The influence of gamma radiation on the dielectric relaxation behaviour of isotactic polypropylene: The α relaxation

Polymer Degradation and Stability 95 (2010) 164e171

Contents lists available at ScienceDirect

Polymer Degradation and Stability journal homepage: www.elsevier.com/locate/polydegstab

The influence of gamma radiation on the dielectric relaxation behaviour of isotactic polypropylene: The a relaxation E. Suljovrujic*, S. Trifunovic, D. Milicevic Vinca Institute of Nuclear Sciences, P.O. Box 522, 11001 Belgrade, Serbia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 13 October 2009 Received in revised form 9 November 2009 Accepted 18 November 2009 Available online 26 November 2009

The high-temperature a relaxation in gamma irradiated isotactic polypropylene (iPP) was studied over the temperature (298e406 K), frequency (103e106 Hz) and absorbed dose (0e700 kGy) ranges by means of dielectric spectroscopy. The multiple a relaxation was resolved from the b relaxation by curve fitting and its parameters were determined. Its position, intensity and activation energy were found to be strongly dependent on the changes in the structural and morphological parameters attributed to the exposure of the samples to radiation. Wide angle X-ray diffraction (WAXD) was used to investigate radiation-induced changes in the crystalline structure and degree of crystallinity, since this relaxation is connected with the crystal phase. Infrared (IR) spectroscopy and gel measurements were used to determine the changes in the oxidative degradation and the degree of network formation, respectively; the polar (carbonyl and/or hydroperoxide) groups that were introduced by irradiation were considered as tracer groups. Conclusions derived according to different methods were compared. The results reveal uncommon a relaxation behaviour with gamma radiation and confirm the multiple nature of this process, together with high dielectric and/or relaxation sensitivity of iPP to the radiation-induced changes. Ó 2009 Elsevier Ltd. All rights reserved.

Keywords: Polypropylene Relaxation Gamma radiation Dielectric spectroscopy Polar groups

1. Introduction Polypropylene (PP) is a thermoplastic polymer which belongs to the family of polyolefins. It has excellent mechanical and dielectric properties and therefore a wide variety of industrial applications, including electrical ones. Due to the low dielectric loss and good heat resistance it has been widely used as electrical insulation, e.g. for cables and as a dielectric in power capacitors [1e3]. Considering the molecular structure of non-polar hydrocarbon polymers (such as polyethylene, polypropylene, etc.), the dipole moments of the (CeH apolar) groups contained in these polymers are very low (in the order of 0.1 Debye) and hardly detectable by the usual dielectric techniques [4]. Despite this, apolar polymers exhibit measurable dielectric spectra corresponding to the transitions measured by the mechanical relaxation techniques. The measurable dielectric relaxations and losses are generally ascribed to impurities and to the fact that these polymers are always slightly oxidized and thus contain polar carbonyl, peroxy or hydroperoxy groups. Among impurities, residual catalysts and antioxidants have been reported to affect the dielectric properties. However, for the electrical

* Corresponding author. Tel./fax: þ381 11 3408 607. E-mail address: [email protected] (E. Suljovrujic). 0141-3910/$ e see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.polymdegradstab.2009.11.034

application of such polymers it is of essential interest to understand the dielectric phenomena in them. Furthermore, dielectric measurements can give valuable information about the structure and dynamics of materials. It is well known that the dielectric response can be used as an indicator of condition and ageing occurring in polymer insulation [5]. The mechanical and especially dielectric relaxation of PP have been relatively less investigated than in the case of polyethylene (PE). The nature and origins of the molecular relaxations in PP were studied mainly by mechanical and nuclear magnetic measurements and less by dielectric spectroscopy. The reality of the basic relaxation processes is more or less clear and the general trends are known. Despite this fact, the molecular origin and the morphological assignment of each molecular relaxation are still a matter of discussion because of the variety of results and interpretations proposed by different authors. Owing to the literature data [6e22], the relaxation spectrum of virgin iPP as a function of temperature is well known; it exhibits four relaxation regions, which, for the sake of simplicity and practical reasons, we label as a, b, g and d, in the order of decreasing temperature. The a process is localized below the melting temperature. Its position varies with many factors (structure, density, thermal history, etc.). The origin of this process is controversial in the literature, but it undoubtedly requires the presence of a crystalline phase. According to Jourdan et al. [6], it is

E. Suljovrujic et al. / Polymer Degradation and Stability 95 (2010) 164e171

due to a relaxation of defects in the crystalline phase, but the amorphous phase in the neighbourhood of the crystallites also contributes to the a process. Multiple nature of the mechanical a relaxation, consisting of two components, was also observed [16,18]. It was found by Pluta and Kryszewski [18] that the morphology and structure differentiation significantly influence the nature and the number of components of the mechanical a relaxation. The presence of a smectic phase as well as the decrease of both the spherulite size and the structure perfection lead to enhancement of the mobility of crystallites, and consequently to an increase of the contribution from the intra-lamellar regions to the a relaxation. Two clearly evident components of the mechanical a relaxation were found for non-spherulitic samples, crystallized from the glassy state. The low-temperature component is mainly attributed to the stress relaxation of the fraction of the non-crystalline phase containing strained molecules and segments of molecules belonging to the specific (irregular) arrangement of the surface layer of the crystallites. The high-temperature component of the a relaxation is connected with the viscous slip of the crystalline elements within the non-crystalline phase. The intensity of this process increases with the decrease of crystallite size. The presence of the smectic phase in the non-spherulitic structure enhances the extent of that process. Furthermore, an impressive study of the mechanical a relaxation as a two-component process was made by Tiemblo and collaborators for the case of iPP exposed to thermo-oxidation [15,16]; they proposed that the a relaxation makes it possible for crystals to act as radical scavengers and thus stabilize iPP exposed to thermo-oxidation at temperatures under 115  C. On the other hand, the dielectric a relaxation is almost unexplored. Actually, very few experimental results are available, and most of them are made by TSDC measurements applied to investigate charge traps in crystalline parts [23,24]. Comparison with the a relaxation in PE was made, too [25]. Boyd and Mansfield have proposed that the dielectric a process in PE can be represented by the propagation of a twisted defect along the chain within crystal lattice, leading to reorganization of the crystal surface [26]. On the other hand, the mechanically active a process in PE, although it requires the presence of the crystal phase, has the relaxation strength assigned to the amorphous component and involves softening or deformation of the latter. However, in the case of PP the relaxation times for the a process follow the Arrhenius law due to less cooperative motions than in the case of the glass rubber transition, i.e. b relaxation. Activation energies for this process usually range from 90 to 170 kJ mol1 [6,20], but much higher values (up to 350 kJ mol1) are also reported [7]. The intermediate b peak situated between 250 and 300 K has been attributed to the glass transition of the amorphous phase. Due to the broadness of this relaxation, some authors proposed two transitions involving unconstrained regions of the amorphous phase and regions constrained by crystallites (which may depend on the crystallite size [20]) [6,7]. The VogeleTammanneFulcher (VTF) or its equivalent, the WilliamseLandeleFerry (WLF) temperature dependence observed for this relaxation indicates cooperative behaviour related to the glass transition with the reported activation energies of 350e400 kJ mol1 [6,20,27]. The g process, according to different literature data and conditions, appears between 150 and 230 K; this relaxation is usually assigned to the local (most probably crankshaft type) motions in the amorphous phase [10,17,18]. The dynamic mechanical investigation has indicated that the initiation of thermal oxidation is concomitant with a partial vanishing of the g relaxation [15e17]. In dielectric relaxation measurements, iPP may also exhibit a fourth relaxation, mainly below 100 K, which is named the d process; this relaxation is weak or absent and is attributed to the hindered rotation of CH3 groups [8,9]. An Arrhenius temperature dependence is observed for low-temperature

165

relaxations and reported activation energies for the g and d relaxations are about 25 kJ mol1 and 5 kJ mol1, respectively [9]. The molecular structure and macroscopic properties of polymers can be significantly modified by ionising radiation. For the case of polypropylene the interest in such modifications lies in its numerous applications [28e31]. In general, industrial applications of radiation processing of plastic include polymerization, grafting, crosslinking and degradation. In the wire and cable industry a major application of high-energy radiation is crosslinking of insulation; crosslinking to a gel content of 55% was shown to be beneficial to cable insulation [32]. By linking the macromolecules into a network, the toughness, impact resistance, chemical resistance, and working temperatures can be considerably improved [33]. It is well known that the interaction of gamma rays with polymeric materials gives rise to free radicals which can stabilize in several ways. The main molecular effects are chain scission, chain branching and crosslinking. Usually, all these effects take place simultaneously and depend on different parameters, such as the chemical structure and morphology of the polymer, as well as the experimental irradiation conditions, including the post-treatment [34e37]. The chemical structure affects the evolution of free radicals towards stable species. Thus, polypropylene is largely destroyed by radiation in the presence of oxygen, as a consequence of the abundance of tertiary carbon or hydrogen atoms in its structure that favour scission of the main chain [38]. The morphology is also important in determining the final molecular structure of irradiated polymers where amorphous or more disordered structures undergo chain branching and crosslinking reactions more easily [35]. In fact, it is generally accepted that in a semicrystalline polymer the amorphous interfacial structure makes a major contribution to gel formation. The mechanical entanglement of the amorphous interfacial regions between different crystallites may enhance the possibility of intermolecular crosslinking. In general, the major effect of irradiation, either electron beam or gamma rays, on the crystalline region are some imperfections [39]. The macromolecules have very small mobility and the oxygen is almost unable to diffuse in crystalline regions; diffusion constants for crystalline regions are small, 8e9 orders of magnitude smaller than in the amorphous region [40]. Because of that, radiation-induced processes take place mostly in the amorphous region and on boundary layers. Among experimental irradiation conditions a key role can be played by the irradiation parameters, i.e. the total absorbed dose and the irradiation dose rate, because they can affect the concentration of the reactive species and consequently the kinetics of the reactions involved [36]. Therefore, radiation-induced structure modifications can be useful tools not only for applications, but also for highlighting some fundamental processes and properties of polymers. Although the effects of radiation on PP have received considerable attention in the past, the effects of radiation on molecular relaxations and especially dielectric relaxation behaviour have not been investigated to appreciable extent. Actually, very few experimental results are available about dielectric relaxations in polypropylene exposed to ionizing radiation [41e43]. In the previous work [19] the response of iPP to gamma radiation in different media (air, deionised distilled (DD) water and acetylene) was studied. Furthermore, the evolution of lowtemperature dielectric relaxations with gamma-irradiation was investigated, too. The connection between the oxidative degradation and dielectric properties was well established. The amount of carbonyl, hydroperoxide and other polar groups is much higher for the irradiation in air than in other media, leading to much higher dielectric losses in this medium. Complete “vanishing” of the g relaxation in iPP samples irradiated in air is connected with a large radiation-induced oxidative degradation in this medium and is reported for the first time.

E. Suljovrujic et al. / Polymer Degradation and Stability 95 (2010) 164e171

A plot of dielectric loss versus temperature for virgin (quenched) iPP, at two different frequencies (100 kHz and 1 MHz) and in a wide temperature range (50e406 K), is shown in Fig. 1a. The a and b relaxation transitions are seen as prominent loss peaks, while the other two less visible peaks at lower temperatures are assigned to the g and d transitions. The b process is well resolved from the g relaxation at moderate frequencies. However, this is not true with respect to the a process. The a process occurs at a higher temperature than the b, but its activation energy is lower. This means that the a and b processes become better resolved in isochronal scans as frequency increases. Furthermore, while the magnitude of the b process decreases with frequency, the magnitude of the a process shows opposite behaviour. Taking in consideration all these facts,

4

a

100 kHz

2

tanδ [10-4]

0

1 MHz 2 0 100

200

300

400

T[K] 4

b

4

300

c

350

4

1. 2. 3. 80 4. 5.

1 MHz T =350 K

300

d

350

e

0 kGy 25 kGy 100 kGy 300 kGy 700 kGy

5. 4.

60

400

2 0

100

T =330.5 K

2 0

100 kHz

3. tanδ [10-4]

The polymer used in this study was a stabilized iPP, HIPOL MA2CR type C-7608 (Mw ¼ 136 000, Mw/Mn ¼ 4.95, r ¼ 0.906 g/cm3). Isotropic sheets were prepared by 20 min compression moulding in a Carver laboratory press at 190  C and gradual increment pressure up to 3.28 MPa. The moulded sheets were quenched in an ice-water mixture. The samples were wrapped in Al-foil and irradiated in a 60Co radiation facility, in air, at room temperature, at a dose rate of 9 kGy/h, to various absorbed doses up to 700 kGy. Furthermore, it should be emphasized that irradiated samples were annealed (at 95  C) for 3 h in air, in order to minimize post-irradiation oxidation. The annealed irradiated samples, all 0.28  0.02 mm thick, were used for further investigations. Both pristine and irradiated sample surfaces were studied by Attenuated Total Reflectance Infrared spectroscopy (FT-IR/ATR). This technique functions by passing a radiation beam through a crystal made of a high-refractive index infrared-transmitting material, which is then totally internally reflected at the surface. The sample is brought into contact with the totally reflecting surface of the ATR crystal; the evanescent wave is attenuated in regions of the infrared spectrum where the sample absorbs energy. FT-IR/ATR spectra were recorded at room temperature in the wavenumber range of 4000e400 cm1 on a NICOLET PC FT-IR spectrometer (Model 380), with 4 cm1 resolution. The absorbance at 1715 cm1 was determined from these spectra. The carbonyl index was defined as the ratio of A1715(D)/A1715(0), in which A1715(D) and A1715(0) are the absorbencies at 1715 cm1 for different absorbed doses and before gamma-irradiation, respectively. Wide-angle X-ray diffractograms of the samples were obtained using a Bruker D8 Advance Diffractometer (in normal mode, with Cu Ka emission). The parallel beam optics was adjusted by a parabolic Göbel mirror (push plug Ni/C) with horizontal grazing incidence soller slit of 0.12 and LiF monochromator; diffractometer scans were taken in the angular range of 2q ¼ 10 e45 , at a step of 0.02 , with 10 s exposition per step. Furthermore, crystallinity was evaluated from diffraction curves by resolving multiple peak data into individual crystalline peaks and an amorphous halo. Quantitative analysis and fitting of multiple peaks in experimental spectra were performed using standard software with asymmetric pseudoVoigt functions. For the gel measurements, the samples were inserted into a 200 mesh stainless steel cloth and immersed in xylene with 0.5 wt.-% antioxidant (Irganox 1010). The gel content was determined by the measurement of weight loss of the samples after solvent extraction in the boiling xylene for 17 h, followed by drying the samples for 4 h in a vacuum oven at 60  C. Presented results are average values of

3. Results and discussion

tanδ [10-4]

2. Experimental

five identically prepared samples. Furthermore, sol-gel calculations are based on the CharlesbyePinner (CeP) expressions. The dielectric loss tangent (tan d) of the samples in the form of discs (1.3 cm in diameter) was measured on a Digital LCR Meter 4284 A, as a function of temperature (298e406 K) and in the frequency range 103e106 Hz. Dielectric measurements were taken at increments of approximately 2 K during a heating run from 298 to 406 K, with a heating rate of 1.7 K/min between equilibrated temperatures. At each equilibrated temperature, measurements of capacitance and tan d were taken at several frequencies from 1 kHz to 1 MHz; data acquisition over the frequency range required about 5 min.

tanδ [10-4]

The aim of this work is to investigate the high-temperature dielectric relaxation behaviour in virgin and gamma irradiated iPP. In most electrical applications, exploitation temperatures coincide with the a relaxation zone and information about this process can play a significant role. Dielectric relaxation spectroscopy (DRS) and gamma radiation were used as powerful methods for characterization and modification of structure, respectively. In the case of dielectric relaxation measurements, the polar (mainly carbonyl and hydroperoxide) groups that were introduced by radiation-induced oxidation in apolar iPP were considered as tracer groups whose motion reflected the motion of the polymer chains. A variety of supplementary measurements were made to determine radiationinduced changes in the structure and morphology. Results obtained by WAXS, IR and gel measurements were compared with the changes in (the intensity, position and activation energy of) the dielectric a relaxation and some interesting and uncommon new features were observed.

tanδ [10-4]

166

40

2.

20

400 4

1b. annealed

1 MHz T =330 K 2

2

1a. 1 MHz

0 300

350

T[K]

400

0 300

350

400

T[K]

Fig. 1. (a) Dielectric loss tangent versus temperature for virgin iPP sample at two different frequencies (100 kHz and 1 MHz); Experimental dielectric loss tangent spectrum (,), fitted dielectric spectrum (solid line) and fitted components (dash and dash-dot lines) of dielectric spectrum (b (Tg), a1, a2 and a3) for two different frequencies: 100 kHz (b) and 1 MHz (c and d); (e) dielectric loss tangent versus temperature, at f ¼ 1 MHz, for iPP samples irradiated in air to different absorbed doses.

E. Suljovrujic et al. / Polymer Degradation and Stability 95 (2010) 164e171

380

a

60

b

360

T [K]

tan [10-4]

90

30

max

max

340

100 kHz 1 MHz

100 kHz 1 MHz 0

0

300

Absorbed dose [kGy]

600

320

0

300

600

Absorbed dose [kGy]

Fig. 2. The intensity (a) and position (b) of the dielectric a relaxation as a function of absorbed dose measured at two different frequencies: 100 kHz and 1 MHz.

a

5.

5.

4.

4.

3.

3.

2.

2.

1.

1.

3600 3200 2800

c

1715 1640

A1715(D)/ A1715(0)

3340

30

b

Absorbance

objective values for the temperatures of the relaxation peaks were obtained using curve fitting. This analysis is more complicated at higher frequencies despite the fact that the a process is shifted further from the b process. Namely, while the a process at 100 kHz can be successfully fitted by one peak (Fig. 1b), at higher frequencies it is necessary to use two or even three peaks to fit the experimental dielectric loss tangent spectrum (Fig. 1c and d). Multiple nature of the a process is even more pronounced in the case of an annealed quenched iPP sample (Fig. 1e); although the dielectric a process is much broader in relaxation times in the case of annealed sample, the central positions for this relaxation are very similar to those in the virgin (quenched) sample. Radiation additionally complicates this situation. By comparing the dielectric loss scans, for the virgin and irradiated samples (Fig. 1e), it can be observed that the radiation significantly changes the a relaxation zone; the changes in the magnitude and position are more than evident and are presented in Fig. 2a and b, respectively. The radiation-induced oxidation introduces carbonyl and/or hydroperoxide groups as statistically distributed tracer groups whose motion reflects the motion of the polymer chains. For this reason, the dielectric a relaxation monitors the changes in the polar content caused by the radiation-induced oxidation, but it arises only from the polar groups which are connected with the molecules contributing to this relaxation. Thus, some connections between the radiation-induced oxidation and the magnitude of the dielectric relaxation can be established. Exposure of iPP films to radiation in air results in a significant increase in oxidation and the formation of hydroperoxides and/or carbonyls as major products [44]. The mechanism commonly invoked in the radiation-induced oxidation of PP is the attack of free radicals on the polymer chains. The alkyl radicals generated in this way react with oxygen, giving rise to hydroperoxides, alcohols, and carbonyl compounds such as carboxylic acids, ketones, and esters, whose relative percentage depends on oxidation conditions [34,45]. The quantification of the various oxidation products of PP (as a result of gamma, photo, and thermal oxidation) has been done by Lacoste et al. [46,47]. The major oxidation products in irradiated iPP are the tertiary hydroperoxides, which are unstable and undergo radical decomposition to other more stable oxidative products with an increase in radiation dose and their concentration. This is a major source of carbonyl compounds, especially at higher doses. The radiation-induced modifications that occur in the hydroxyl and carbonyl region, measured by FT-IR/ATR technique, are shown in Fig. 3a and b, respectively. Only one broad band around 3340 cm1 appears in the hydroxyl region of the annealed irradiated samples and it can be attributed to the formation of hydroperoxides (around 3420 cm1), alcohols (around 3430 cm1)

167

1800 1500

Wavenumber [cm-1]

20

10

0 0

300

600

Absorbed dose [kGy]

Fig. 3. FT-IR/ATR spectra in the hydroxyl (a) and carbonyl (b) regions for unirradiated samples (1) and for those irradiated in air to 25 kGy (2), 100 kGy (3), 300 kGy (4) and 700 kGy (5); (c) the carbonyl index A1715(D)/A1715(0) as a function of absorbed dose.

and carboxylic acids (around 3210 cm1). The magnitude of this broad band first increases for low absorbed doses (100 kGy) and then levels off for median doses and starts decaying with a further increase in the radiation dose (300 kGy). On the other hand, in the carbonyl region the increase of absorbance in the range 1725e1715 cm1 indicates the formation of ketones. Saturated carboxylic acids are generally observed at 1755 and 1718e1710 cm1, while unsaturated acids absorb around 1700 cm1. The absorptions around 1780, 1740 and 1725 cm1 are usually ascribed to lactones, esters and aldehydes, respectively [44]. Ketones and carboxylic acids, responsible for the absorption maxima at 1715 cm1, are the main radiation-induced oxidation products observed in the carbonyl region. The evolution of carbonyl content (through the carbonyl index at 1715 cm1) with absorbed dose is presented in Fig. 3c. A linear dependence of the carbonyl content is evident for lower doses, while for higher ones (>200 kGy) an intense deviation (saturation) from linear curve occurs. The saturation in the carbonyl content for the iPP samples irradiated in air coincides with the start of gelation (Fig. 4a). Gavrila and Gosse have also found for the iPP gamma irradiated in air that the amount of carbonyl groups declines sharply at the gel point [48]. However, the generation of oxidizing species can probably be suppressed by intensive crosslinking behaviour and the formation of net structure, but more probably the real reason for the saturation in carbonyl and hydroperoxide contents is the fact that the oxygen present in the bulk and consumed due to radiation-induced reactions is not supplied fast enough by diffusion at higher doses. Apparently, the radiation-induced oxidation is limited by insufficient diffusion rate of oxygen and the accessibility of free radicals to atmospheric oxygen [49]. Post-radiation annealing also plays a significant role and makes contribution to the saturation in carbonyl and hydroperoxide contents. Annealing at elevated temperature introduces a much higher rate of thermal recombination of free radicals than the rate of oxygen diffusion from the surface into the film; this effect is more pronounced for the higher absorbed doses i.e. for the higher concentrations of free radicals and in the samples with higher crystallinity, since the free radicals trapped in the crystalline area are the main cause of the post-radiation oxidation observed in the case of gamma irradiated PP [44]. However, similarity between

168

E. Suljovrujic et al. / Polymer Degradation and Stability 95 (2010) 164e171

60

b

a

0 kGy 0 kGy annealed

Intensity [a.u.]

Gel [ ]

40

2.0

20

s+s 0.5

1.5

25 kGy

100 kGy 300 kGy

1.0

0.5

0

0

0.5

700 kGy

1

100/dose [kGy-1]

0

300

15

600

Absorbed dose [kGy]

c

40 35 30

0

300

45

600

Absorbed dose [kGy]

15

Crystal size [nm]

Crystallinity [ ]

45

30

2 [degree]

d

10

(110) 5

0

300

600

Absorbed dose [kGy]

Fig. 4. (a) The development of gel content with radiation dose. Shown by the Insert is the CharlesbyePinner (CeP) plot; (b) the WAXD diffractograms of iPP samples irradiated to various absorbed doses (25, 100, 300 and 700 kGy); (c) crystallinity as a function of radiation dose calculated from WAXD data; (d) crystal size as a function of radiation dose calculated for the most intense diffraction peak (110) of the a form at 2q ¼ 14.1.

the intensity of the dielectric a relaxation (Fig. 2a) and FT-IR/ATR data (Fig. 3) is evident, but it is necessary to bear in mind that in the case of the dielectric a relaxation there is a contribution only from the polar groups which are connected with the molecules contributing to this relaxation and the phase in which the relaxation occurs, while FT-IR/ATR data give information about the polar hydroperoxide and/or carbonyl groups in the thin surface layer. The radiation-induced increase in the amount of carbonyl, hydroperoxide and other polar groups in apolar iPP causes increase in the polymer polarity and thus such a large increase observed in the magnitude of the dielectric a process (Fig. 2a) can be clearly connected with the radiation-induced oxidation. Contrary to this fact, the explanation for the large radiation-induced shift in the position of the dielectric a relaxation is not so simple. The position of the a relaxation is significantly shifted to higher temperatures at low doses (25 and 50 kGy). For higher doses (100 kGy) the saturation and decay in the temperature at which the dielectric a relaxation occurs is evident (Figs. 1e and 2b). In order to identify the origin of the observed behaviour, the evolution of crystallinity, crystal size and gel content with absorbed dose may be helpful. There are three basic crystalline forms for melt crystallized iPP: monoclinic, pseudo-hexagonal, and triclinic. Moreover, rapid cooling (quenching) of an iPP melt to low-temperature produces a mixture of amorphous and smectic (mesomorphic) phase. Previous studies on quenched iPP have shown that its structure is very sensitive to thermal annealing [50e53]. Regardless of the doubts concerning the structure of the smectic phase, the common conclusion of the above mentioned studies was that the annealing of quenched iPP leads to the development of order. The smectic

phase is stable at room temperature for long periods of time, but it transforms into the monoclinic form when annealing temperature is above 60  C [52]. Furthermore, two limiting structures have been postulated for this crystal form: the limiting disorder modification (low melting point) and the limiting order modification (high melting point) [53]. Diffraction spectra in Fig. 4b show that quenched (virgin) iPP exhibits smectic phase, while the well known monoclinic form dominates in all annealed samples (unirradiated and irradiated). Annealing-induced transformation from smectic to monoclinic form is followed by the increase in the degree of crystallinity and in the crystal size. All these facts should be more than sufficient for the shifting of the dielectric a relaxation, but the comparison between the dielectric spectra of the quenched and annealed quenched samples does not confirm this. Although the shape of the dielectric a process is much broader for the case of the annealed quenched sample, the temperatures of the relaxation maxima (at different frequencies) are very similar to those in the quenched sample (Fig. 1e); from the mentioned it can be concluded that the shift in the position of the dielectric a relaxation should be connected with the changes in iPP structure induced by radiation. Crystallinity and crystal size as functions of radiation dose are presented in Fig. 4c and d; it is evident that the crystallinity and crystal size of iPP increase for low radiation doses and decrease for higher ones (100 kGy). In many studies the increase in crystallinity has been attributed to lamellar thickening and/or the increase in crystal perfection or to the formation of new lamellae by recrystallisation of small fractions which are produced by chain scission during irradiation [54,55]. On the other hand, the decrease in crystallinity was attributed to the formation of crosslinking [56] and to the radiation-induced defects within the crystals, as well as to those at the lateral grain boundaries [32,57]. The decrease in crystallinity is concomitant with the increase in the gel content observed in Fig. 4a. It can be noticed that iPP undergoes scission for the absorbed doses D < 200 kGy. Irradiation of iPP with higher doses, due to more intense crosslinking behaviour, leads to the significant gel formation (three-dimensional network). In general, many studies indicate that in the absence of an effective crosslinking co-agent and/or acetylene as a crosslinking medium, much higher doses (250 kGy) are required before the dose to incipient gelation is reached and the effects of significant levels of crosslinking begin to show [28,39,58e61]. The maximum crosslinking of iPP irradiated in air takes place at a dose of 500 kGy. Irradiation for higher doses (D > 500 kGy) leads to saturation and even to a small decay in the gel content. The CharlesbyePinner (CeP) plot is shown by the insert in Fig. 4a and estimated values from the CeP equation (DG ¼ 214 kGy, G(S) ¼ 0.362, G(X) ¼ 0.257, G(S)/G(X) ¼ 1.41, gmax ¼ 77.2 and fc ¼ 0.963) discussed in our previous paper [19] are in good agreement with the literature data. Herein, the dose required to reach the gel point is gelation dose (DG), G(S) and G(X) are radiation yields of scission and crosslinking, gmax is the calculated gel fraction for infinite dose and fc is the correlation coefficient of linear regression. The shift in the position of the dielectric a relaxation is most intensive for lower doses (100 kGy) at which gel content is zero and oxidative degradation dominates. Thus, it is not possible to make the correlation between the shift in the position of the dielectric a relaxation and the crosslinking despite the fact that in the literature crosslinking and gel formation are the main suspects for the shift of relaxations to higher temperatures [4,62]. Furthermore, changes in crystallinity and crystal size can also influence the position of the a relaxation. An important feature of the a process in PE is that its location depends primarily on the crystal lamellar thickness; it has been demonstrated by Popli et al. [63], Mansfield et al. [26] and Nitta et al. [64] that the temperature of this transition increases with the crystallite thickness for a series of branched,

E. Suljovrujic et al. / Polymer Degradation and Stability 95 (2010) 164e171

D=100 kGy

0

tan [10-3]

8 4 0

tan [10-3]

8 4 0

tan [10-3]

T =357.5 K

100 kHz

300

c

350

T =374 K

1 MHz

300

350

e 1 MHz

300

350

T[K]

400

T =346.5 K

b

350

400

Activation energy [kJ/mol]

4 0 kGy 25 kGy 100 kGy 300 kGy 700 kGy

b

150

120

90 2.8

3.0

-1

1000/ T [K ]

0

300

600

Absorbed dose [kGy]

Fig. 6. (a) Loss map for the a process; (b) activation energy (Ea) for the dielectric a process versus absorbed dose.

and a larger difference between the temperatures of each component at different frequencies and/or absorbed doses confirmed this fact (Figs. 1 and 5). In spite of this, in many cases it is almost impossible to determine activation energies for each component. The loss map for the a processes is shown in Fig. 6a. An Arrhenius temperature dependence is observed and calculated activation energies are presented in Fig. 6b. It is evident that the activation energy of the dielectric a relaxation significantly increases with radiation at lower doses (100 kGy), while saturation, even a small decay in the activation energy is evident at higher doses. The evolution of the activation energy with absorbed dose is relatively similar to that observed for the position of the dielectric a relaxation and probably has the same origin. The shift in the position and increase in the activation energy of the dielectric a relaxation is most intensive for the low absorbed doses at which oxidative degradation dominates; an explanation for this can be found in the prevalence of high-temperature components probably characterized by a higher activation energy. Deviation from such behaviour is evident at higher doses at which crosslinking dominates over the oxidative degradation and a significant level of net formation is achieved.

6

1 MHz

4. Conclusions T =364 K

d

6 0

400

T =374 K

100 kHz

0 300 12

400

tan [10-3]

4

a

D=700 kGy 12

300 12

tan [10-3]

tan [10-3]

8

5

3

180

a

6

log f

linear and metallocene catalyzed PEs, as well as for ethylene-vinyl acetate copolymers and hydrogenated polybutadienes. This is true both dielectrically and mechanically and the observation holds for both bulk crystallized PE specimens and single crystal mats [65]. In our case, it appears that the radiation-induced changes in crystallinity and crystal size can have some, but not decisive influence on the position of the a relaxation. The crystallinity and crystal size for highly irradiated samples are smaller than those for unirradiated ones (Fig. 4c and d), but the dielectric a relaxation still occurs at much higher temperatures (Fig. 1e). There exists a more probable explanation for the observed relaxation shift with radiation and is connected with the complex and multiple nature of this relaxation, confirmed by the presence of shoulders in Fig. 1. Analysis and fitting of multiple peaks in experimental spectra of irradiated samples are presented in Fig. 5, for two different absorbed doses. Comparison with the virgin iPP (Fig. 1) revealed two interesting facts. At the lower frequencies, in all irradiated samples (Fig. 5a and b) the dielectric a relaxation can be resolved in two fitted components, which is not the case with unirradiated samples (Fig. 1b). Furthermore, while the low-temperature component (a1) is dominant in unirradiated (quenched and annealed quenched) samples, for irradiated ones high-temperature components are prevalent especially at lower doses (Fig. 5c and e). With further increase in absorbed dose, the decay in the position of the dielectric a relaxation can be explained by partial recovery of the low-temperature component (a1), probably due to the intensive crosslinking behaviour (Fig. 5d and f). Thus, the shift in the position of the dielectric a relaxation, observed in Fig. 2b, can presumably be explained by the prevalence of high-temperature components in this relaxation due to radiation-induced oxidative degradation. The reason for a more intensive increase in the high-temperature component with absorbed dose is difficult to nominate at this stage of investigation, but it is well known that each relaxation and each relaxation component can have a different sensibility to the oxidative degradation and introduction of specific polar groups, as well as to the crosslinking and net formation [4,66,67]. Each relaxation component is also characterized by different relaxation times and different activation energies, and in some cases by different origin; more visible shoulders, a better resolution

169

350

1 MHz

T =364 K

400

f

6 0

300

350

400

T[K]

Fig. 5. Experimental (,) and fitted dielectric loss tangent spectrum (solid line) and fitted components (dash and dash-dot lines) of dielectric spectrum (b (Tg), a1, a2 and a3) for two different absorbed doses: 100 kGy (a, c, e) and 700 kGy (b, d, f) at two different frequencies: 100 kHz (a, b) and 1 MHz (c, d, e and f).

Dielectric relaxation spectroscopy and gamma radiation were used as powerful methods for characterization and modification of iPP structure, respectively. The high sensitivity of iPP to gamma radiation is confirmed through the influence of absorbed dose on the dielectric relaxation behaviour. In our previous paper, the disappearance of the g relaxation in gamma irradiated iPP is connected with a strong radiation-induced oxidative degradation in air. In this paper, presented results reveal uncommon a relaxation behaviour with gamma radiation and confirm the multiple nature of this process. The radiation-induced increase in the amount of carbonyl, hydroperoxide and other polar groups in apolar iPP causes an increase in polarity, and thus the increase observed in the magnitude of the dielectric a process can be clearly connected with the radiation-induced oxidation. Contrary to this fact, the explanation for the large radiation-induced shift in the position of the dielectric a relaxation is not so simple. This shift is most intensive

170

E. Suljovrujic et al. / Polymer Degradation and Stability 95 (2010) 164e171

for lower doses (100 kGy) at which the gel content is zero and oxidative degradation dominates. Thus, is not possible to make the correlation between the shift in the position of the dielectric a relaxation and the crosslinking. Furthermore, WAXD data indicated that the radiation-induced changes in crystallinity and crystal size can have some but not decisive influence on the position of the a relaxation. The most probable explanation for the observed shift with radiation is connected with the complex and multiple nature of this relaxation. The low-temperature component (a1) is dominant in unirradiated (quenched and annealed quenched) samples, while for irradiated ones high-temperature components are prevalent, especially at lower doses. With further increase in absorbed dose, the decay in the position can be explained by partial recovery of the low-temperature component (a1), probably due to the intensive crosslinking behaviour. Thus, the shift in the position of the dielectric a relaxation can presumably be explained by the prevalence of high-temperature components in this relaxation due to the radiation-induced oxidative degradation. The reason for a more intensive increase in the high-temperature component with absorbed dose is difficult to nominate, but it is well known that each relaxation component can have a different sensibility to the oxidative degradation and introduction of specific polar groups, as well as to the crosslinking and net formation. An Arrhenius temperature dependence is observed and the changes in activation energies with absorbed dose are relatively similar to those observed for the position of the dielectric a relaxation. Acknowledgement This work has been supported by the Ministry of Science and Technology of the Republic of Serbia (grant No. 141013). The authors thank Dr. Miodrag Mitric for his help with WAXD studies. References [1] Zhuravlev SP, Zhuravleva NM, Polonskij YuA. Deformation characteristics of polypropylene film and thermal stability of capacitor insulation made on the base of polypropylene film. Elektrotekhnika 2002;11:36e40. [2] Fournie R. All film power capacitors. Endurance tests and degradation mechanisms. Bulletin de la Direction des etudes et recherches Serie B, Reseaux Electriques, Materiels Electriques 1990;1:1e31. [3] Montanari GC, Fabiani D, Palmieri F, Kaempfer D, Thomann R, Mülhaupt R. Modification of electrical properties and performance of EVA and PP insulation through nanostructure by organophilic silicates. IEEE Transactions on Dielectrics and Electrical Insulation 2004;11(5):754e62. [4] Hedvig P. Dielectric spectroscopy of polymers. Budapest: Academia Kiado; 1977. [5] Fouracre RA, MacGregor SJ, Judd M, Banford HM. Condition monitoring of irradiated polymeric cables. Radiation Physics and Chemistry 1999;54 (2):209e11. [6] Jourdan C, Cavaille JY, Perez J. Mechanical relaxations in polypropylene. A new experimental and theoretical approach. Journal of Polymer Science, Part B: Polymer Physics 1989;27(11):2361e84. [7] Umemura T, Suzuki T, Kashiwazaki T. Impurity effect of the dielectric properties of isotactic polypropylene. IEEE Transactions on Electrical Insulation 1982;EI-17(4):300e5. [8] Brandrup J, Immergut EH. Polymer handbook. New York: Wiley; 1975. [9] Starkweather HW, Avakian P, Matheson RR, Fontanella JJ, Wintersgill MC. Ultralow temperature dielectric relaxations in polyolefins. Macromolecules 1992;25(25):6871e5. [10] Quijada-Garrido I, Barrales-Rienda JM, Pereña JM, Frutos G. Dynamic mechanical and dielectric behavior of erucamide (13-Cis-Docosenamide), isotactic poly(propylene), and their blends. Journal of Polymer Science, Part B: Polymer Physics 1997;35(10):1473e82. [11] Sakai A, Tanaka K, Fujii Y, Nagamura T, Kajiyama T. Structure and thermal molecular motion at surface of semi-crystalline isotactic polypropylene films. Polymer 2005;46:429e37. [12] Castejón ML, Tiemblo P, Gómez-Elvira JM. Photo-oxidation of thick isotactic polypropylene films. II. Evolution of the low temperature relaxations and of the melting endotherm along the kinetic stages. Polymer Degradation and Stability 2001;71(1):99e111. [13] Perepechko II. Svoistva polimerov pri nizkih temperaturah. Moskva: Khimiya 1977.

[14] Hara T. Dielectric property of some polymers in low temperature region. Japanese Journal of Applied Physics 1967;6:147e50. [15] Hoyos M, Tiemblo P, Go cmez-Elvira JM. The role of microstructure, molar mass and morphology on local relaxations in isotactic polypropylene. The a relaxation. Polymer 2007;48(1):183e94. [16] Tiemblo P, Gomez-Elvira JM, García Beltrán S, Matisova-Rychla L, Rychly J. Melting and a relaxation effects on the kinetics of polypropylene thermooxidation in the range 80e170  C. Macromolecules 2002;35(15):5922e6. [17] Olivares N, Tiemblo P, Gomez-Elvira JM. Physicochemical processes along the early stages of the thermal degradation of isotactic polypropylene I. Evolution of the G Relaxation Under Oxidative Conditions Polymer Degradation and Stability 1999;65(2):297e302. [18] Pluta M, Kryszewski M. Studies of alpha-relaxation process in spherulitic and non-spherulitic samples of isotactic polypropylene with different molecular ordering. Acta Polymerica 1987;38(1):42e52. [19] Suljovrujic E. Gel production, oxidative degradation and dielectric properties of isotactic polypropylene irradiated under various atmospheres. Polymer Degradation and Stability 2009;94:521e6. [20] Arranz-Andrés J, Peña B, Benavente R, Pérez E, Cerrada ML. Influence of isotacticity and molecular weight on the properties of metallocenic isotactic polypropylene. European Polymer Journal 2007;43(6):2357e70. [21] McCrum NG. Density-independent relaxations in polypropylene. Journal of Polymer Science Part B: Polymer Letters 1964;2(5):495e8. [22] Read BE. Mechanical relaxation in isotactic polypropylene. Polymer 1990;30 (8):1439e45.  ski P, Kryszewski M. The effect of spherulite structure of polypropylene [23] Myslin on the thermally stimulated electret depolarization current. Polymer Bulletin 1980;2(11):761e8. [24] Yamashita T, Ikezaki K. A method for correlating charge traps of polypropylene to its morphology. Journal of Electrostatics 2005;63(6e10):559e64. [25] Boyd RH. Relaxation processes in crystalline polymers: experimental behavior e a review. Polymer 1985;26(3):323e47. [26] Mansfield M, Boyd RH. Molecular motions, thea relaxation, and chain transport in polyethylene crystals. Journal of Polymer Science: Polymer Physics Edition 1978;16(7):1227e52. [27] Roy SK, Kyu T, Manley RStJ. Physical and dynamic mechanical properties of ultradrawn polypropylene films. Macromolecules 1988;21(2):499e504. [28] Dole M. The radiation chemistry of macromolecules, vol. 2. New York: Academic Press; 1973. [29] Gao J, Lu Y, Wei G, Zhang X, Liu Y, Qiao J. Effect of radiation on the crosslinking and branching of polypropylene. Journal of Applied Polymer Science 2002;85 (8):1758e64. [30] Lugao AB, Otaguro H, Parra DF, Yoshiga A, Lima LFCP, Artel BWH, et al. Review on the production process and uses of controlled rheology polypropyleneGamma radiation versus electron beam processing. Radiation Physics and Chemistry 2007;76(11e12):1688e90. [31] Lugao AB, Artel BWH, Yoshiga A, Lima LFCP, Parra DF, Bueno JR, et al. Production of high melt strength polypropylene by gamma irradiation. Radiation Physics and Chemistry 2007;76(11e12):1691e5. [32] Chodak I. Properties of crosslinked polyolefin-based materials. Progress in Polymer Science (Oxford) 1995;20(6):1165e99. [33] Dadbin S, Frounchi M, Saeid MH, Gangi F. Molecular structure and physical properties of e-beam crosslinked low-density polyethylene for wire and cable insulation applications. Journal of Applied Polymer Science 2002;86:1959e69. [34] Krause B, Haußler L, Voigt D. Comparison of the molecular properties and morphology of polypropylenes irradiated under different atmospheres and after annealing. Journal of Applied Polymer Science 2006;100(1):634e9. [35] Valenza A, Piccarolo S, Spadaro G. Influence of morphology and chemical structure on the inverse response of polypropylene to gamma radiation under vacuum. Polymer 1999;40(4):835e41. [36] Spadaro G, Valenza A. Influence of the irradiation parameters on the molecular modifications of an isotactic polypropylene gamma-irradiated under vacuum. Polymer Degradation and Stability 2000;67(3):449e54. [37] Lugao AB, Hutzler B, Ojeda T, Tokumoto S, Siemens R, Makuuchi K, et al. Reaction mechanism and rheological properties of polypropylene irradiated under various atmospheres. Radiation Physics and Chemistry 2000;57 (3e6):389e92. [38] Setnescu R, Jipa S, Setnescu T, Podina C, Osawa Z. Chemiluminescence study on the oxidation of several polyolefins: II. Chemiluminescence from g-irradiated polymers. Polymer Degradation and Stability 1998;61(1):109e17. [39] Mark EJ. Physical properties of polymers handbook. Woodbury, New York: American Institute of Physics; 1996. [40] Singh A, Silverman J, editors. Radiation processing of polymers. Munich: Hanser Publishers; 1991. [41] Banford HM, Fouracre RA, Faucitano A, Buttafava A, Martinotti F. The influence of g-irradiation and chemical structure on the dielectric properties of PP. Radiation Physics and Chemistry 1996;48:129e30. [42] Banford HM, Fouracre RA, Faucitano A, Buttafava A, Martinotti F. The influence of chemical structure on the dielectric behavior of polypropylene. IEEE Transactions on Dielectrics and Electrical Insulation 1996;3(4):594e8. [43] Abdel-Hamid HM. Effect of electron beam irradiation on polypropylene filmsdielectric and FT-IR studies. Solid-State Electronics 2005;49(7):1163e7. [44] Rivaton A, Lalande D, Gardette J-L. Influence of the structure on the g-irradiation of polypropylene and on the post-irradiation effects. Nuclear Instruments and Methods in Physics Research B 2004;222:187e200.

E. Suljovrujic et al. / Polymer Degradation and Stability 95 (2010) 164e171 [45] Bertoldo M, Bronco S, Cappelli C, Gragnoli T, Andreotti L. Combining theory and experiment to study the photooxidation of polyethylene and polypropylene. The Journal of Physical Chemistry B 2003;107(11880e11888): 11880. [46] Lacoste J, Vaillant D, Carlsson DJ. Gamma-, photo-, and thermally-initiated oxidation of isotactic polypropylene. Journal of Polymer Science, Part A: Polymer Chemistry 1993;31(3):715e22. [47] Lacoste J, Vaillant D, Chmela S. Gamma-, photo- and thermally-initiated oxidation of polyolefines used in packaging. Journal of Polymer Engineering 1996;15(1e2):139e52. [48] Gavrila DE, Gosse B. Post-irradiation degradation of polypropylene. Journal of Radioanalytical and Nuclear Chemistry 1994;185(2):311e7. [49] Shinde A, Salovey R. Irradiation of ultrahigh-molecular-weight polyethylene. Journal of Polymer Science: Polymer Physics Edition 1985;23(8): 1681e9. [50] Ferrero A, Ferracini E, Mazzavillani A, Malta V. pp. / New X-ray study of the quenched isotactic polypropylene transition by annealing. Journal of Macromolecular Science - Physics 2000;B39(1):109e29. [51] Russo R, Vittoria V. Determination of intrinsic birefringence of smectic phase in isotactic polypropylene. Journal of Applied Polymer Science 1996;60 (7):955e61. [52] Cohen Y, Saraf RF. A direct correlation function for mesomorphic polymers and its application to the 'smectic' phase of isotactic polpropylene. Polymer 2001;42(13):5865e70. [53] Dudi c D, Kostoski D, Djokovi c V, Drami canin MD. Formation and behaviour of low-temperature melting peak of quenched and annealed isotactic polypropylene. Polymer International 2001;51(2):111e6. [54] Alariqi SAS, Kumar AP, Rao BSM, Singh RP. Effect of g-dose rate on crystallinity and morphological changes of g-sterilized biomedical polypropylene. Polymer Degradation and Stability 2009;94(2):272e7. [55] Suljovrujic E. Radiation modification of the physical properties of polyolefins. Belgrade: University of Belgrade; 2000. [56] Stojanovic Z, Kacarevic-Popovic Z, Galovic S, Milicevic D, Suljovrujic E. Crystallinity changes and melting behavior of the uniaxially oriented iPP exposed

[57]

[58]

[59]

[60]

[61]

[62]

[63]

[64] [65] [66]

[67]

171

to high doses of gamma radiation. Polymer Degradation and Stability 2005;87 (2):279e86. Suljovrujic E. The influence of molecular orientation on the crosslinking/ oxidative behaviour of iPP exposed to gamma radiation. European Polymer Journal 2009;45(7):2068e78. Krause B, Voigt D, Haußler L, Auhl D, Münstedt H. Characterization of electron beam irradiated polypropylene: influence of irradiation temperature on molecular and rheological properties. Journal of Applied Polymer Science 2006;100(4):2770e80. Lugao AB, Cardoso ECL, Lima LFCP, Hustzler B, Tokumoto S. Characterization study of gamma-irradiated, high melt-strength polypropylene. Nuclear Instruments and Methods in Physics Research, Section B: Beam Interactions with Materials and Atoms 2003;208(1e4):252e5. Han DH, Shin S-H, Petrov S. Crosslinking and degradation of polypropylene by electron beam irradiation in the presence of trifunctional monomers. Radiation Physics and Chemistry 2004;69(3):239e44. Jones RA, Cail JI, Stepto RFT, Ward IM. Atomistic and flory-stockmayer analyses of irradiated i-PP gel fractions and comparison with results from PE. Macromolecules 2000;33(20):7337e44. Suljovrujic E. Dielectric studies of molecular b-relaxation in low density polyethylene: the influence of drawing and ionizing radiation. Polymer 2002; 43(22):5969e78. Popli R, Mandelkern L, Benson RS. Dynamic mechanical studies of a andb relaxations of polyethylenes. Journal of Polymer Science Part A-2, Polymer Physics 1984;22(3):407e48. Nitta K-H, Tanaka A. Dynamic mechanical properties of metallocene catalyzed linear polyethylenes. Polymer 2001;42(3):1219e26. Boyd RH. Relaxation processes in crystalline polymers: molecular interpretation - a review. Polymer 1985;26(8):1123e33. Sáenz de Juano-Arbona VS, Vallés-Lluch A, Contat-Rodrigo L, Ribes-Greus A. Chemical and thermal characterization of high- and low-density irradiated polyethylenes. Journal of Applied Polymer Science 2002;86(8):1953e8. Suljovrujic E. Some aspects of structural electrophysics of irradiated polyethylenes. Polymer 2005;46(17):6353e9.