm for C3. The increase of SEBS loading from 5

wt% to 10 wt% had almost no influence on the torque

value, a similar behavior being reported in the case of

PP–PS blend compatibilized with 5 and 7% SEBS [34].

The low influence of SEBS on the melt viscosity of PP/

NS composites could result from the small concentration

and/or the good dispersion of SEBS in PP.

Thermal Properties

Figure 2 shows the DSC heating curves corresponding

to the melting process of nanocomposites and references.

SEBS was entirely amorphous, showing no thermal event

in the considered temperature range. PP and PPM showed

similar endothermic events, characteristic to a crystal

phase melting. The heating thermograms of composites

exhibited endothermic melting peak at a lower tempera-

ture than PP (Table 2), probably because of the decreased

lamellar thickness of the polymer crystals and the lower

perfection of crystal structure induced by NS. Broadened

endothermic peaks were obtained for C2 and C3 as com-

pared to C1, indicating lower perfection of crystal struc-

ture as a result of SEBS addition. The influence of SEBS

that tend to diffuse in the PP matrix by its ethylene-butyl-

ene (EB) midblocks and to hinder PP chains access to a

more perfect crystalline structure was reported elsewhere

[35] and could explain this behavior. Initial and final

melting temperatures (Tmi and Tmf) were calculated using

TA Instruments Universal Analysis soft to estimate the

lamellar thickness range (Table 2). Lamellar thickness (L)

was calculated using Gibbs–Thomson equation [36, 37]:

Tm ¼ T0ð1� 2s=L DHoÞ (2)

where T0 is the equilibrium melting point of a perfect

crystalline PP (460 K), the melting enthalpy DHo ¼ 184

� 106 J/m3, and the surface free energy s ¼ 0.0496 J/m2

[36, 37].

A decrease in the average lamellar thickness and a nar-

row L range were detected in NS containing composites

as compared to neat PP and PPM. Analyzing the influence

of SEBS in nanocomposites, one can observe a broader

lamellar thickness distribution with the increasing amount

of SEBS, especially for C3, in the range of thinner lamel-

lae. It can be postulated that SEBS is responsible for the

formation of smaller less ordered crystallites.

All the nanocomposites showed higher crystallinity

when compared to neat PP or PPM, probably because

of the nucleating effect of NS, which was signaled in

several publications [3, 38]. When compared with each

other, nanocomposites containing SEBS showed smaller

degree of crystallinity than C1 (without SEBS). The

decrease of PP crystallinity in PP–SEBS blends has

been signaled by other authors and explained by the

hindering effect of SEBS during the crystallization pro-

cess of PP [39].

The DSC curve of SEBS is shown in Fig. 3. One can

see a sharp glass transition event coupled with a broad

endothermic peak, both extending from 2658C to 358C.

The glass transition of the EB blocks takes place at Tg1 ¼256.18C and that of the styrenic blocks was estimated at

Tg2 ¼ 85.58C. Similar results were obtained considering

the midpoint of the heat capacity jump as Tg. The endo-

thermic event with a temperature peak of �17.58C and

around 3 J/g heat flow can be ascribed to locally ordered

EB domains. Although there are only few attempts to

determine Tg for SEBS blocks by DSC, our results are

very close to already reported Tg values determined by

DSC and other techniques [40–42]. Glass transition forFIG. 2. DSC diagrams of SEBS, PP, PPM, and nanocomposites from

408C to 1958C.

TABLE 2. DSC results for PP, PPM, and nanocomposites.

Samples

Tm 6 0.5%

(8C)

Lma

(nm)

LMa

(nm)

La

(nm)DHm 6 1%

(J/g) XC 6 1%

PP 164.6 6.5 14.7 11.1 96.7 46.7

PPM 164.4 6.5 14.5 11.0 98.6 47.6

C1 162.7 7.1 14.2 10.2 103.8 52.8

C2 163.4 7.0 14.4 10.5 94.8 50.9

C3 163.8 6.6 14.4 10.7 91.2 51.8

a Lm is the minimum lamellar thickness; LM is the maximum lamellar

thickness; L is the average lamellar thickness obtained by using the peak

temperature (Tm).


SEBS blocks in nanocomposites could not be detected by

DSC probably because of the small concentration of

SEBS (�10 wt%) in these composites.

Glass transition of PP in neat PP, PPM, and nanocom-

posites was determined from the heat capacity jump, as

illustrated in the inset of Fig. 2. For neat PP, Tg was

210618C, for PPM Tg was 27.58C and for nanocompo-

sites it was 29.08C (C1), 28.98C (C2), and 210.68C(C3), respectively. The Tg value obtained for PP is very

close to other reported results [24]. The glass transition

shifted to a higher temperature in PPM, which indicates a

restriction of the polymer chain mobility due to the misci-

bility between PP and MA-PP. The addition of NS deter-

mined a small decrease of Tg value in the limit of experi-

mental error (1.58C), but the simultaneously addition of

NS and 10 wt% SEBS has as result a noticeable decrease

of Tg (�38C) showing a shift of the glass transition tem-

perature of PP toward the lower Tg value characteristic to

EB blocks of SEBS.

Dielectric Properties

The real (e0r) and imaginary (e00r ) parts of relative com-

plex permittivity of PPM, SEBS, and PP, measured from

1022 Hz to 106 Hz, at 333 K are shown in Fig. 4. e0r(f)was almost constant on a large range of frequencies, only

a weak relaxation process close to 1 Hz, more visible in

e00r (f), was detected for all the samples. The relaxation

process could be assigned to a-relaxation coming from

the local molecular fluctuations. For PP and PPM, this

could be the ac relaxation of the crystalline phase [43–

45]. Although information regarding the frequency range

of main relaxation processes in PP and PP modified with

MA-PP in the published literature is often inconsistent,

the behavior of PPM is closed to that reported by Bohn-

ing et al. [21] on MA-PP. They assigned this relaxation

process to localized molecular fluctuations of the polar

MAH groups. The influence of polar groups in PPM and,

especially of water molecules located in the vicinity of

the MAH groups, as observed by Bohning et al. [21],

FIG. 3. DSC diagram of SEBS from 21008C to 1508C.

FIG. 4. Real er’ and imaginary e00r parts of relative complex permittivity

in function of frequency (f) for PP, PPM and SEBS measured at 333 K.

FIG. 5. Real er’ and imaginary e00r parts of relative complex permittivity

in function of frequency (f) for nanocomposites and PPM, measured at

303 K (a) and 353 K (b).


could determine the difference between PPM and PP

dielectric spectra. The dielectric behavior of SEBS is

more difficult to be explained because of the lack of in-

formation regarding molecular dynamics in this block

copolymer. Chen et al. [46] studied SEBS by dielectric

spectroscopy in a large range of temperatures and

observed a relaxation process (denoted as a0) located

between that assigned to the glass transitions of EB and

styrene blocks, in the temperature range from 283 to 323

K, close to our test conditions. They offered several

explanations for the origin of this relaxation, one related

to the molecular motions in mixed styrene–EB interphase

regions of high thicknesses and another involving a hard

domain/soft domain interfacial polarization relaxation

[46]. Local motions inside EB blocks, which probably

contain polar groups because of oxidative degradation

could be, in our opinion, at the origin of this relaxation

process. This affirmation is in line with DSC observa-

tions. The contribution of interfacial polarization should

be considered in all the samples because they are com-

mercial products and contain additives and catalyst resi-

dues being heterogeneous materials.

Figure 5a and b shows the real and imaginary parts of

relative complex permittivity measured at 303 K (Fig. 5a)

and 353 K (Fig. 5b) for the nanocomposites and PPM

(dashed line in Fig. 5). All nanocomposites have shown

higher e0 values than PPM. An explanation of this behav-

ior could be the high affinity of NS particles to water by

‘‘physical adsorption’’ and ‘‘chemisorption’’ [47]. Both the

water and silanol groups on the surface of NS modify the

dielectric properties of nanocomposites because of their

high polarity and mobility [47]. A recent study on the

effects of water absorption in nanocomposites reported

that nanocomposites absorb significantly more water than

the unfilled polymer in environmental conditions [48].

This was explained by the high nanoparticle-matrix inter-

face, which could be the preferred location for water mol-

ecules. When compared with each other, SEBS-containing

nanocomposites (C2 and C3) showed lower e0r values than

C1 (without SEBS). It seems that SEBS slowed down the

dynamic motion in PP composites probably because of

the new intermolecular interactions arising between com-

ponents. Interactions between styrenic blocks of SEBS

and MA-PP on the one hand and between EB blocks of

SEBS and PP on the other hand have already been sig-

naled in these complex systems [13].

Almost two times higher e00r values were obtained for

nanocomposites as compared to PPM, showing higher

dielectric losses that must be related to the NS addition as

mentioned above. Two relaxation processes were detected

in e00r (f) curves of nanocomposites at all tested tempera-

tures. The first was in the same frequency domain as for

the references, between 0.05 and 10 Hz when the temper-

ature increased from 303 to 353 K and is probably related

to a-relaxation of polymer components together with

interfacial polarization. At 3038C, the second relaxation

process was located close to 102 Hz for C1 and 10–50 Hz

for C2 and C3 and shifted to higher frequencies with the

increase of the temperature. Moreover, the height and

broadness of both peaks slightly changed in the tested

temperature range: the first peak became less intense and

broader and the second peak became more intense and

narrower with the increase of temperature, showing

changes in dielectric strength and in the distribution of

relaxation times. The first peak was more intense for

FIG. 6. e00r —experimental and fitted curves—in function of frequency f

for 303 K, 318 K, 333 K, and 353 K, for nanocomposites C1 (a), C2

(b), and C3 (c).


SEBS containing composites (C2 and C3) as compared to

C1, but the dielectric losses seem to be not related to the

amount of SEBS in nanocomposites, which could indicate

the preferred location of NS in PP not in SEBS.

The origin of the second relaxation process, which

appeared only in nanocomposites and was more intense

than the first a-relaxation could be related to MWS polar-

ization in heterogeneous systems due to the accumulation

of charges at the large interfaces created between

nanoparticles and the matrix. Moreover, Bohning et al.

[21] found a MWS polarization process in MA-PP/clay

nanocomposites at high temperatures ([330 K) and low

frequencies (\10 Hz).

Quantitative evaluation of the relaxation parameters

could help explaining the dielectric relaxations. The dielec-

tric loss spectra e00r (f) were fitted to the empirical Havri-

liak–Negami (HN) model function [24]:

e00r ¼ �ImDe

ð1þ ðiotHNÞaÞb

( )þ s

evo(3)

where De and tHN correspond to the relaxation strength

and the HN mean relaxation time of the relaxation

process, respectively, s is the dc-conductivity of the sam-

ple, ev is the dielectric permittivity of vacuum, and o ¼2pf, f being the frequency of the applied electric field.

The curve shape parameters, a and b (0 , a � 1, 0 , ab� 1), describe the broadness of the relaxation peak and

its asymmetry, respectively. Considering the very low dc-

conductivity of all nanocomposites (under 1015 S/cm), the

second term has no noticeable influence on the values of

e00r , that estimate, in our case, the relaxation contribution

of the imaginary part of permittivity.

Both relaxation processes were fitted using a compila-

tion of Eq. 3 [49], which is given below (Eq. 4) and the

results of the fitting are shown in Fig. 6a–c (solid lines—

experimental data, dotted lines—fitting curves according

to the HN equation):

e00 ¼ De 1þ 2ðotHNÞa cosap2

� �þ ðotHNÞ2a

h i�b=2

sin b arctgsinðap=2Þ

ðotHNÞ�a þ cosðap=2Þ

� �� (4)

A good description of the experimental data was

obtained using HN function meaning that at least four

parameters (Table 3) are needed for the complete descrip-

tion of the relaxation processes. The Havriliak–Negami

relaxation time is related to the maximum relaxation time

(smax) by the equation [50]:

tmax ¼ tHN

sin pab2ðbþ1Þ

� �sin pa

2ðbþ1Þ

� �24

35

1=a

(5)

The relaxation time data corresponding to peak max-

ima for the three nanocomposites are given in an Arrhe-

TABLE 3. HN fitting parameters for the relaxation processes for two temperatures, 303 K and 353 K.

Sample

Temperature

(K)

First relaxation Second relaxation

De sHN, (s) a b De sHN, (s) a b

C1 303 0.20 12.5 0.58 0.62 0.37 5.0 3 1024 0.50 0.70

353 0.16 5.0 3 1022 0.58 0.80 0.31 7.0 3 1026 0.51 0.68

C2 303 0.23 5.4 0.51 0.80 0.36 6.2 3 1023 0.36 0.80

353 0.27 3.6 3 1022 0.64 0.68 0.37 3.6 3 1025 0.50 0.72

C3 303 0.27 10.5 0.57 0.72 0.27 2.0 3 1022 0.37 0.75

353 0.23 0.1 0.57 0.69 0.29 5.1 3 1025 0.51 0.60

FIG. 7. Log relaxation time (smax) versus 1/T for the first (top) and sec-

ond relaxation (bottom) in nanocomposites.

TABLE 4. Values of activation energy calculated via Eq. 6 for the first

relaxation process and fitting parameters of Eq. 5 for the second

relaxation process.

Sample

First relaxation Second relaxation

s! (s) Ea (kJ/mol) s! (s) Ea (kJ/mol) TV (K)

C1 5.5 3 10215 86.2 9.9 3 10210 23.3 206.8

C2 1.2 3 10215 90.1 6.6 3 10210 28.4 206.7

C3 4.5 3 10214 81.6 1.3 3 10210 33.8 207.0


nius representation in Fig. 7. The first relaxation process

follows the Arrhenius type equation [24]:

tðTÞ ¼ t1 expEa

RT

� �(6)

where the pre-exponential factor t! is the relaxation time

at very high temperature, Ea is the ‘‘Arrhenius’’ activation

energy in kJ/mol, and R is the universal gas constant. The

dashed line in Fig. 7 represents the fitting equation. The

first relaxation process had similar behavior in nanocom-

posites as in the references, and the temperature depend-

ence of the relaxation times obeyed the Arrhenius law,

also characteristic to a-relaxation processes. The values of

Ea (Table 4) obtained for PP nanocomposites, between 80

and 90 kJ/mol, are close to the activation energy of the

a-relaxation process of the crystalline PP phase measured

by other methods, usually from 90 to 170 kJ/mol [20].

Higher activation energy and lower relaxation time were

obtained for C2 as compared to the other two nanocom-

posites. Many factors could influence this behavior related

to the degree of crystallinity, the proportion of the defects

in the crystalline phase, and NS and SEBS dispersion and

location in the matrix and others. Sato reported a shift to

longer time of the relaxation peak with an increase in

crystallinity in polyethylene [51]. The lowest degree of

FIG. 8. AFM images (height, phase, and peak force error from left to right) of C1 (a), C2 (b), and C3 (c):

scan area 5 lm 3 5 lm.


crystallinity from all the composites was obtained in the

case of C2, as shown by DSC results (Table 2). This

lower crystallinity could contribute to the shift to smaller

relaxation time in analogy with Sato results. The slight

increase of the activation energy is not a consequence of

increased crystallinity in C2 than in C1 or C3 as shown

by DSC results and could be an effect of the better

dispersion of NS and SEBS in the matrix, the higher

degree of structural homogeneity enhancing the resistance

to the transport of charge carriers through the polymer

matrix [28].

The Arrhenius plot of the second relaxation process

was significantly bent for all nanocomposites, which is a

typical feature of relaxations that are related to glass

transition. The temperature dependence of the second

peak relaxation times was fitted by the Vogel–Fulcher–

Tammann (VFT) equation [24]:

tðTÞ ¼ t1 expEV

RðT � TVÞ

� �(7)

where EV is the ‘‘Vogel’’ activation energy in kJ/mol, and

TV is the extrapolated (Vogel) glass temperature. The

corresponding fitting parameters are shown in Table 4.

They are in good agreement with previously reported

results for neat PP [24, 52–53] and PP blends and compo-

sites [21, 44], although the values of these parameters

reported in the literature are rather different. Sengers

FIG. 9. AFM images (height, phase, and peak force error from left to right) of C1 (a), C2 (b), and C3 (c

and d): scan area 3 lm 3 3 lm.


et al. [24] obtained for neat PP and its blends with poly-

styrene activation energies ranging from 22.5 to 31.6 kJ/

mol. Ideal glass transition temperatures were found to be

with 55–57 K below Tg values determined by DSC, in

agreement with previously reported results [54, 55].

The second process has the features of a b relaxation

because the temperature dependence of the relaxation

times obeyed the VFT equation. However, the high

dielectric losses obtained for a relaxation associated with

glass rubber transition in the case of a polymer with low

concentration of dipoles, like PP, is uncommon. Several

works emphasized the particular influence of water on

the dielectric properties of polymer nanocomposites. Lau

et al. [48] studied the effect of water absorption in poly-

ethylene–NS composites, and their results suggest that

water molecules can act as effective dielectric probes,

enhancing and/or shifting the structural relaxation. More-

over, the physically adsorbed water on the NS surface

cannot be completely removed even at 6008C [47],

which is well above the melt processing temperature of

our nanocomposites. Therefore, the peculiar behavior

related to the second relaxation process can be explained

by the high concentration of polar groups from the water

and silanol groups brought in our nanocomposites on the

NS surface. On the other hand, considering MWS polar-

ization together with the dynamical behavior of the sur-

rounding polymer phase at the origin of the second

relaxation process, an explanation for this uncommon

behavior could be provided by a particle-core/polymer-

shell configuration. Further work is clearly needed to

elucidate the origin of this process.

Atomic Force Microscopy

Atomic force microscopy images (scan area 5 lm 3 5

lm) of nanocomposites are shown in Fig. 8a–c (height,

phase, and peak force error images from left to right). NS

is relatively homogenous dispersed in C1 and C2, as

detected in Fig. 8a and b. NS is easily observed as

brighter dots because of its higher stiffness compared to

that of the polymer matrix. It is important to note that

dark dots are observed near NS particles in phase images,

indicating a soft material that could be the compatibilizer,

MA-PP. The crystalline structure of the matrix is hardly

visible at this magnification, but similar features can be

detected for both C1 and C2. Different surface character-

istics are detected for C3 in Fig. 8c. Larger zones with a

different texture as compared to that of PP from the pre-

vious images can be observed in this case. They could be

ascribed to SEBS because of their wormlike structure,

which was mentioned in several studies on TEM analysis

of SEBS and other block copolymers [16, 56, 57].

To get some insight into the influence of SEBS and NS

on the morphology of PP, samples were investigated by

AFM using higher magnification (Fig. 9a–d). The crystal-

line structure of PP can be observed in Fig. 9a (C1) over

the entire analyzed surface, on large areas in Fig. 9b (C2)

and in the bottom of Fig. 9c (C3). The lamellar structure

of PP, which is observed in Fig. 9a (height, phase, and

peak force error images from left to right), has the charac-

teristics of a-form PP and consists of cross-hatched daugh-

ter and mother lamellae with a thickness of 10–14 nm, in

accordance with the lamellar thickness calculated from

DSC and with other reported results [28]. Smaller lamellae

cannot be detected with this AFM mode (ScanAsyst). The

cross-hatched structure of PP can be detected on wider

areas in the case of C2 (Fig. 9b) having similar lamellar

thickness as in the case of C1. Several areas seem to be

covered by a thin layer of SEBS, which determines the

change of morphological features. On these areas, the den-

sity of NS is lower, suggesting the preferential dispersion

of NS in PP. In the case of C3, the lamellar structure of

PP can be detected in the bottom of Fig. 9c in all the

images (height, phase, and peak force error). It is a highly

oriented structure, probably a result of the high tensile

forces underwent by PP during the cooling of compressed

films because of higher contraction of the surrounding

areas rich in SEBS. In the top of these images, a zone with

worm-like structure, rich in SEBS, must be noted. NS is

less visible in these images probably because of the layer

of SEBS detected on the surface of C3. It is interesting to

observe the interfacial boundary that exists between the

two phases, PP and SEBS, in Fig. 9c. A detailed image of

this interface is shown in Fig. 10. It can be seen that the

transition from a phase rich in SEBS (in the top of the

image) to the one containing mainly PP (in the bottom) is

smooth, without disturbances. This suggests good interac-

tion between phases probably because of similar chemical

structure of PP and EB midblock of SEBS [58, 59]. Setz

et al. [13] found that EB midblock of SEBS tends to dif-

fuse into the PP matrix under formation of small micelles.

FIG. 10. AFM image (peak force error) of C3—scan area 1.5 lm 3

1.5 lm.


CONCLUSIONS

Combined effects of SEBS and NS on PP thermal and

dielectric properties in correlation with morphological

aspects were investigated. All nanocomposites showed

higher crystallinity when compared to PP. SEBS hindered

the crystallization process of PP in PP/NS nanocomposites,

leading to a smaller degree of crystallinity and lower perfec-

tion of crystal structure, in accordance with AFM observa-

tions. Broader lamellar thickness distribution was obtained

in nanocomposites containing SEBS, especially in the com-

posite with the larger amount of elastomer. Small difference

was detected by DSC between the Tg value of neat PP and

that of PP in nanocomposites with or without SEBS, only

the simultaneously addition of NS and 10 wt% SEBS result-

ing in a decrease of �38C. Almost two times higher dielec-

tric losses were obtained for nanocomposites as compared

to PP reference probably because of the high polarity and

mobility of water and silanol groups on the surface of NS,

being known that nanocomposites absorb significantly more

water than the unfilled polymer. Two relaxation processes

were detected in e00r (f) curves of nanocomposites at all tested

temperatures, and the empirical Havriliak–Negami model

function was used to fit dielectric data. The first relaxation

was in the same frequency domain as for the references and

was assigned to a-relaxation of polymer components to-

gether with interfacial polarization. The temperature

dependence of the relaxation times obeyed the Arrheniuslaw, and the activation energy obtained for PP nanocompo-sites, between 80 and 90 kJ/mol, was close to that reportedfor the a-relaxation process of PP measured by other meth-ods. MWS polarization was expected to determine thesecond relaxation process. Surprisingly, the temperature de-pendence of the relaxation times obeyed the VFT equation,

which is a characteristic of structural relaxation related to

glass transition. Likewise, the high dielectric losses obtained

for a relaxation associated with Tg in the case of a

polymer with low concentration of dipoles, like PP, is

uncommon. An explanation was given considering the

high concentration of polar groups from the water and

silanol groups brought on the NS surface but further

investigations are needed to elucidate the origin of this

process. The incorporation of SEBS induced several

changes in the dielectric response of the composites,

which support its tendency to hinder the motional proc-

esses of the matrix also observed by DSC. A smooth tran-

sition from a phase rich in SEBS to one containing mainly

PP was detected in the AFM image of PP/NS composite

containing the larger amount of SEBS, emphasizing the

good compatibility at PP/SEBS interface.

REFERENCES

1. M.F. Omar, H.M. Akil, and Z.A. Ahmad, Polym. Compos.,32, 565 (2011).

2. A. Haghtalab and R. Marzban, Adv. Polym. Technol., 30,

203 (2011).

3. H.J. Zhou, M.Z. Rong, M.Q. Zhang, W.H. Ruan, and K.

Friedrich, Polym. Eng. Sci., 47, 499 (2007).

4. O.H. Lin, H.M. Akil, and Z.A.M. Ishak, Polym. Compos.,32, 1568 (2011).

5. F. Zoukrami, N. Haddaoui, C. Vanzeveren, M. Sclavons,

and J. Devaux, Polym. Int., 57, 756 (2008).

6. D.N. Bikiaris, G.Z. Papageorgiou, E. Pavlidou, N. Vourout-

zis, P. Palatzoglou, and G.P. Karayannidis, J. Appl. Polym.Sci., 100, 2684 (2006).

7. D.M. Panaitescu, Z. Vuluga, C. Radovici, and C. Nicolae,

Polym. Test., 31, 355 (2012).

8. Y. Liu and M. Kontopoulou, Polymer, 47, 7731 (2006).

9. K. Chatterjee and K. Naskar, Polym. Eng. Sci., 48, 1077

(2008).

10. A. Bassani and L.A. Pessan, J. Appl. Polym. Sci., 88, 1081

(2003).

11. F. Stricker, Y. Thomann, and R. Mulhaupt, J. Appl. Polym.Sci., 68, 1891 (1998).

12. F.O.M.S. Abreu, M.M.C. Forte, and S.A. Liberman, J. Appl.Polym. Sci., 95, 254 (2005).

13. S. Setz, F. Stricker, J. Kressler, T. Duschek, and R.

Mulhaupt, J. Appl. Polym. Sci., 59, 1117 (1996).

14. Matsuda Y., Hara M. T. Mano, K. Okamoto, and M.

Ishikawa, Polym. Eng. Sci., 45, 1630 (2005).

15. H. Mae, M. Omiya, and K. Kishimoto, J. Solid Mech.Mater. Eng. 2, 254 (2008).

16. F. Kremer and A. Schonhals, Eds., Broadband DielectricSpectroscopy, Springer-Verlag, Berlin (2003).

17. Z-M. Dang, J-K. Yuan, J-W. Zha, T. Zhou, S-T. Li, and

G-H. Hu, Prog. Mater. Sci., 57, 660 (2012).

18. K. Kessairi, S. Napolitano, S. Capaccioli, P. Rolla, and M

Wu1bbenhorst, Macromolecules, 40, 1786 (2007).

19. A. Gitsas and G. Floudas, Macromolecules, 41, 9423

(2008).

20. E. Suljovrujic, Polym. Bull., 68, 2033 (2012)

21. M. Bohning, H. Goering, A. Fritz, K.-W. Brzezinka, G.

Turky, A. Schonhals, and B. Schartel, Macromolecules, 38,

2764 (2005).

22. T. Furukawa, H. Yuruzume, T. Kimura, and S. Matsuura,

Proc. 8th International Symposium on Electrets, Dielectric

investigation of polypropylene grafted with maleic anhy-

dride, 450 (1994).

23. C. Pandis, E. Logakis, 1 V. Peoglos, P. Pissis, M. Omastova,

M. Mravcakova, A. Janke, J. Pionteck, Y. Peneva, and L.

Minkova, J. Polym. Sci. Part B: Polym. Phys., 47, 407 (2009).

24. W.G.F. Sengers, O. van den Berg, M. Wubbenhorst, A.D.

Gotsis, and S.J. Picken, Polymer, 46, 6064 (2005).

25. E. Laredo, M. Grimau, A. Bello, F. Sanchez, M.A. Gomez,

C. Marco, I. CampoY, and J.M. Arribas, J. Polym. Sci. PartB: Polym. Phys., 43, 1408 (2005).

26. Y. Wang and S-W. Huang, Polymer-Plastics Techn. Eng.,46, 1039 (2007).

27. S. George, K.T. Varughese, and S. Thomas, J. Appl. Polym.Sci., 73, 255 (1999).

28. A. Mirzadeh, P.G. Lafleur, M.R. Kamal, and C. Dubois,

Polym. Eng. Sci., 50, 2131 (2010).

29. S. Sahin and P. Yayla, Polym. Test., 24, 1012 (2005).


30. P. Supaphol and J.E. Spruiell, J. Appl. Polym. Sci., 75, 44

(2000).

31. N. Furgiuele, A.H. Lebovitz, K. Khait, and J.M. Torkelson,

Polym. Eng. Sci., 40, 1447 (2000).

32. M. Hemmati, H. Nazokdast, and H.S. Panahi, J. Appl.Polym. Sci., 82, 1129 (2001).

33. B. Cheng, C. Zhou, W. Yu, and X. Sun, Polym. Test., 20,

811 (2001).

34. W. Brostow, T. H. Grguric, O. Olea-Mejia, V. Rek, and J.

Unni, e-Polymers, 33 1–9 (2008).

35. S.C. Tjong, S.P. Bao, and G.D. Liang, J. Polym. Sci. PartB: Polym. Phys., 43, 3112 (2005).

36. V. Virkkunen, P. Laari, P. Pitkanen, and F. Sundholm,

Polymer, 45, 4623 (2004).

37. J. Kang, F. Yang, T. Wu, H. Li, D. Liu, Y. Cao, and M.

Xiang, J. Appl. Polym. Sci., 125, 3076 (2012).

38. G.Z. Papageorgiou, D.S. Achilias, D.N. Bikiaris, and G.P.

Karayannidis, Thermochim. Acta, 427, 117 (2005).

39. Z. Ahmad, K.D. Kumar, M. Saroop, N. Preschilla, A. Bis-

was, J.R. Bellare, and A.K. Bhowmick, Polym. Eng. Sci.,50, 331 (2010).

40. R-M. Ho, A. Adedeji, D.W. Giles, D.A. Hajduk, C.W.

Macosko, and F.S. Bates, J. Polym. Sci. Part B: Polym.Phys., 35, 2857 (1997).

41. G.S. Rajan, K.A. Mauritz, S.L. Stromeyer, T. Kwee, P.

Mani, J.L. Weston, D.E. Nikles, and M. Shamsuzzoha,

J. Polym. Sci. Part B: Polym. Phys., 43, 1475 (2005).

42. F. Picchioni, M. Aglietto, E. Passaglia, and F. Ciardelli,

Polymer, 43, 3323 (2002).

43. W-G. Hu and K. Schmidt-Rohr, Acta Polym., 50, 271

(1999).

44. S.A. Madbouly and J.U. Otaigbe, Polymer, 48, 4097

(2007).

45. E. Suljovrujic, S. Trifunovic, and D. Milicevic, Polym.Degrad. Stab., 95, 164 (2010).

46. H. Chen, M.K. Hassan, S.K. Peddini, and K.A. Mauritz,

Eur. Polym. J., 47, 1936 (2011).

47. L. Peng, W. Qisui, L. Xi, and Z. Chaocan, Colloids Surf A:Physicochem. Eng. Aspects, 334, 112 (2009).

48. K.Y. Lau, A.S. Vaughan, G. Chen, and I.L. Hosier, IEEEConference on Electrical Insulation and Dielectric Phenom-ena, Dielectric response of polyethylene nanocomposites:

the effect of surface treatment and water absorption, Mon-

treal, 275 –278 (2012).

49. A. Schonhals and F Kremer, in Broadband Dielectric Spec-troscopy Cap. 3: Analysis of Dielectric Spectra, F. Kremer

and A. Schonhals, Eds., Springer Verlag, Berlin 59 (2003).

50. A. Nalawade, M.K. Hassan, W.A. Jarrett, K.A. Mauritz, and

M.H. Litt, Polym. Int., 61, 55 (2012).

51. Y. Sato, J. Appl. Polym. Sci., 26, 27 (1981).

52. P.G. Santangelo, K.L. Ngai, and C.M. Roland, Macromole-cules, 29, 3651 (1996).

53. O. van den Berg, W.G.F. Sengers, W.F. Jager, S.J. Picken,

and M. Wubbenhorst, Macromolecules, 37, 2460 (2004).

54. P. Voudouris, N. Gomopoulos, A. Le Grand, N. Hadjichris-

tidis, G. Floudas, M.D. Ediger, and G. Fytas, J. Chem.Phys., 132, 074906 (2010).

55. K. Fukao and Y. Miyamato, Phys. Rev. E Stat. Nonlin. Soft.Matter Phys. 64 (1 Pt.1): 011803 (2001).

56. T-I Yang and P. Kofinas, Polymer, 48, 791 (2007).

57. M.A. Paglicawan, J.K. Kim, and D-S Bang, Polym. Com-pos., 31, 210 (2010).

58. M. Denac, I. Smit, and V. Musil, Composites: Part A, 36,

1094 (2005).

59. O. Balkan, A. Ezde, and H. Demirer, Polym. Compos., 31,

1265 (2010).


The effect of poly[styrene- b -(ethylene- co -butylene)- b -styrene] on dielectric, thermal, and morphological characteristics of polypropylene/silica nanocomposites

Documents