Crystallization, structural and mechanical properties of PA6/PC/NBR ternary blends: effect of NBR-g-GMA compatibilizer and organoclay

2.1 Materials

PA6 with the trade name Akulon from the DSM Co. (The Netherlands), PC with the trade name Makrolon-2800 from the Bayer Co. (Germany), and NBR containing 34% with the trade name Kosyn-KNB 35L from Acrylonitrile from Kumho Petrochemical Co. (Korea) were employed to prepare ternary blend. Glycidile methacrylate (GMA), dicumylperoxide, and styrene comonomer were supplied by Merck (USA) and used in the synthesis compatibilizer according to previous studies [25]. Organically modified montmorillonite with the trade name Cloisite^®30B was supplied by Southern Clay Product Inc. (USA).

2.2 Specimens preparation

All components were dried at 80°C in vacuum oven (MMM-Vacucell, Germany) for 24 h prior to mixing in an internal mixer Brabender (Germany). Preparation of nanocomposites was performed at one step of mixing process, with die temperature of 250°C and screw speed of 60 rpm at 8 min. All ingredients were mixed manually and then charged simultaneously into the chamber of the internal mixer. In order to investigate the role of compatibilizer, NC content, as well as the effect of blend composition, nine samples were prepared by dry blending the whole ingredients, as illustrated in Table 1. As a reference, neat PA6 (P0) and one unfilled sample of PA6/PC/NBR (P1) were also prepared under the same condition for comparison purposes. The samples were prepared in a 300-ml Brabender machine with 80% of fill factor, of which the temperature and chamber velocity were set at 175°C and 50 rpm, respectively. Pelletizing and compression molding were done in a laboratory hot press (GUYER AG, Denmark) at 260°C and 100-bar pressure.

2.3 Crystal structure

Differential scanning calorimetry (DSC) was carried out using 200F3 MAIMA DSC according to ASTM D3418 to study the crystallization behavior of the samples. For a DSC scan, the prepared sample was heated from room temperature (25°C) to 300°C with the heating rate of 15°C/min and held at 300°C for about 1 min, then the heated sample was cooled to 25°C with the heating rate of 10°C/min. A Leica MATS polarized optical microscope (POM) was employed to investigate the spherulitic crystallite morphologies of the ternary nanocomposite samples. Before the optical microscopy test, the samples were heated with 15°C/min rate to 250°C for 2 min and cooled down at the same rate.

2.3 Structural characterization

X-ray diffractions (XRDs) were obtained using a Philips X’pert MPD X-ray diffractometer (The Netherlands) with Cu Kα radiation (λ=1.56 Å) at a generator voltage of 50 kV and 40-mA current at room temperature. The samples were cut as circular discs of 25-mm diameter and 2.5-mm thickness from the compression-molded plaques. A Philips CM-200 transmission electron microscopy (TEM), manufactured in The Netherlands, with 200 kV of voltage was applied to study the morphology of nanocomposites. The phase morphology of nanocomposites was studied by a VegaII XMU SEM equipped with a LEO-440 energy-dispersive X-ray (EDX), manufactured in UK. Prior to the SEM test, the samples were cryogenically fractured in liquid nitrogen and then etched by toluene at room temperature for 24 h to extract the PC phase. The treated samples were coated with a thin layer of gold in sputtering machine. The analysis of the SEM micrographs was carried out using image analysis software. Tensile properties were evaluated using an Instron 6025 universal testing machine (UK) according to ASTM D638 and cross head speed of 10 mm/min. Izod impact strength of specimens was evaluated using a Zwick-5102 Impact Tester (Germany), according to ASTM D256. The depths of notch on specimens are fixed at 2.00±0.01 mm. Dumbbell-shaped specimens for tensile test and standard notched specimens of impact test were cut from the compression-molded plaques.

3.1 Investigation of crystallization

The crystallization kinetics was investigated considering OC and NBR-g-GMA additives’ role on crystallinity and consequent mechanical properties. Diffusion and nucleation were reported as the two main processes that determine the overall crystallization process. These phenomena were influenced by the composition of the polymeric phases [26]. The DSC thermograms in Figure 1 showed the accelerated crystallization process of P1 sample compared to neat PA6. Despite the negative effect of NBR on crystallization process, PC facilitated the crystallization process, leading to higher crystallization temperature. By the addition of NBR-g-GMA as compatibilizer to PA6/PC/NBR blend, it was expected that the compatibilized blend is more homogeneous with relatively improved crystallization process [27]. As an evidence of this expression, increase in the initial crystallization temperature (T_C) of PG3, PG5, and PG7 samples was clearly observed in DSC thermograms. It could be related to the nucleating role of NBR-g-GMA compatibilizer [28] in ternary blend. According to DSC thermograms, acceleration effect of OC in crystallization process was observed by the increase in OC up to 7 wt%.

Figure 1:

DSC thermograms of neat PA6 and PA6/PC/NBR blend with and without OC modifier and compatibilizer.

The summarized results of DSC thermograms can be seen in Table 1. It was observed that the effect of compatibilizer (NBR-g-GMA) on crystallization behavior of the blend was more significant compared with that of OC. The T_C values of PN3, PN5, and PN7 samples did not show any considerable shift in comparison with neat P1. With more attention in DSC thermograms, it was found that T_C decreased with the increase of OC to 7 wt%.

For better consideration of the dependence of T_C, NBR-g-GMA compatibilizer was added to one case of nanocomposite with 5 wt% of OC (sample PNG5). As it could be observed from the DSC thermogram of PNG5, the presence of NBR-g-GMA did not compensate the negative effect of OC for the blends with sufficient amount, as the T_C value was still smaller than in unfilled blends. The reported values of melting temperature (T_m) in Table 1 (curves were not shown here) showed a slight increase in P1 blend compared to the neat PA6 due to the presence of PC in the matrix. However, OC and NBR-g-GMA did not have important effect on T_m. The reported content of crystallinity (X_c) in Table 1 is calculated using the following:

Crystallinity (%)=(ΔHfobs/ΔHfo)×100,

where ΔH_f^obs was the heat of fusion values obtained by DSC and ΔH_f^o was the heat of the fusion value for 100% crystalline PA6 (240 J/g) [19]. The calculated values of X_c in Table 1 revealed higher values of X_c in OC-filled ternary blends even much more than the compatibilized ones. A slight increase in X_c was observed in nanocomposites by increasing the OC content up to 7 wt%, whereas the increase in X_c with compatibilizer was not so obvious. The simultaneous use of OC and compatibilizer in PNG5 sample resulted in a significant reduction on X_c. In order to get proper view of spherulites structures and to confirm the DSC results, the POM micrographs of unfilled PA6/PC/NBR blend were compared with the micrographs of compatibilized (PG5) and OC-filled blend (PN5) in Figure 2. As expected, based on literature [1], NBR-g-GMA and OC caused a reduction in the number average diameter of spherulites from ~17 μm in unfilled PA6/PC/NBR to ~4 and ~7 μm in PN5 and PG5 samples, respectively. In fact, the presence of OC layers with well exfoliation in PA6 matrix hindered the crystallization process and homogenous structure (Figure 2B), but with the presence of compatibilizer, structure was more homogenous, and the average diameter of spherulites showed a slight increase (Figure 2C). For the PNG5 sample, the compatibilizer improved the uniformity of spherulites structure of nanocomposite with small size of spherulites to 5-μm average diameter.

Figure 2:

The images of polarized optical micrographs of (A) P1, (B) PN5, (C) PG5, and (D) PNG5.

3.2 Investigation of OC exfoliation

XRD patterns of the OC (Cloisite^®30B) and PA6/PC/NBR/OC nanocomposites of regions at the surface were shown in Figure 3. The diffraction peak at 2θ=4.9° corresponds to a basal spacing of 18.26 Å, which corresponds to the (d₀₀₁) plane of the Cloisite^®30B.

Figure 3:

X-ray diffractograms of Cloisite^®30B, and PN3, PN5, PN7, and PNG5 nanocomposites.

It could be observed that the diffraction peak of OC in the ternary PA6/NBR/PC nanocomposites shifted to lower 2θ: 4.79, 4.75, 4.69, and 4.66° for PN3, PN5, PNG5, and PN7, respectively. Therefore, the intragallery spacing of nanocomposites (calculated according to Bragg’s law, d=λ/2sinθ_max [12]) was shifted to higher values from 18.68 Å in PN3 sample to 19.18 Å in PN7 sample. The enlarged intragallery of OC could be related to the intercalation of alkyl ammonium and the decline in the electrostatic interactions of clay layers during melt mixing, which facilitates the diffusion of macromolecular chains into the intragalleries [29]. The intensity of the diffraction peak showed a drastic decrease in all nanocomposite samples. From this result, it could be concluded that partially intercalated/exfoliated structure of OC was obtained for all the nanocomposites. Also it could be claimed that the presence of PC and NBR phase in PA6 matrix did not affect the microstructure of the OC.

The mixing torque as an evidence of time-dependent mixture state was measured during the mixing process for all samples (Figure 4). The increase in the mixing torque increased the shear stress and break-up process of the OC agglomerations and caused enhancing in the value of exfoliation. It was observed that by increasing the OC content up to 7 wt%, the steady-state torque increased. The torque result of PNG5 was similar to that of PN5 because NBR-g-GMA in PA6 did not show any special influence the interactions of the matrix and OC [19].

Figure 4:

The mixing torque traces of P1 blend and PN3, PN5, and PN7 nanocomposites.

Complementary to the XRD results, TEM images of the nanocomposites were illustrated in Figure 5 to get a better view of the OC dispersion. A full exfoliation of OC into individual layers along the flow direction during melt mixing was clearly obvious for nanocomposites, except for a few agglomeration or tactoids. Due to higher polarity of PA6 compared to PC and NBR phases, it was expected that OC became fully exfoliated in the PA6 matrix. TEM micrographs for all nanocomposites showed a good dispersion of OC in the each of the three phases, especially the PA6 phase. With increasing OC content (PN7), some agglomerations of OC platelets appeared in TEM micrographs. Also there was no nodule of NBR and PC in the PA6 matrix, which may be due to the improvement in the adhesion and interfacial interaction between matrix and minor polymer phases in the presence of OC [30]. With the addition of NBR-g-GMA compatibilizer to PNG5 sample (Figure 5D), the dispersion and homogeneity of NBR were improved and quantity of agglomerations and tactoids was decreased by enhanced interaction.

Figure 5:

TEM images of (A) PN3, (B) PN5, (C) PN7, and (D) PNG5 samples.

3.3 Investigation of morphology quantification

The SEM micrographs together with respective particle size distributions of prepared samples were shown in Figure 6. The dispersion of PC particles in PA6 matrix is found by etching PC with toluene prior to SEM test. The number average of PC particle size (đ) was calculated from the curves of respective particle size distributions. The calculated particle size (đ) for unfilled blend in Figure 6A was ~1.12 μm. It showed that PC and NBR particles were dispersed very well in the PA6 matrix with a relatively small number average of PC particle size. These observations could be related to high die temperature during process of 250°C, which increased the interfacial adhesion between PC and NBR with PA6 matrix. Due to the increasing effect of OC on the interfacial adhesion and enhanced interaction between ingredients, homogeneity was improved and đ was decreased to ~1.01 and 0.79 μm for PN5 and PN7 samples, respectively.

Figure 6:

SEM images of fracture surfaces of (A) P1, (B) PN5, (C) PN7, (D) PG5, (E) PG7, and (F) PNG5 along with their the curves of respective particle size distributions.

SEM micrographs in Figure 6 showed rougher fractured surface for nanocomposites than in unfilled blend, in relation to improved microstructure, which needs more impact energy to fracture the surface. It can also be observed in the SEM micrograph of PN7 that formation of clay agglomerations caused an increase in the number of stress concentrations in the PA6 matrix. In general, it can be concluded that OC had significant influence on the morphology of PA6/PC/NBR nanocomposites.

The effect of NBR-g-GMA compatibilizer on the number average of PC particle size was similar to OC. The đ values of ~1.03 and 0.87 μm were reported for PG5 and PG7 samples, respectively. However, less rough surface of PG5 and PG7 than that of nanocomposite samples showed the small effect of NBR-g-GMA on interaction of ingredients and resulted in impact energy. The simultaneous effect of OC and NBR-g-GMA could be seen in a SEM micrograph of PNG5 sample (Figure 6F). The roughest surface along with the smallest number average of PC particle size (đ~0.62 μm) was obtained in the PNG5 sample.

Figure 7 represented EDX micrographs of fracture surface of nanocomposites prepared with 3, 5, and 7 wt% of OC. As a confirmation that toluene did not affect the morphology, the SEM images of non-etched surface of PN3, PN5, and PN7 samples were done with EDX micrographs. The light domains represented the position of OC (Si particles) in the matrix. The element content of Si of PN3, PN5, and PN7 was 0.4, 0.58, and 1.13 wt%, respectively. It is observed that the number of OC particles increased by enhancing the content of OC up to 7 wt%. Dispersion of OC was proved to be critical for improving the toughness of nanocomposite [13, 15]. The formation of OC agglomerations was obvious in PN7 samples (Figure 7C). Therefore, reduction in the impact strength was expected by increasing the stress concentration.

Figure 7:

EDX micrographs of (A) PN3, (B) PN5, and (C) PN7 samples along with their SEM images of non-etched surface.

3.4 Investigation of mechanical properties

The results of yield stress for all prepared samples in Figure 8 showed that OC particles caused a slight improvement in the yield stress of unfilled blend by increasing OC content to 7 wt%. In contrast to OC, NBR-g-GMA compatibilizer had a negative effect on the yield stress of PA6/PC/NBR blend, whereas the presence of OC in PNG5 sample relatively compensated the decreasing effect of compatibilizer in the yield stress.

Figure 8:

Yield stress of neat PA6 and PA6/PC/NBR blend with and without OC modifier and compatibilizer.

Comparing the tensile modulus values in Figure 9 showed the intense enhancing effect of OC in the tensile modulus. It was shown that in exfoliated or intercalated morphology, the formation of hydrogen bonds between OC layers and polymer matrix could produce a strong interaction and cause improvement in the tensile modulus of blends [31]. These interactions promoted the value of tensile modulus and the yield stress by declining the segmental mobility of the polymer chains near the surface of OC particles [31]. As represented in Figure 9, tensile modulus had an intense enhancement from 3 wt% to 5 wt% of OC, but in the PN7 sample, tensile modulus did not notably improve with regard to the PN5 sample. It could be related to the formation of OC agglomeration in the PN7 sample. The effect of NBR-g-GMA compatibilizer on increasing tensile modulus of PA6/PC/NBR blend was less than in NC. But with the presence of OC and compatibilizer together in the PNG5 sample, tensile modulus showed the highest value of tensile modulus.

Figure 9:

Tensile modulus of neat PA6 and PA6/PC/NBR blend with and without OC modifier and compatibilizer.

As it can be observed in Figure 10, the values of impact strength reduced with the increase in OC content to 7 wt%. It was found that impact strength of PA6 depended on the presence of rubbery phase in the PA6 matrix, size, and shape of the dispersed phases [32]. In contrast to the improving effect of rubber phase in impact strength (P1 sample compared to P0 sample), OC particles had a decreasing effect on the impact strength of blends, which could be due to stress concentration sites in the matrix. Increase in the content of OC hindered the stress transfer from the dispersed phase to the polymer matrix phase, causing a higher impact energy [33]. The observations in Figure 10 showed that the impact strength of PA6/PC/NBR blends decreased by addition of OC. In contrast to OC, NBR-g-GMA compatibilizer slightly enhanced the impact strength of PA6/PC/NBR blend. The decrease in the impact strength observed in the PNG5 sample with regard to compatibilized blend was an evidence of the negative effect of OC in the impact strength.

Figure 10:

Impact strength of neat PA6 and PA6/PC/NBR blend with and without OC modifier and compatibilizer.