Fracture toughness and failure mechanisms in un-vulcanized and dynamically vulcanized PP/EPDM/MWCNT blend-nanocomposites

Abstract

The fracture toughness and deformation mechanisms of un-vulcanized and dynamically vulcanized polypropylene/ethylene–propylene–diene terpolymer (PP/EPDM) blends and polypropylene/ethylene–propylene–diene terpolymer/multi walled carbon nanotube (PP/EPDM/MWCNT) blend-nanocomposites were investigated using the essential work of fracture (EWF) methodology followed by detailed microscopy analysis. The effect of maleic anhydride grafted polypropylene (PP-g-MA) on the morphology and fracture toughness of the multicomponent system was also investigated. It was found that both the dynamic vulcanization and compatibilization using PP-g-MA increased the fracture toughness of the blend and blend-nanocomposite systems. The results illustrated that the dominant fracture mechanism related to the EPDM dispersed phase in un-vulcanized samples was the formation of dilatation bands due to cavitation and/or debonding of dispersed EPDM rubber particles. In vulcanized samples, the development of dilatation shear bands, resulting from repeated particle debonding, was suppressed and the formation of nanovoids and cavitation in rubber particles led to promoting shear yielding of the adjacent matrix and dense plastic deformation. Incorporation of MWCNTs into the PP/EPDM blend reduced the essential work of fracture (w_e) and enhanced the non-essential work of fracture (βw_p). In the blend-nanocomposites, two mechanisms induced by the MWCNTs were observed. While large MWCNT aggregates acted as favored sites for crack initiation, the individual MWCNT impregnated fibrils arrested the crack propagation. The presence of PP-g-MA diminished the negative effect of MWCNTs and further enhanced their positive effect through decreasing the size of large aggregates (favored sites to initiate large cracks) and increasing the number of dispersed individual MWCNT ropes (increasing the potential of impregnated fibril formation), respectively.

---

1. Introduction

Because of the notch sensitive nature and poor fracture toughness of polypropylene (PP) on exposure to sever conditions such as low temperature or high deformation rate, the blending of PP with elastomeric polymers such as styrene–ethylene–butylene–styrene (SEBS), ethylene–propylene rubber (EPR) and ethylene–propylene–diene terpolymer (EPDM) is an efficient strategy to increase its toughness.^1–4 Among these rubbers, EPDM is considered as one of the most effective impact modifiers for PP.³ However, incompatibility of PP and EPDM results in poor interfacial adhesion and consequently unsatisfactory mechanical properties of simple PP/EPDM blends.⁵ Mechanical properties of these blends can be improved by using compatibilizers^5–7 and/or dynamic vulcanization processes.^7–9 Based on the fact that rubber toughening usually sacrifices the modulus and stiffness of PP, simultaneously reinforcing and toughening PP is very important to maintain its stiffness-to-toughness balance. The incorporation of micro- and nano-size fillers into PP/elastomer blends can restore the required stiffness and strength. Among the fillers, carbon nanotubes (CNTs) in their different types (multi walled carbon nanotubes (MWCNTs) and single walled carbon nanotubes (SWCNTs)) have effective reinforcing effects, because of their large aspect ratio and high tensile strength.^10,11

The challenges for developing high performance polymer/CNT nanocomposites include homogeneous dispersion of CNTs in the polymeric matrix and providing the strong interfacial interactions to ensure efficient load transfer from the polymeric matrix to the CNTs. CNTs are usually present in bundles and exhibit a highly aggregated state in polymeric matrixes because of the strong inter-tube van der Waals interactions, which hold the bundles together.¹⁰ Aggregated CNTs could play the role of stress concentrators leading to the creation of microscopic defects whilst homogeneous dispersion of CNTs prevents the stress concentration and their large surface area provides greater interaction with the polymeric matrix and enhances the stress-transfer mechanisms. Moreover, the fracture strain of CNTs that is estimated to be 10–30%, allows extensive bending and buckling.¹² This high flexibility most probably provides an additional source of high impact strength for the nanocomposites.¹³

The mechanical properties and fracture behavior of PP/CNT nanocomposites have extensively been studied recently. Prashantha et al.¹⁴ and Zhao et al.¹⁵ reported that PP/MWCNT nanocomposites show ductile behavior in lower loadings (1 wt%) and brittle behavior with breaking after the yield point at higher loadings (2 wt%). Hemmati et al.¹⁶ reported that the formation of some MWCNT aggregates causes a decrease in the tensile and impact strengths of PP/MWCNT nanocomposites. Seo et al.¹⁷ showed that the impact strength of PP/MWCNT nanocomposites steadily increases with the MWCNT content at low filler content and gradually decreases after passing through a maximum with further increasing of the filler content because of CNT agglomeration. The fracture toughness of PP/MWCNT nanocomposites was studied by Satapathy et al.¹⁸ and they observed higher crack propagation resistance at 0.5 wt% MWCNT nanocomposite compared to pure PP. Prashantha¹⁹ used PP-g-MA as a compatibilizer to improve the dispersion of CNTs and showed that the Young’s modulus, yield stress and impact strength of PP/MWCNT/PP-g-MA nanocomposites are higher than those of PP/MWCNT nanocomposites. Valentini et al.²⁰ reported that the toughness of PP improved by incorporation of EPDM and it was further enhanced by addition of MWCNTs. It was also shown by Liu et al. that MWCNTs did not have a significant effect on the fracture toughness of a polypropylene/ethylene–vinyl acetate copolymer (PP/EVA) blend with a weight ratio of 80 : 20, while the impact strength of the PP/EVA (60/40) blend greatly increased with MWCNTs.²¹

Our previous research work⁷ showed that MWCNTs increased the impact strength of un-vulcanized and dynamically vulcanized PP/EPDM (80/20) blends, in which the presence of PP-g-MA had an effective role in increasing the fracture toughness of these blend-nanocomposites. The involved deformation micro-mechanisms were discussed in that work. In this study the fracture behavior of un-vulcanized and dynamically vulcanized PP/EPDM blends and PP/EPDM/MWCNT blend-nanocomposites was investigated using the essential work of fracture (EWF) method. The effect of PP-g-MA as a compatibilizer on the toughness of the blends and blend-nanocomposites was also investigated. The main focus of this study was to get insight into the micro- and nano-deformation mechanisms induced by different dispersed components which were involved in the toughening of such blends and blend-nanocomposites.

2. Experimental

2.1. Materials

Isotactic polypropylene (PP – SI080) with a melt flow index of 8.0 g/10 min (230 °C, 2.16 kg) was obtained from Polynar Petrochemical Company, Tabriz, Iran. EPDM-KEP270 with a density of 0.96 g cm⁻³ (ENB content = 4.5 wt%, ethylene content = 57 wt%) was supplied by Kumho Polychem Company, South Korea. The used multi walled carbon nanotubes (MWCNTs) were provided by Nanocyl. The specifications of the MWCNTs are as follows: the average diameter is 9.5 nm, the average length of the nanotubes is 1.5 μm and the purity is higher than 90%. The vulcanization ingredients, such as mercaptobenzothiazole disulfide (MBTS), tetramethylthiuram disulfide (TMTD), stearic acid and sulfur were supplied by Merck. PP-g-MA with a MFI value of about 80.0 g/10 min (180 °C, 2.16 kg) and 1.2 wt% of grafted maleic anhydride were used in this study.

2.2. Blend preparation

The blends, nanocomposites and blend-nanocomposites were prepared by melt mixing in an internal mixer (Brabender W50EHT) with a 55 ml mixing chamber at a starting temperature of 180 °C and a rotor speed of 60 rpm. The chamber was always charged with 45 ml of polymers and mixing was done for 12 min. The PP phase (PP or PP + PP-g-MA) was fed into the internal mixer followed by the addition of EPDM for un-vulcanized blends and further MWCNTs in the melt mixture for un-vulcanized blend-nanocomposites. For different vulcanized samples, 1.5 phr stearic acid and 5 phr zinc oxide were added after 4 min from melt mixing of the PP phase with EPDM and the MWCNTs (for the blend-nanocomposites). Then, after 1 min of mixing, accelerators (including 2.1 phr TMTD and 0.9 phr MBTS) were charged and then 1 phr sulfur was added after 30 s. Typical torque–time curves for the un-vulcanized and vulcanized blends together with the blend-nanocomposites are shown in Fig. 1. Neat PP and PP phase/MWCNT nanocomposites were also similarly processed for comparison. Prior to blending, the MWCNTs and PP-g-MA were dried at 80 °C under vacuum for 12 h. For all the blends and blend-nanocomposites the weight ratio of PP to EPDM was set at 80 : 20. For all the nanocomposites and blend-nanocomposites, 0.5 wt% of MWCNTs was used. To study the effect of PP-g-MA on the MWCNT dispersion and compatibilization of different blends, 2 wt% of PP-g-MA was used in the preparation of some samples. To prevent thermo-mechanical degradation during melt blending, 0.1 wt% of a thermal stabilizer (Irganox 1010) was used.

Fig. 1 Schematic mixing torque–time diagrams of the un-vulcanized blend and blend-nanocomposite (a and b) and the vulcanized blend and blend-nanocomposite (c and d).

2.3. Instruments

Samples for the EWF tests were prepared by compression molding according to ISO 293 and cooled to ambient temperature by water with a 15 ± 2 °C min⁻¹ average cooling rate. The dimension of the EWF test samples was 80 × 23 × 1 mm. The notches were inserted using a sharp razor blade to obtain double edge notched tension (DENT) specimens with a series of ligament lengths ranging from 13 to 19 mm. At least five specimens were tested for each ligament length and data reduction followed the recommendation of the ESIS-TC4 group.²² All the fractures for the EWF test were performed at room temperature with a crosshead speed of 5 mm min⁻¹ and using a Zwick/Roell tensile testing machine (Z 010).

To study the morphology and micro-mechanical deformation processes three preparation and investigation techniques were used:

(1) the cryofractured surfaces were subjected to pre-treatment to remove the EPDM phase in the un-vulcanized samples and also to remove the PP phase in the vulcanized samples. In the un-vulcanized cryofractured samples, the EPDM phase was etched by cyclohexane at room temperature for 24 h and in the vulcanized cryofractured samples, the PP phase was etched by boiling xylene for 30 s. The treated samples were then sputter-coated with gold prior to analysis;

(2) single edge notched (SEN) semi thin films (approximately 40 μm thickness) with dimensions of 80 × 23 mm were prepared by the melt-pressing method and inserting a notch using a sharp razor blade. The prepared film samples were conducted under a tensile test with a crosshead speed of 5 mm min⁻¹ and the test was terminated when a load of 120 N was reached. Also the un-notched films (of the same dimensions) were conducted under a stress–strain test with a crosshead speed of 5 mm min⁻¹ until complete failure. In the case of the un-notched films, plastically deformed zones close to the fracture plane, and the arrested crack-tip damage zone for single edge notched films, were evaluated by optical microscopy. Although the stress–strain state in the semi thin sections is different from that in the bulk material, the nature and mode of deformation cannot be changed and the micro-mechanical deformation mechanisms are comparable in both cases;²³

(3) the morphology of the surface and sub-surface of the EWF test fractured samples was also investigated. The area of sub-surface interest (marked with an arrow in Fig. 2a) is perpendicular to the fracture surface and perpendicular to the crack growth direction.

Fig. 2 (a) Locations inside the plastic deformation zone that were observed by scanning electron microscopy, and (b) the deeply double edge notched specimen and the different energy dissipation zones involved.

The morphology of the fractured surfaces was examined with a HITACHI S-4160 field emission scanning electron microscope (FE-SEM). The deformed and fractured films were studied using an Olympus BX60 optical microscope equipped with a Sony camera (model: SSC-DC58AP).

2.4. Essential work of fracture (EWF) methodology

The EWF test method has gained popularity due to its experimental simplicity as the method avoids measurements of the current crack advance as well as detection of the cracking initiation. The EWF concept that was proposed first by Broberg²⁴ and further developed by Cotterell and Reddel²⁵ and Karger-Kocsis^26–29 has been used extensively to study the fracture behavior of ductile polymeric materials. In this method the total energy required to fracture (W_f) can be divided into the essential work of fracture (W_e) consumed in the inner fracture process zone (IFPZ) to create a new surface and the non-essential work of fracture (W_p) dissipated in the outer plastic deformation zone (OPDZ). The related zones are shown schematically in the case of a deeply double edge notched tension specimen in Fig. 2b.

The relationship between the work of fracture and its components can be written as follows:

W_f = W_e + W_p (1)

By assuming that W_e is proportional to the ligament area, and W_p is proportional to the volume of the plastic zone, the relationship between the specific terms can be defined as follows:

W_f = w_e × lt + βw_p × l²t (2)

w_f = w_e + βw_pl (3)

where w_f is the specific total work of fracture, l is the ligament length, t is the thickness of the sample and β is a plastic shape factor depending on the geometry of the plastic zone.

The total work of fracture, W_f, can be determined by integration of the area under force–displacement curves for each specimen. According to eqn (3), the specific total work of fracture is a linear function of the ligament length. Therefore w_e, that indicates the resistance to crack initiation, can be determined from the interception of the linear regression line fitted to the w_f − l graphs with the y-axis, and the slope of this line, βw_p, indicates the resistance against crack propagation.

As discussed by Karger-Kocsis et al.^30,31 the total work of fracture can be partitioned at the maximum load as the summation of two contributes: a term W_y related to the yielding of the ligament area and another term W_n,t associated to the subsequent necking and tearing. Therefore it can be written that:

W_f = W_y + W_n,t (4)

The specific terms (w_y and w_n,t) can be expressed as a function of the ligament length similar to eqn (2), as follows:

w_f = w_y + w_n,t = (w_e,y + β_yw_p,yl) + (w_e,n,t + β_nw_p,n,tl) (5)

where w_e,y is the specific essential yielding-related work of fracture, w_e,n,t is the specific essential necking and tearing work, w_p,y is the volumetric energy dissipated during yielding and w_p,n,t is the dissipated plastic work during necking and tearing. β_y and β_n,t are geometry factors related to the shape of the plastic zone during the yielding and necking and tearing, respectively. By comparing eqn (3) and (5), it can be concluded that:

w_e = w_e,y + w_e,n,t (6)

βw_p = β_yw_p,y + β_nw_p,n,t (7)

3. Results and discussion

3.1. Phase morphology

The SEM micrographs of the cryofractured surfaces of the PP/MWCNTs with and without PP-g-MA are shown in Fig. 3. It is obvious that the MWCNTs are present in the PP matrix mainly in the form of large aggregates. This indicates imperfect mixing of the MWCNT phase in the PP matrix. In the presence of PP-g-MA reasonably uniform dispersion and good distribution of the MWCNTs with smaller aggregate size were observed. This suggests that the presence of PP-g-MA improves the dispersion state of the MWCNTs in the PP matrix and limits the formation of MWCNT aggregates.¹⁹

Fig. 3 SEM micrographs of (a) PP/MWCNT and (b) PP/MWCNT/PP-g-MA. MWCNT aggregates are shown by arrows.

Fig. 4 shows the SEM images of the cryofractured surfaces of un-vulcanized PP/EPDM and PP/EPDM/MWCNTs with and without PP-g-MA at two different magnifications. The dark holes in the micrographs represent the etched EPDM particles. For the blends and blend-nanocomposites, the matrix-dispersed type phase morphology could be observed. By comparing Fig. 4a–b′ it seems that the incorporation of PP-g-MA into the PP/EPDM blend leads to a reduction in the size of dispersed EPDM particles through compatibilization of the PP and EPDM phases.⁷

Fig. 4 SEM micrographs of different samples. (a and a′) Un-vulcanized PP/EPDM, (b and b′) un-vulcanized PP/EPDM/PP-g-MA, (c and c′) un-vulcanized PP/EPDM/MWCNT and (d and d′) un-vulcanized PP/EPDM/PP-g-MA/MWCNT. The EPDM phase was etched by cyclohexane.

It can be seen in Fig. 4c and c′ that the minor EPDM and MWCNT components are mainly separately distributed in the PP matrix in the PP/EPDM/MWCNT blend-nanocomposite sample. Although the addition of PP-g-MA into the PP/EPDM/MWCNT blend-nanocomposite reduced the size of the MWCNT aggregates, no obvious effect was observed for the size of the dispersed EPDM particles. The size reduction of the MWCNT aggregates due to the presence of PP-g-MA in the PP/EPDM/MWCNT sample was also reported in our previous work, as revealed via optical microscopy analysis.⁷ By considering that a constant amount of PP-g-MA was used in all the blends and blend-nanocomposites, the less evident effect of PP-g-MA on the EPDM particles in the blend-nanocomposite may be related to the higher interfacial surface area (PP–EPDM and PP–MWCNT interfaces) in this system than in the blend (PP–EPDM interface) which lowers the efficiency of PP-g-MA in decreasing the size of the EPDM particles in the blend-nanocomposites.

The cryofractured surfaces of the vulcanized samples did not provide a proper insight into the phase morphology (see the ESI†), because of agglomeration of the EPDM dispersed particles during the etching of the PP phase (with boiling xylene for 30 s). This agglomeration originates from the high surface area of the EPDM dispersed particles and the high temperature used for dissolving the PP matrix.³²

3.2. Load–displacement curves and EWF parameters

The load–displacement curves of the DENT specimens for all the samples during the EWF tests are shown in Fig. 5. This figure demonstrates that the PP, PP/PP-g-MA and related nanocomposites exhibit brittle-like behavior, without a stable crack propagation stage. The trend of the load–displacement curves indicates that fast crack propagation takes place in the PP and PP/PP-g-MA samples without yielding of the ligament whilst PP/MWCNT and PP/MWCNT/PP-g-MA show a sharp yielding peak, followed by fast crack propagation and sudden failure. This yielding peak is related to the localized plasticity at the roots of pre-cracks of the EWF test specimens.

Fig. 5 Load–displacement curves of the DENT specimens with different ligament lengths in the EWF tests. (a) PP, (b) PP/PP-g-MA, (c) PP/MWCNT, (d) PP/PP-g-MA/MWCNT, (e) un-vulcanized PP/EPDM, (f) vulcanized PP/EPDM, (g) un-vulcanized PP/EPDM/PP-g-MA, (h) vulcanized PP/EPDM/PP-g-MA, (i) un-vulcanized PP/EPDM/MWCNT, (j) vulcanized PP/EPDM/MWCNT, (k) un-vulcanized PP/EPDM/PP-g-MA/MWCNT and (l) vulcanized PP/EPDM/PP-g-MA/MWCNT.

In the case of the un-vulcanized and vulcanized blends as well as the blend-nanocomposites, well developed crack initiation and crack propagation regions are observed on the load–displacement curves. The load reached a maximum as a definite yield point occurred, and after the peak, the load dropped steadily until failure of the specimen. This trend is characteristic of ductile fracture behavior as the ligament fully yields and the crack propagates in a stable manner.

Fig. 5 also indicates that the presence of MWCNTs in PP/EPDM blends causes some instability during the crack propagation stage of both un-vulcanized and vulcanized systems. These instabilities are believed to be due to non-homogeneous dispersion of the MWCNTs and the presence of carbon nanotube aggregates in the matrix.³³ It is also obvious from Fig. 5 that the addition of PP-g-MA into different blend-nanocomposites diminished the level of instability during the fracture process. This may be attributed to the size reduction of the MWCNT aggregates due to the presence of PP-g-MA as discussed earlier and reported in our previous work.⁷

The load–displacement curves for all the blends and blend-nanocomposites show that the shape of the curves remains the same as the ligament length increases. This is one of the most important prerequisites of the EWF test method and indicates that the crack propagates in the same manner for different ligament lengths, with the fracture mode of the test sample being unchanged with the ligament length. To ensure that the EWF data were obtained under plane stress conditions for the blends and blend-nanocomposites, and to remove data where fracture occurred prior to full ligament yielding, the net-section stress at the maximum σ_max (σ_max = F_max/lt, where F_max is the maximum load in the load–displacement curves) was calculated for different ligament lengths and plotted against the ligament length for different samples (Fig. 6). Accordingly, an average value of maximum stress (σ_mean) was calculated and the results are assumed to be valid if the maximum stress is within the range of 0.9–1.1σ_mean.

Fig. 6 Maximum stress (σ_max) against ligament length for (a) un-vulcanized PP/EPDM, (b) vulcanized PP/EPDM, (c) un-vulcanized PP/EPDM/PP-g-MA, (d) vulcanized PP/EPDM/PP-g-MA, (e) un-vulcanized PP/EPDM/MWCNT, (f) vulcanized PP/EPDM/MWCNT, (g) un-vulcanized PP/EPDM/PP-g-MA/MWCNT and (h) vulcanized PP/EPDM/PP-g-MA/MWCNT.

Fig. 6 clearly shows that the data for all the blends and their nanocomposites lie within the range of 0.9σ_mean and 1.1σ_mean, indicating that the EWF tests have been conducted under plane stress conditions recommended by the ESIS protocol.²² Another important prerequisite of the EWF test method is that the plastic zone has not spread over the lateral boundaries of the specimens.

According to Fig. 7, there is no evidence for spreading of the yielded zones to the lateral boundaries of the EWF test samples. The above mentioned criteria confirm the applicability of the EWF test method to the blends and blend-nanocomposites with ductile fracture behavior in this work and also the validity of the obtained data.

Fig. 7 Fractured DENT samples under the EWF test. (a) Un-vulcanized PP/EPDM, (b) vulcanized PP/EPDM, (c) un-vulcanized PP/EPDM/PP-g-MA, (d) vulcanized PP/EPDM/PP-g-MA, (e) un-vulcanized PP/EPDM/MWCNT, (f) vulcanized PP/EPDM/MWCNT, (g) un-vulcanized PP/EPDM/PP-g-MA/MWCN and (h) vulcanized PP/EPDM/PP-g-MA/MWCNT.

It should be noted that the lack of prerequisites of the EWF test method invalidates the use of this approach for evaluation of the fracture toughness of PP, PP/PP-g-MA and related nanocomposites that display unstable crack propagation. Although the fracture energies obtained for these samples are apparent values and not the intrinsic properties, the crack resistance of these materials was determined by the EWF methodology to provide a qualitative comparison with those calculated for ductile samples.³⁴

The values of w_e and βw_p for PP, PP/PP-g-MA and their nanocomposites obtained from the interception extrapolated to zero ligament length and the slope of the straight lines, respectively, are listed in Table 1. The results of Table 1 indicate that almost the total fracture energy of the PP, PP/PP-g-MA, PP/MWCNT and PP/MWCNT/PP-g-MA samples is dissipated in the inner region near the fracture surface (w_e) and a negligible energy is dissipated in the outer plastic zone (βw_p). The results also show that the addition of PP-g-MA into PP reduced the w_e and βw_p values. Although the extent of reduction is low, it could be attributed to the poor mechanical properties of PP-g-MA relative to PP, owing to its lower molecular weight. On the other hand, incorporation of MWCNTs increased the w_e and βw_p values and these parameters further increased by addition of PP-g-MA into the PP/MWCNT nanocomposites. While the dominant fracture mechanism in these brittle-like samples seems to be crazing, the increased w_e and βw_p values of the MWCNT filled nanocomposites may be due to the nucleation of multiple crazing promoted by the MWCNT aggregates. It should be noted that the MWCNTs (nanotube ropes and nanotube aggregates) reduce the spherulite size of PP through acting as a nucleating agent.³⁵ This leads to a smaller inter-spherulitic boundary that is favored for crack initiation sites and also could allow the matrix to deform more easily, since polymers with larger spherulites tend to be more embrittle.³⁶ Therefore, promoting multiple crazing and reduction of the spherulite size could be responsible for the increased fracture toughness of the PP-based nanocomposites as compared with pure PP, which in turn led to yielding of the material at the notch tip. The slight increase in the work of fracture parameters of PP/MWCNTs upon the addition of PP-g-MA could be attributed to the better dispersion of the MWCNTs in the matrix, which is discussed in the “phase morphology” section.

Table 1 The essential (w_e) and non-essential (βw_p) work of fracture for PP, PP/PP-g-MA and related nanocomposites (R² for all the samples is higher than 0.93)

Samples	we, N mm^−1	βwp, N mm^−2
PP	4.6 ± 0.2	0.61 ± 0.06
PP/PP-g-MA	4.5 ± 0.2	0.51 ± 0.05
PP/MWCNT	5.03 ± 0.2	0.81 ± 0.07
PP/MWCNT/PP-g-MA	5.2 ± 0.2	0.85 ± 0.06

---

A summary of the values of the w_e and βw_p parameters together with the results for splitting the essential and non-essential work of fracture into the yielding and necking terms for different blends and blend-nanocomposites, obtained by plotting w_y − l and w_n − l, are listed in Table 2.

Table 2 The results of essential and non-essential work of fracture and splitting the essential and non-essential work of fracture into yielding and necking and tearing for all the blends and blend-nanocomposites (R² for all the samples is higher than 0.97)

		Samples	we, N mm^−1	βwp, N mm^−2	we,y, N mm^−1	βywp,y, N mm^−2	we,n, N mm^−1	βnwp,n, N mm^−2
Un-vulcanized	PP/EPDM		9.7 ± 0.5	2.3 ± 0.15	1.28 ± 0.1	0.28 ± 0.03	8.47 ± 0.5	2.08 ± 0.15
	PP/EPDM/PP-g-MA		18.7 ± 1	3 ± 0.12	1.42 ± 0.1	0.41 ± 0.0.3	17.3 ± 1	2.52 ± 0.15
	PP/EPDM/MWCNT		7.8 ± 0.6	2.8 ± 0.2	2.2 ± 0.12	0.6 ± 0.05	5.6 ± 0.3	2.2 ± 0.18
	PP/EPDM/MWCNT/PP-g-MA		13.2 ± 0.6	3.3 ± 0.2	3.6 ± 0.12	0.8 ± 0.05	9.6 ± 0.5	2.5 ± 0.2
Vulcanized	PP/EPDM		47.3 ± 2	5 ± 0.2	3.8 ± 0.2	0.89 ± 0.05	43.5 ± 2	4.1 ± 0.25
	PP/EPDM/PP-g-MA		50.8 ± 2	5.2 ± 0.25	4.2 ± 0.23	0.98 ± 0.05	46.6 ± 2	4.2 ± 0.25
	PP/EPDM/MWCNT		30.5 ± 2	5.4 ± 0.3	4.8 ± 0.25	1.1 ± 0.06	25.7 ± 2	4.3 ± 0.3
	PP/EPDM/MWCNT/PP-g-MA		32.5 ± 2.5	5.6 ± 0.3	5.89 ± 0.3	1.2 ± 0.06	26.6 ± 2.5	4.4 ± 0.3

---

Comparing the results of Table 2 with those of Table 1 shows that incorporation of 20 wt% EPDM into PP increased the w_e and βw_p parameters simultaneously, as expected. This is attributed to the toughening mechanisms such as internal void formation, matrix shear yielding and/or interfacial debonding related to the dispersed rubber particles. The addition of PP-g-MA enhanced the essential and non-essential work of fracture of the PP/EPDM blend, contrary to pure PP. The compatibilization effect of PP-g-MA in the PP/EPDM blend with subsequent improvement in the interfacial adhesion between the PP and EPDM phases could be the main reason for the increased fracture toughness of the PP/EPDM/PP-g-MA blend. The compatibilization role of PP-g-MA has been reported in the literature^6,7 and also is confirmed in the previous section.

According to the results of Table 2, it is obvious that the incorporation of MWCNTs into the un-vulcanized PP/EPDM blend reduced the w_e value along with an increase in βw_p. Similar to the PP/EPDM blend, the presence of PP-g-MA in the un-vulcanized PP/EPDM/MWCNT nanocomposite increased both the w_e and βw_p values. However, the extent of improvement in the w_e value for the un-vulcanized PP/EPDM blend upon the introduction of PP-g-MA is much higher than that obtained by addition of PP-g-MA into its blend-nanocomposite counterpart. This might be due to the fact that in the compatibilized PP/EPDM/MWCNT blend-nanocomposite, the PP-g-MA distributes at the interfacial region between both the dispersed components (MWCNTs and EPDM) and the matrix which could reduce its effective concentration at the PP/EPDM interface as compared to the unfilled PP/EPDM binary blend.

The data in Table 2 show that the dynamic vulcanization greatly improved the specific EWF and specific plastic work values of the blends and blend-nanocomposites, so that the vulcanized samples exhibited significantly higher work of fracture parameters than their un-vulcanized counterparts. This is primarily due to the fact that the vulcanization process improves the interfacial interaction between the PP and EPDM phases and also increases the cohesive strength of the EPDM rubber particles. Similar to the un-vulcanized samples, the presence of MWCNTs in the vulcanized PP/EPDM blends (with and without PP-g-MA) decreased the w_e value and increased the βw_p parameter. The reason behind these variations and involved micro-mechanisms is more evidenced by splitting the different terms as appears below.

3.3. Constituting terms of the essential and non-essential work of fracture

In the case of neat PP and the PP-based samples with brittle-like fracture behavior, the w_e and βw_p values reported in Table 1 are almost entirely consumed during yielding up to the crack initiation stage of the fracture process (w_e ≈ w_e,y and βw_p ≈ β_yw_p,y) with negligible contribution from the crack propagation stage (w_e,n,t ≈ 0 and β_nw_p,n,t ≈ 0), as expected. But for the un-vulcanized and vulcanized blends and the blend-nanocomposites which behaved in a ductile manner with stable crack propagation, it is apparent that the contribution made by the necking and tearing stage is much greater than that of the yielding stage for both the w_e and βw_p parameters, as measured and recorded in Table 2. This is characteristic of completely stable crack propagation in these samples. Very interesting results can be observed at a first look to Table 2. The presence of PP-g-MA significantly increased the energy dissipation during the necking and tearing stage (w_e,n,t) of the fracture process for the blends and blend-nanocomposites, especially for the un-vulcanized samples. The results of Table 2 also demonstrate that on one hand, the un-vulcanized and vulcanized PP/EPDM/MWCNT blend-nanocomposites have a higher w_e,y and lower w_e,n than the un-vulcanized and vulcanized PP/EPDM blends, respectively, and on the other hand, the incorporation of PP-g-MA into the un-vulcanized and vulcanized PP/EPDM/MWCNT blend-nanocomposites increases both w_e,y and w_e,n. Moreover by considering that the presence of MWCNTs has a negative effect on the fracture toughness (w_e) of the un-vulcanized and vulcanized PP/EPDM samples (with and without PP-g-MA), it may be concluded that the MWCNT aggregates facilitate the tearing process, while the presence of PP-g-MA due to the reduction in the size of the MWCNT aggregates increases the resistance to tearing and crack propagation. Overall, the results of Table 2 show that the fracture toughness of all the samples is controlled by the necking and tearing components of the work of fracture. This is more evidenced and discussed using detailed fractography studies in the next section.

3.4. Toughening and failure mechanisms

SEM images of PP, PP/PP-g-MA, PP/MWCNT and PP/MWCNT/PP-g-MA fractured surfaces under EWF test conditions are depicted in Fig. 8. Smooth fractured surfaces without shear yielding and/or plastically deformed materials imply that little energy is dissipated during the fracture process of these samples (Fig. 8a–d). These morphologies confirm the brittle-like failure behavior of the PP, PP/PP-g-MA, PP/MWCNT and PP/MWCNT/PP-g-MA samples with low fracture toughness.

Fig. 8 SEM micrographs of the fracture surface of the EWF test specimens. (a) PP, (b) PP/PP-g-MA, (c) PP/MWCNT and (d) PP/MWCNT/PP-g-MA.

As stated before, the presence of MWCNTs not only reduces the size of spherulites of the PP matrix,³⁶ but also the MWCNT aggregates could serve as the sites for the initiation of craze-like features throughout the PP-based nanocomposites. Moreover, the MWCNT aggregates could act as crack deviations and therefore could change the direction of crack growth. These mechanisms were also reported for halloysite nanotubes filled with PA6.³⁷ The toughening mechanisms, induced by MWCNTs in the PP/MWCNT nanocomposite, are demonstrated in Fig. 9. The optical microphotograph of the PP/MWCNT nanocomposite shows that the MWCNT aggregates act as both craze initiation sites (Fig. 9a) and crack deviations (Fig. 9b).

Fig. 9 Optical photomicrograph of PP/MWCNT: (a) MWCNT aggregate acting as a craze initiation site, and (b) MWCNT aggregate hindering a crack patch. The black arrows in (b) show the crack path.

Optical microscopy images of deformation zones developed ahead of the arrested crack-tip of SEN films for PP, and the un-vulcanized and vulcanized PP/EPDM blends are shown in Fig. 10. These microphotographs were obtained by transmission optical microscopy (TOM) in the bright-field mode, so the voided materials appear dark in the micrographs. Little deformation surrounding the crack-tip of the PP sample is observed in Fig. 10a. This image indicates that the crack-tip plastic zone of PP consists mainly of a small number of long micro-cracks. When the plastic zone reaches a critical size and fulfils the unstable condition of the deformation, the macro-crack develops and the brittle fracture takes place.

Fig. 10 Optical microscopy images of deformation zones that developed ahead of the arrested crack-tip. (a) PP, (b) un-vulcanized PP/EPDM and (c) vulcanized PP/EPDM.

In contrary with PP, the crack-tip damage zone of the un-vulcanized PP/EPDM blend (Fig. 10b) is composed of a large number of dilatation band features, so that a large damage zone has been developed in front of the crack-tip. It is believed that the dilatation bands are formed by sequential (repetitive) cavitation and/or debonding of the dispersed EPDM rubber particles.^2,38 The presence of EPDM particles leads to relaxation of the stress concentration due to the release of strain constraint by internal void formation of the EPDM particles and/or interfacial debonding from the surrounding matrix. Therefore, the nucleation of the catastrophic crack is suppressed and the fracture toughness (w_e) is improved as compared with pure PP.

It is clearly visible in Fig. 10c that dynamic vulcanization further intensifies the different deformation processes infront of the crack-tip. As can be seen, a larger crack-tip damage zone composed of a much more dense bundle of dilatation bands is developed ahead of the crack-tip for the vulcanized PP/EPDM blend as compared to the un-vulcanized blend. It seems that a larger number of bands with much lower thicknesses have been developed in the vulcanized blend. Moreover, the vulcanized PP/EPDM blend showed a large crack-tip yielded zone at close proximity of the pre-crack, visible as a dark dog-bone region which is absent in the un-vulcanized PP/EPDM blend. Therefore, a larger volume of the material participates in energy absorption and/or dissipation processes in the vulcanized sample than the un-vulcanized sample, which results in a greatly improved fracture toughness of the vulcanized blend.

Fig. 11 shows optical microscopy images taken from sub-surface damaged zones of un-notched films along with SEM images taken from the sub-surface of the EWF fractured sample, for the un-vulcanized and vulcanized PP/EPDM blends. Fig. 11a shows dilatation bands in the sub-surface deformed zone which is apparently constituted of a row of cavitated and/or debonded rubber particles. It is clear from this image that the extent of deformation decreases when moving away from the fractured surface in the specimen. The formation of these bands confirms the hypothesis of a limited adhesion between PP and the dispersed EPDM particles. During the band propagation, the ligaments of the polymer between the micro-voids are allowed to deform and stretch. This plastic deformation and stretching of the ligaments is actually the real source of toughness associated with the formation of dilatational shear bands³⁸ and is responsible for the higher toughness of the un-vulcanized PP/EPDM blend than the neat PP. Fig. 11b also shows some elongated voids formed due to the formation of internal cavities in the rubber particles and/or debonding at the interface of PP and the EPDM particles. It can interestingly be seen in this figure that the cavities are arranged preferentially along a straight line. This is in accordance with the results of optical microscopy (Fig. 11a) and emphasizes again that the deformation mechanism in the un-vulcanized PP/EPDM blend is massive dilatation shear banding. Similar mechanisms were reported by Zebarjad for a simple PP/EPR (80/20) blend.^2,38,39

Fig. 11 Optical microscopy images of sub-surface damaged zones for un-notched films (a and c) and SEM micrographs taken from the sub-surface region of the EWF test specimen (b and d). Un-vulcanized PP/EPDM (a and b) and vulcanized PP/EPDM (c and d).

It can be seen from Fig. 11c and d that the size of the voids has decreased in the vulcanized blend compared with those in the un-vulcanized blend. This could be related to the improved cohesive energy of the EPDM particles and also interfacial adhesion of the PP and EPDM phases due to the vulcanization process.⁹ However, nanovoid formation and cavitation in rubber particles could lead to promoting shear yielding and plastic deformation of adjacent matrix material. A dense and extensive damage zone of the vulcanized blend indicates severe deformation in this area in comparison with the un-vulcanized blend. Therefore one could claim that in the vulcanized blend, because of improved interfacial adhesion between the PP and EPDM phases, the development of dilatation shear bands through repeated particle debonding is suppressed and here it seems that the main mechanism is shear yielding of the PP matrix. This is also evident from more developed stress whitening effects observed ahead of the arrested crack-tip of the vulcanized sample than the un-vulcanized one (Fig. 10c and see also Fig. 7a and b).

Fig. 12 shows SEM images of fractured surfaces of the un-vulcanized and vulcanized PP/EPDM blends with and without PP-g-MA, under the EWF test at two magnifications. As shown in this figure, plastic deformation and stretching of the matrix ligaments followed by necking and twisting are the dominant deformation mechanisms of these samples. Shear-induced size reduction during vulcanization,⁴⁰ and also improved interfacial adhesion between the PP and dispersed EPDM phases along with more resistive EPDM particles against tearing for the vulcanized samples compared with the un-vulcanized ones, greatly increase the resistance to crack initiation and subsequent crack propagation in the form of larger w_e and w_e,n,t values for the former system than the latter one (Table 2). The above mentioned difference in the deformational micro-mechanisms of the un-vulcanized and vulcanized PP/EPDM blends causes a different fracture pattern on the surface of the deformed samples, as can be seen in Fig. 12.

Fig. 12 SEM micrographs of the fracture surface of the EWF test specimens. (a and a′) Un-vulcanized PP/EPDM, (b and b′) vulcanized PP/EPDM, (c and c′) un-vulcanized PP/EPDM/PP-g-MA and (d and d′) vulcanized PP/EPDM/PP-g-MA under the EWF test at two magnifications.

It is relatively hard to differentiate in fracture mechanisms that occurred in samples with and without PP-g-MA based on the micrographs in Fig. 12. However, higher fracture toughness of PP/EPDM/PP-g-MA compared with the PP/EPDM blend is probably due to higher interfacial adhesion between the PP and EPDM phases in the former blend than the latter one.

Optical microscopy images of deformation zones developed ahead of the arrested crack-tip for the un-vulcanized and vulcanized PP/EPDM/MWCNT blend-nanocomposites are shown in Fig. 13. Similar to the unfilled samples, vulcanization caused to an intensified and large damage zone infront of the crack-tip of the MWCNT filled blend. Fig. 13 reveals that the MWCNT aggregates act as stress concentrator sites, so that large MWCNT aggregates located in a highly deformed zone are debonded from the matrix (shown by arrows). Therefore, large MWCNT aggregates could have the main role in unstable crack growth through crack nucleation in weakened zones caused by debonding from the matrix.

Fig. 13 Optical microscopy images of deformation zones that developed ahead of the arrested crack-tip. (a) Un-vulcanized PP/EPDM/MWCNT and (b) vulcanized PP/EPDM/MWCNT.

The SEM micrographs taken from the EWF test fractured surfaces of different blend-nanocomposites are shown in Fig. 14. The low magnification micrographs of the MWCNT filled samples display less disturbed surfaces as compared with the unfilled samples (Fig. 12). This is probably due to hindrance of the fibril yielding process imposed by the MWCNTs. A similar observation has been reported for PP-based nanocomposites containing 0.5 wt% MWCNTs.¹² This process manifests itself in the form of a larger yielding-related specific work of fracture (w_e,y) for all the blend-nanocomposites than their unfilled counterparts, as depicted in the results of Table 2.

Fig. 14 SEM micrographs of the fracture surface of the EWF test specimens. (a and b) Un-vulcanized PP/EPDM/MWCNT, (c and d) vulcanized PP/EPDM/MWCNT, (e and f) un-vulcanized PP/EPDM/MWCNT/PP-g-MA and (g and h) vulcanized PP/EPDM/MWCNT/PP-g-MA under the EWF test at two magnifications.

A decrement of w_e and an increment of βw_p for the un-vulcanized and vulcanized blend-nanocomposites compared with the un-vulcanized and vulcanized blends, respectively, could be related to two different micro- and nano-mechanisms induced by the MWCNTs: on one hand, the large sized MWCNT aggregates act as crack initiation sites which reduce the crack initiation resistance of the material (Fig. 15a and b) and on the other hand, developing the individual MWCNT impregnated fibrils, which have high strength, arrests the crack propagation (Fig. 15a and c). The former micro-mechanism leads to a decrease of w_e (the term related to the resistance to crack initiation) while the latter nano-mechanism leads to an increase in βw_p (the term related to the resistance to crack propagation) through blocking the growing voids. It should be noted that this type of crack propagation is irregular, which is evident from the waviness in the load–displacement curves of the blend-nanocomposites during the crack propagation stage (Fig. 5). Arresting of the crack propagation by nanoclay impregnated fibrils has been reported by Saminathan.⁴¹

Fig. 15 SEM micrographs of the fracture surface of the EWF test specimens. (a) Un-vulcanized PP/EPDM/MWCNT and (b and c) vulcanized PP/EPDM/MWCNT. Individual MWCNTs are shown by arrows and a MWCNT aggregate is marked by the dotted circle in (a).

Noting the above discussed micro- and nano-mechanisms, it is reasonable that PP-g-MA increases both the w_e and βw_p parameters of the blend-nanocomposites through decreasing the size of large aggregates (favored sites to initiate large cracks) and increasing the number of dispersed individual MWCNT ropes. On the other hand, the results of Table 2 indicate that w_e,n decreases in the presence of MWCNTs. This suggests, despite the presence of the MWCNT impregnated fibrils that potentially could have the main role in the increment of tear resistance of the sample, the MWCNT aggregates act as defects and facilitate the tearing process.

It is worth noting that since in PP-based nanocomposites the dominant fracture mechanism is crazing, the MWCNT aggregates increase the fracture toughness of the resulting materials, through acting as nucleation sites for crazes/micro-cracks, whereas in blend-nanocomposites, with shear banding followed by shear yielding as the main source of energy dissipation, the presence of MWCNT aggregates has a negative impact on the fracture toughness of the resulting systems.

Fig. 16 shows SEM images of the sub-surface for the un-vulcanized PP/EPDM/MWCNT blend-nanocomposite sample, fractured in the EWF test. According to this figure, it can be concluded that the presence of MWCNTs in the un-vulcanized PP/EPDM blend has not changed the nature of the micro-mechanisms of deformations related to the EPDM particles and also dilatation shear banding occurs in the un-vulcanized blend-nanocomposites similar to the unfilled counterpart.

Fig. 16 SEM micrographs taken from the sub-surface region of the EWF test specimen for un-vulcanized PP/EPDM/MWCNTs with different magnifications.

4. Conclusions

Fracture behavior and deformation mechanisms of MWCNT filled PP/EPDM blends were studied using the EWF methodology and evidenced using detailed electron and optical microscopy analysis. The results showed that dynamic vulcanization as well as the addition of PP-g-MA improved the essential and plastic work of fracture of the PP/EPDM blends and the PP/EPDM/MWCNT blend-nanocomposites. The results illustrated that the main fracture mechanism in the un-vulcanized samples was the formation of dilatation bands, while in vulcanized samples, the development of dilatation shear bands due to repeated debonding was suppressed and in these samples formation of nano-scale voids in the rubber particles led to promoting shear yielding of the adjacent matrix and a dense plastic.

In the blend-nanocomposites, two conflicting MWCNT induced mechanisms were observed: large MWCNT aggregates acted as favored sites for crack initiation that led to a decrease of w_e and on the other hand, individual MWCNT impregnated fibrils arrested the crack propagation that caused an increase in βw_p through blocking the growing voids. In this case, PP-g-MA displayed a beneficial role on the fracture toughness of the blend-nanocomposites. The energy partitioning approach revealed an increase in both the energy consumed for yielding (w_e,y) and the energy dissipated during ductile tearing (w_e,n) of the compatibilized systems compared with the uncompatibilized ones.

References

1. A. K. Gupta and S. N. Purwar, J. Appl. Polym. Sci., 1986, 31, 535–551.
2. S. M. Zebarjad, A. Lazzeri, R. Bagheri, S. M. Seyed Reihani and M. Frounchi, Mater. Lett., 2003, 57, 2733–2741.
3. X.-G. Tang, R.-Y. Bao, W. Yang, B.-H. Xie, M.-B. Yang and M. Hou, Eur. Polym. J., 2009, 45, 1448–1453.
4. F. Chen, B. Qiu, B. Wang, Y. Shangguan and Q. Zheng, RSC Adv., 2015, 5, 20831–20837.
5. B. C. Kim, S. S. Hwang, K. Y. Lim and K. J. Yoon, J. Appl. Polym. Sci., 2000, 78, 1267–1274.
6. I. Hejazi, F. Sharif and H. Garmabi, Mater. Des., 2011, 32, 3803–3809.
7. M. Khodabandelou and M. K. Razavi Aghjeh, Study on Impact Behaviour of Un-vulcanized and Dynamically Vulcanized PP/EPDM/CNT nanocomposites, presented at The 5th International Congress on Nanoscience & Nanotechnology (ICNN2014), Tarbiat Modares University, Tehran, Iran, 22–24 October 2014.
8. N. K. Gupta, A. K. Jain, R. Singhal and A. K. Nagpal, J. Appl. Polym. Sci., 2000, 78, 2104–2121.
9. M. Ishikawa, M. Sugimoto and T. Inoune, J. Appl. Polym. Sci., 1996, 62, 495–1502.
10. J. Huang and D. Rodrigue, Mater. Des., 2015, 65, 974–982.
11. E. Logakis, E. Pollatos, C. Pandis, V. Peoglos, I. Zuburtikudis, C. G. Delides, A. Vatalis, M. Gjoka, E. Syskakis, K. Viras and P. Pissis, Compos. Sci. Technol., 2010, 70, 328–335.
12. M. Ganß, B. K. Satapathy, M. Thunga, R. Weidisch, P. Potschke and D. Jehnichen, Acta Mater., 2008, 56, 2247–2261.
13. D. Bikiaris, Materials, 2010, 3, 2884–2946.
14. K. Prashantha, J. Soulestin, M. F. Lacrampe, P. Krawczak, G. Dupin and M. Claes, Compos. Sci. Technol., 2009, 69, 1756–1763.
15. P. Zhao, K. Wang, H. Yang, Q. Zhang, R. Du and Q. Fu, Polymer, 2007, 48, 5688–5695.
16. M. Hemmati, G. H. Rahimi, A. B. Kaganj, S. Sepehri and A. M. Rashid, J. Macromol. Sci., Part B: Phys., 2008, 47, 1176–1187.
17. M.-K. Seo, J.-R. Lee and S.-J. Park, Mater. Sci. Eng., 2005, 404, 79–84.
18. B. K. Satapathy, M. Ganß, R. Weidisch, P. Pötschke, D. Jehnichen, T. Keller and K. D. Jandt, Macromol. Rapid Commun., 2007, 28, 834–841.
19. K. Prashantha, J. Soulestin, M. F. Lacrampe, M. Claes, G. Dupin and P. Krawczak, eXPRESS Polym. Lett., 2008, 2, 735–745.
20. L. Valentini, J. Biagiotti, M. Kenny and M. A. L. Manchado, J. Appl. Polym. Sci., 2003, 89, 2657–2663.
21. L. Liu, Y. Wang, Y. Li, J. Wu, Z. Zhou and C. Jiang, Polymer, 2009, 50, 3072–3078.
22. Fracture mechanics testing method for polymers, adhesive and composites, ed. E. Clutton, D. R. Moore, A. Pavan and J. G. Williams, Elsevier Science Ltd, 2001, pp. 177–195.
23. G.-M. Kim and G. H. Michler, J. Appl. Polym. Sci., 1996, 60, 1391–1403.
24. K. B. Broberg, Int. J. Fract., 1968, 4, 11–18.
25. B. Cotterell and J. K. Reddel, Int. J. Fract., 1977, 13, 267–277.
26. J. Karger-Kocsis and E. J. Moskala, Polym. Bull., 1997, 39, 503–510.
27. J. Karger-Kocsis, T. Czigany and E. J. Moskala, Polymer, 1997, 38, 4587–4593.
28. J. Karger-Kocsis, Polym. Bull., 1996, 37, 119–126.
29. T. Barany, J. Karger-Kocsis and T. Czigany, Polym. Degrad. Stab., 2003, 82, 271–278.
30. D. E. Mouzakis, J. Karger-Kocsis and E. J. Moskala, J. Mater. Sci. Lett., 2000, 19, 1615–1619.
31. J. Karger-Kocsis and D. Ferrer-Balas, Polym. Bull., 2001, 46, 507–512.
32. C. Yukun, X. Chuanhui, C. Liming, W. Yanpeng and C. Xiaodong, Polym. Test., 2012, 31, 728–736.
33. J. Karger-Kocsis, V. M. Khumalo, T. Bárány, L. Mészáros and A. Pegoretti, Compos. Interfaces, 2013, 20, 395–404.
34. M. Mehrabi Mazidi and M. K. Razavi Aghjeh, RSC Adv., 2015, 5, 47183–47198.
35. A. R. Bhattacharyya, T. V. Sreekumar, T. Liu, S. Kumar, L. M. Ericson, R. H. Hauge and R. Smalley, Polymer, 2003, 44, 2373–2377.
36. H. Zhang and Z. Zhang, Eur. Polym. J., 2007, 43, 3197–3207.
37. K. Prashantha, H. Schmitt, M. F. Lacrampe and P. Krawczak, Compos. Sci. Technol., 2011, 71, 1859–1866.
38. S. M. Zebarjad, R. Bagheri, A. Lazzeri and S. Serajzadeh, Mater. Des., 2004, 25, 247–250.
39. S. M. Zebarjad, R. Bagheri, S. M. Seyed Reihani and A. Lazzeri, J. Appl. Polym. Sci., 2003, 90, 3767–3779.
40. A. K. Jain, A. K. Nagpal, R. Singhal and N. K. Gupta, J. Appl. Polym. Sci., 2000, 78, 2089–2103.
41. K. Saminathan, P. Selvakumar and N. Bhatnagar, Polym. Test., 2008, 27, 453–458.

---

Footnote

† Electronic supplementary information (ESI) available: The SEM micrographs from the cryofractured surface of vulcanized PP/EPDM in different magnifications. See DOI: 10.1039/c5ra12087j

---