Morphological, dynamic mechanical, rheological and impact strength properties of the PP/PVB blends: the effect of waste PVB as a toughener

Abstract

In this study, compatibilized PP/PVB blends were prepared and their morphological, dynamic mechanical, rheological and Izod impact strength properties were investigated. Thermal and crystallization behavior of the blends were also studied. SEM micrographs showed a reduction in the particle size and size distribution of the PVB domains by adding PP-g-MA up to a certain content, beyond which the particle diameter increased. This is ascribed to saturation of the interface with compatibilizer and micelle formation in the matrix. Furthermore, the migration of some fraction of plasticizer from the PVB phase into the PP matrix was revealed by dynamic mechanical analysis (DMA). Differential scanning calorimetry (DSC) analysis indicated the immiscibility of the blend components and also the role of the PVB particles as nucleating agents for crystallization of the PP. Based on the rheological properties, the role of the compatibilizer in improving compatibility and homogeneity was more pronounced for the blends containing 15 wt% PVB. This is originated from the plasticizer effect on the interactions between the compatibilizer and the blend components. Finally, Izod impact test results showed that the PP/PVB blends had a higher impact strength compared to the pure PP. This implies the fact that waste PVB film could act as an elastomer and improve the impact resistance of brittle polymers. The results of the present work showed that waste PVB could be used in a new application instead of disposal or incineration, which is important from economical and environmental points of view.

---

Introduction

Poly(vinyl butyral) (PVB) is a random copolymer which is produced by the acetalization reaction of poly(vinyl alcohol) (PVA) and butanal. PVA is also prepared by the hydrolysis of poly(vinyl acetate) (PVAc) and as a result, there is a small percent (1–3 wt%) of unreacted vinyl acetate in PVB. Therefore, depending on the stoichiometry of reactants and the extent of hydrolysis in the PVA, a macromolecular chain of PVB includes different molar fractions of vinyl butyral (VB), vinyl alcohol (VA) and vinyl acetate units. The VB units are hydrophobic and increase the toughness of the polymer chain, whereas the VA ones are hydrophilic and provide strong adhesion to many substrates.^1–4 Based upon the molar ratio of the mentioned units in this random terpolymer, PVB can show various properties. For instance, it was shown that higher contents of VB in the chain give rise to lower glass transition temperatures (T_g) (due to the decrease in hydrogen bonding between hydroxyl groups) and higher thermal stability. Also, only PVBs with a high percentage of VA form crystalline structures. Therefore, most commercial PVBs are amorphous.^5,6

PVB has been used in many applications, such as adhesives, coatings, binders, lacquers, etc., but the most important use of this polymer is in laminated safety glasses for automobiles and buildings. In this case, PVB film that contains plasticizer is inserted between sheets of glass. It can absorb mechanical energy and prevent scattering of the glass fragments during crashes by possessing a high level of toughness and adhesion to the glass.^7–10

Considering the production of cars in the world (about 90 million per year)¹¹ and the application of PVB films in the building industry, it could be estimated that about 100 000 tons of PVB film is consumed each year. As a result, large amounts of used laminated safety glasses are disposed of in landfills or incinerated. On the other hand, waste PVB film (trim) is collected in the production line of windscreens.¹² Due to its high price and environmental concerns, it is necessary to find an appropriate method for the recycling and reuse of waste PVB film in suitable applications. The presence of the hydrophobic and hydrophilic functional groups on the PVB polymer chain provides an opportunity to blend it with different polymers. There are several reports^13–17 of blends of pure PVB (without plasticizer) with other polymers. Also, several works have been presented in the literature for preparing blends containing plasticized PVB film. Bendaoud et al.¹⁸ studied the preparation of plasticized PVB/PVC blends for flooring applications as a method for the reuse of waste PVB. In other work, the mechanical properties of recycled PVB/thermoplastic starch (TPS) blends were investigated by Sita et al.¹⁹ Also, Cha et al.^20,21 reported the preparation of PA-6/PVB blends in which waste PVB film was used as an impact modifier for nylon-6. They showed that the blends possess higher impact resistance relative to the pure PA-6. These findings imply that waste PVB film has a similar ability to elastomers, to perform as a toughener for polymers that suffer from low impact properties.

Polypropylene (PP) is a thermoplastic polymer with good processability, mechanical and thermal properties, stiffness, etc., but it shows poor impact resistance.^22,23 Regarding the good mechanical properties of the PVB film, it appears that there is a possibility for using waste PVB and related trims for the toughening of PP. This is very important from economical and environmental points of view. Therefore, the present work was undertaken to find another plausible solution for waste PVB film and to reuse it in new blends. In this study, PP blends containing waste PVB film (trim) were prepared and their morphological, rheological, dynamic mechanical and Izod impact strength properties were investigated. Moreover, to evaluate the compatibility between blend components and the crystallization behavior of the samples, differential scanning calorimetry (DSC) measurements were performed.

Experimental

Materials

PVB film (under the trade name of TROSIFOL, Germany) in the form of trim was used as one of the blend components. The density and melt flow index (MFI) of the PVB film were 1.07 g cm⁻³ and 2 g/10 min (190 °C/2.16 kg), respectively. Commercial PP, SF-060 with an MFI of 6 g/10 min (230 °C/2.16 kg), and a density of 0.9 g cm⁻³ was purchased from Polynar Co. (Tabriz, Iran). Maleic anhydride grafted polypropylene (PP-g-MA) with 0.1 wt% grafted maleic anhydride, which was provided by Solvay (Belgium), was used as a reactive compatibilizer. It had a density of 0.9 g cm⁻³ and an MFI of 64 g/10 min (230 °C/2.16 kg).

Sample preparation

The PP/PVB blends were prepared via melt compounding using a Brabender internal mixer (WTH 55, Germany). The temperature, mixing time and screw speed were 200 °C, 10 min and 50 rpm, respectively. The samples for different tests were prepared by compression molding in a Toyosiki Press (Japan) at 200 °C and a pressure of 25 MPa. The compositions of the prepared blends are presented in Table 1.

Table 1 Compositions of the blends

Sample	PP (wt%)	PVB (wt%)	PP-g-MA^a (wt%)
a Weight percent of the PP-g-MA is on the basis of the blend content.
PP	100	0	0
PP/PVB15	85	15	0
PP/PVB15/3com	85	15	3
PP/PVB15/5com	85	15	5
PP/PVB15/10com	85	15	10
PP/PVB30	70	30	0
PP/PVB30/5com	70	30	5
PP/PVB30/10com	70	30	10

---

Characterization

Morphological analysis. Scanning electron microscopy (SEM) was used to examine the microstructure of the prepared blends. The samples were cryofractured in liquid nitrogen and etched with ethanol for 9 h to extract the PVB phase. Then, the blends were sputter coated with gold and examined using a Vega Tescan model microscope. To analyze the SEM images, JMicroVision software was used. In this way, we calculated number-average (D_n) and volume-average (D_v) diameters and the polydispersity (PD) of the PVB particles using the following equations:where n_i is the number of PVB particles having a diameter D_i.

Fourier transform infrared (FT-IR) spectroscopy. FT-IR spectra from 400 to 4000 cm⁻¹ were recorded using an FT-IR spectrometer (Equinox 55-LSI 01, Bruker). The measurements were performed on thin films of the samples.

Dynamic mechanical analysis (DMA)

Dynamic mechanical analysis (DMA) of the PP/PVB blends was performed in bending mode at a frequency of 1 Hz, a rate of 5 °C min⁻¹ and temperatures ranging from −80 to 155 °C under a N₂ atmosphere using a DMA-TRITON (TRITEC 2000, England) apparatus. The dimensions of the rectangular bars were 30 mm × 10 mm × 1 mm.

Differential scanning calorimetry (DSC) measurements

Thermal properties of the samples were studied by a Mettler DSC (STAR^e SW 12.00). The specimens were first heated from 30 to 200 °C at a rate of 10 °C min⁻¹ under nitrogen flow and annealed at 200 °C for 5 min to eliminate any thermal history. The specimens were then cooled to −30 °C at a cooling rate of 10 °C min⁻¹ and held at −30 °C for 5 min. Finally, they were reheated to 200 °C at the same rate. The crystallinity (X_c (%)) of the PP in the samples was calculated using the following equation:where ΔH_m is the observed melting enthalpy of PP in the samples, ΔH⁰_m is the melting enthalpy of 100% crystallized PP (209 J g⁻¹ (ref. 24)) and ω_PP is the weight fraction of PP in the samples.

Rheological measurements

The rheological properties of the samples were studied using an Anton Paar Physica (MCR 300 model) shear rheometer. Parallel plate geometry with a plate diameter of 25 mm and a gap of 1 mm was used. To determine the linear viscoelastic region, strain sweep experiments were conducted at a frequency of 1 rad s⁻¹. Then, frequency sweep tests were carried out between 0.01 and 600 rad s⁻¹ at a strain amplitude of 5%. All tests were performed at 200 °C under a nitrogen atmosphere.

Notched Izod impact test

The Izod impact strength of the notched samples was measured using a Zwick impact tester (USA) at room temperature according to ASTM D256. The test was conducted with a 1 J pendulum.

Results and discussion

Morphological analysis

SEM micrographs of the PP/PVB blends with various compatibilizer loadings are shown in Fig. 1 and 2. The micrographs indicate that the PVB particles are nearly spherical in shape. As observed, the uncompatibilized blends show a coarse morphology with a broad size distribution of the dispersed phase. It is more pronounced in the case of the PP/PVB30 blend, in which there are particles with a larger size compared to the PP/PVB15 blend. The calculated D_n, D_v and polydispersity (PD) values for the PVB particles in the blends are presented in Table 2. The D_v and PD values for the PP/PVB30 blend are 4.46 μm and 2.69, respectively. The morphology of the uncompatibilized PP/PVB blends implies immiscibility between the dispersed and matrix phases as a result of weak adhesion between PP and PVB.

Fig. 1 SEM micrographs for PP/PVB15 blends containing (a) 0, (b) 3, (c) 5 and (d) 10 wt% compatibilizer. Ethanol was used to etch the fracture surface of the samples.

Fig. 2 SEM micrographs for PP/PVB30 blends containing (a) 0, (b) 5 and (c) 10 wt% compatibilizer. Ethanol was used to etch the fracture surface of the samples.

Table 2 Quantitative values for the PP/PVB blends obtained from SEM micrographs

Sample	Dn (μm)	Dv (μm)	PD
PP/PVB15	1.38 ± 0.69	2.27	1.64
PP/PVB15/3com	0.96 ± 0.50	1.64	1.70
PP/PVB15/5com	1.18 ± 0.65	2.04	1.73
PP/PVB15/10com	0.92 ± 0.47	1.65	1.79
PP/PVB30	1.66 ± 1.47	4.46	2.69
PP/PVB30/5com	1.03 ± 0.52	1.81	1.74
PP/PVB30/10com	1.36 ± 0.84	2.68	1.97

---

Incorporation of PP-g-MA into the blends decreased the particle size of the PVB. For instance, adding 3 wt% compatibilizer to the PP/PVB15 blend led to a reduction of D_v from 2.27 to 1.64 μm. The decrease in particle size and the more homogeneous distribution of the PVB domains in the PP matrix by adding PP-g-MA may be attributed to the enhancement of interfacial bonding between PP and PVB by reactive compatibilization that leads to suppression of coalescence. In fact, the compatibilizer localizes in the interface and a copolymer could be formed by an esterification reaction between the –OH functional groups of the PVB chain and the maleic anhydride groups of PP-g-MA.^17,25 The chemical reaction between PVB and PP-g-MA is illustrated in Fig. 3. The chemical reaction will be investigated in the FT-IR analysis. In addition to the chemical reactions, there are physical interactions in the form of hydrogen bonding and di-polar di-polar interactions between the oxygen containing groups of PVB and maleic anhydride in the interface.¹⁷ On the other hand, olefinic segments of the compatibilizer generate entanglements with the hydrocarbon chains of PP. Owing to the physical and chemical interactions, adhesion between two phases increases and PVB particles disperse in the matrix finely and homogenously.

Fig. 3 Chemical reaction between PVB chains and PP-g-MA.

Based on the SEM observations, it could be found that the addition of a further amount of the compatibilizer to the blend increases the size of the PVB particles. Similar results have been reported for compatibilized PP/PS²⁶ and PP/PET blends.²⁷ The increase was observed at 5 and 10 wt% of the PP-g-MA content for the PP/PVB15 and PP/PVB30 blends, respectively. So, it appears that interface saturation with compatibilizer occurred when the PP-g-MA content was about 3 wt% (for PP/PVB15) and 5 wt% (for PP/PVB30), respectively. Actually, above these critical contents, excess compatibilizer migrates to the PP matrix. This in turn decreases the viscosity and elasticity of the matrix and hence a lower shear stress is imposed on the dispersed phase particles, leading to increase in the average particle size.^26,27 In fact, the presence of PP-g-MA in the PP matrix can affect the blend morphology which is developed in the melt state. For the PP/PVB15/10com blend, the particle size decreased again. This could be attributed to the formation of a network like structure that increases the viscosity. Actually, when the content of the compatibilizer is 5 wt%, a few molecules migrate to the matrix and act as a plasticizer, while at 10 wt%, a higher amount of PP-g-MA is present in the matrix. In this case, micelles of the PP-g-MA create physical interactions with the PP matrix. This leads to an increase in the number of trapped PP chains that in turn form interconnections between PVB particles. In other words, interactions among the PVB dispersed phase, interphase and PP chains generate a network like structure that decreases the particle size of the PVB domains.

FT-IR analysis

The FT-IR spectra of PVB, PP-g-MA and the compatibilized PP/PVB blend are shown in Fig. 4. In the spectrum of PVB, a broad band at 3461 cm⁻¹ corresponds to the hydroxyl group of the VA units, while the peaks at 1053 cm⁻¹ and 1150 cm⁻¹ are the symmetric and asymmetric stretching vibration of C–O–C (butyral ring).^4,28,29 Moreover, the sharp peak at 1733 cm⁻¹ is assigned to the C=O bond related to both of acetate groups in the polymer chain and the plasticizer of the PVB film. In fact, the commercial PVB films contain about 30 wt% of tri(ethylene glycol)-bis-(2-ethylhexanoate) (TEGB) as plasticizer.^8,30 For PP-g-MA, the characteristic peaks of anhydride appear at 1784 cm⁻¹ (symmetric stretching) and 1858 cm⁻¹ (asymmetric stretching).³¹ These peaks are weak due to the low content of the grafted maleic anhydride.

Fig. 4 FT-IR spectra of (a) PP-g-MA, (b) PVB, and (c) compatibilized PP/PVB blend.

To track the reaction between the OH groups of the PVB and MA groups of the compatibilizer and hence, the formation of the ester linkage (Fig. 3), the peaks corresponding to the hydroxyl and anhydride groups were monitored. It is necessary to note that the probable formed ester linkage shows a peak in the region of 1730–1740 cm⁻¹, but this is covered by the sharp peak of the plasticizer. As a result, it is impossible to follow this as an indicator of the occurrence of the reaction. Another important point is that in the blending of PP and PVB, because of the plasticizer migration (which will be discussed later), the intensity of the C=O peak of the plasticizer decreases. During blending of the components, the intensity of the peaks corresponding to the butyral ring (C–O–C) doesn’t change, while the intensity of the OH band decreases due to the chemical reaction. Therefore, the possibility of the chemical reaction and the degree of conversion could be studied by calculating the ratio of intensity of the hydroxyl peak to that of the peaks of the butyral ring (I_O–H/I_C–O–C). A decrease in this ratio indicates the presence of a chemical reaction between components.³² The ratio for the PVB film and PP/PVB/PP-g-MA blend are 0.9 and 0.57, respectively, implying that the hydroxyl groups reacted with the MA groups of the compatibilizer. The degree of conversion (DC) of the OH groups could be calculated in the same way as presented elsewhere³² according to the following equation:

The value of DC was around 37%, suggesting that about one third of the OH groups in the PVB chain were converted to ester linkages. Furthermore, the disappearance of the anhydride peaks in the blend spectrum is another proof for the ring opening of MA in the reaction with OH functional groups.

Dynamic mechanical analysis (DMA)

Investigation of the dynamic response and compatibility between components in the blends could be carried out by dynamic mechanical analysis (DMA). Fig. S1 and S2† show storage modulus and tan δ curves of PVB and PP, respectively. The storage modulus curve of PVB reveals that at low temperatures (below 0 °C), the polymer is in the glassy region (with an E′ value of about 10⁹ Pa). After that, the transition region begins and the storage modulus decreases considerably and finally, a rubbery region is observed at temperatures higher than approximately 35 °C. The E′ values in this region are in the order of 10⁶ Pa. In the transition region, especially 20–40 °C, tan δ values are close to unity. This means that the viscoelasticity of PVB is more prominent in this region.

As seen, the plasticized PVB shows a peak at 29.6 °C in the tan δ curve that corresponds to the glass transition temperature (T_g). The reported T_g in the literature for pure PVB was in the temperature range between 70 and 85 °C, depending on the structure of this terpolymer. Cascone et al.¹⁷ and Arayachukiat et al.³³ observed the T_g for PVB at 87 and 80 °C, respectively. Actually, it can be said that the presence of the plasticizer in the PVB film reduced T_g of the polymer to around 30 °C. Also, the tan δ curve of the PP indicates that T_g is 10.1 °C for this polymer.

The tan δ curves for the blends are displayed in Fig. 5. Based on the tan δ curves, it could be observed that the difference between the T_g values of the PP and PVB doesn’t reduce after blending. This is therefore a hint at the incompatibility of the blend components. On the contrary, T_g of PP shifted to lower temperatures (4.4 °C) and that of PVB moved to higher temperatures (55.2 °C) unexpectedly by adding 15 wt% of the PVB to the PP matrix. In the PP/PVB30 blend, the T_g values for the PP and PVB were −0.3 and 53.5 °C, respectively. These changes in T_g resulted from the migration of the plasticizer from PVB particles to the PP matrix. In fact, at processing conditions (high temperature and shear), plasticizer molecules diffuse through the PVB phase and consequently, some portion evaporates and the other portion localizes in the PP matrix. This phenomenon leads to an increase in the segmental movements of the PP chains and in turn, reduces the T_g of the polymer. On the other hand, the loss of the plasticizer increased the T_g of the PVB domains. The lower T_g value for PP in the PP/PVB30 blends compared to the PP/PVB15 ones indicates that incorporation of higher amounts of PVB in the blend causes greater transfer of the plasticizer into the PP matrix. According to Fig. 5, it could be found that there are similar values of T_g for uncompatibilized and compatibilized blends.

Fig. 5 Temperature dependence of tan δ for the PP/PVB blends. For greater clarity, the curves were shifted from their original locations.

Differential scanning calorimetry (DSC)

To investigate the miscibility between the blend components more precisely and also to determine the influence of addition of the PVB and compatibilizer on PP crystallization and melting, DSC measurements were performed. The DSC traces showing the glass transitions of the PP, PVB and different PP/PVB blends are illustrated in Fig. 6. As can be seen, PVB shows a glass transition at 20.4 °C. The observed T_g for the plasticized PVB is similar to the value reported by Tsonev et al.³⁴ Also, T_g for the PP appeared at −6.2 °C. Since the PP and PVB phases showed their own T_g in the uncompatibilized and compatibilized blends and there wasn’t any decrease in ΔT_g between the components, we could say that the PP/PVB blends are immiscible.

Fig. 6 DSC curves showing glass transitions of the (a) PVB, (b) PP, (c) PP/PVB15, (d) PP/PVB15/3com, (e) PP/PVB15/5com, (f) PP/PVB15/10com, (g) PP/PVB30, (h) PP/PVB30/5com, and (i) PP/PVB30/10com. To show the transitions with more details, the curves of three samples are presented once more in the lower figure.

From the DSC curves of the blends, it could be found that T_g of the PP phase was shifted to lower temperatures, while that of PVB increased to 45–50 °C. As discussed in the DMA section, this behavior results from the migration of some fraction of plasticizer from PVB particles to the PP phase. It is noteworthy that the T_g values of PP and PVB in the PP/PVB30 blends are lower than those of the PP/PVB15 ones. This is attributed to the presence of a higher weight fraction of the PVB in the former systems, leading to a greater amount of plasticizer migration to the PP phase. On the other hand, with increasing PVB content, the particle size increased, as observed in the SEM micrographs. Therefore, diffusion of the plasticizer molecules through particles is restricted to some degree and as a result, less of the plasticizer exudes from each PVB particle. In other words, although the amount of released plasticizer from each PVB particle decreases by coarsening of the dispersed phase, due to the presence of a higher content of PVB, total migration to the PP phase increases. As observed, there is a good agreement between the DSC curves and DMA analysis.

The melting curves of the samples are shown in Fig. 7. According to this figure, PVB doesn’t show an endothermic fusion peak, implying that this polymer is amorphous. In fact, PVBs with a low content (<60 wt%) of vinyl alcohol units in the polymeric chain don’t form a crystalline structure.¹⁵ Considering the fact that the content of the vinyl alcohol in the chain is about 20 wt% in our sample, like other commercial PVB films,^8,30 this observation is expected. From the heating scan for the PP, an endothermic peak of melting at 167.4 °C was seen. Incorporation of PVB into PP had the effect of decreasing the value of T_m of the PP crystals. Similar to T_g of the PP, this decrease is more remarkable for the PP/PVB30 blends, due to the plasticizer migration to the PP phase in the blends. Furthermore, a comparison of T_m of the compatibilized and uncompatibilized blends suggests that the addition of the compatibilizer has a trivial influence on the melting peak temperature of PP.

Fig. 7 Melting behavior of (a) PVB, (b) PP, (c) PP/PVB15, (d) PP/PVB15/3com, (e) PP/PVB15/5com, (f) PP/PVB15/10com, (g) PP/PVB30, (h) PP/PVB30/5com, and (i) PP/PVB30/10com.

Fig. 8 presents the DSC crystallization curves of the PP, PVB and PP/PVB blends with different compatibilizer contents. As expected, the curve for PVB is smooth and no exothermic peak can be seen. Also, neat PP shows an onset temperature (T_onset) of crystallization at 116 °C and a crystallization peak temperature (T_c) at 111.8 °C. The cooling scan for the uncompatibilized PP/PVB blends demonstrated that T_onset and T_c for the PP crystals didn’t change upon the addition of the PVB to the matrix. Besides, the addition of PP-g-MA to the blends affected the nonisothermal crystallization of the PP and considerably increased the aforementioned temperatures. For instance, T_onset and T_c for the PP/PVB15/5com blend are 121.4 and 117.2 °C, respectively. This may be ascribed to the improvement of the interfacial adhesion between phases due to the role of the PP-g-MA. Therefore, the size of the PVB particles reduces and a more homogenous dispersion could be obtained in the PP matrix. The fine PVB particles, which were covered with compatibilizer molecules, could act as heterogeneous nucleation agents, causing an increase in crystallinity. The X_c (%) values for the PP and PP/PVB15/5com blend are 48.9 and 52.2, respectively. These observations are in accordance with results of other reports that studied the effect of the minor phase and compatibilizer on the crystallization behavior of the matrix.^35–39 For HDPE/PA-6,³⁵ HDPE/NBR,³⁶ PP/PA-66,³⁷ PP/Novolac,³⁸ and PP/PET blends,³⁹ it has been found that the presence of HDPE-g-MA and PP-g-MA enhanced the effectiveness of the dispersed phase particles as nucleating agents.

Fig. 8 Crystallization behavior of (a) PVB, (b) PP, (c) PP/PVB15, (d) PP/PVB15/3com, (e) PP/PVB15/5com, (f) PP/PVB15/10com, (g) PP/PVB30, (h) PP/PVB30/5com, and (i) PP/PVB30/10com.

Rheological properties

Fig. 9 shows the linear viscoelastic properties of the PP/PVB15 blends containing different amounts of the compatibilizer. The addition of 3 wt% compatibilizer to the blend increased the complex viscosity, storage and loss moduli. These increases originate from the effect of PP-g-MA in lowering the interfacial tension between the PP and PVB phases that leads to a decrease in interphase slippage, and hence an increase in viscosity.⁴⁰ On the other hand, different chemical and physical interactions of the compatibilizer with blend components at the interface (as described in the morphological analysis), affected the storage modulus of the PP/PVB blend. The compatibilizer effect on the complex viscosity of the blends is more significant at lower frequencies. In fact, the portion of interactions at the interface (such as entanglements of the compatibilizer and PP chains) diminishes at high shear rates and consequently, the degree of interconnectivity between the dispersed phase domains decreases.

Fig. 9 Complex viscosity (top), storage (middle) and loss moduli (bottom) as a function of frequency for PP/PVB 15 blends.

By the further incorporation of compatibilizer, the conditions were changed and the complex viscosity decreased. Similar behavior was evident for the storage and loss moduli of the blend. It appears that saturation of the interface occurred at 3 wt%. So, at higher contents of PP-g-MA, excess compatibilizer molecules migrated to the PP matrix. Generation of compatibilizer micelles is a consequence of migration to the matrix. The micelles could act as a plasticizer and reduce the viscosity and modulus of the blend system.⁴¹ As can be seen, there is a good agreement between the SEM analysis and the rheological properties for these blends.

The rheological responses of the PP/PVB30 blends are exhibited in Fig. 10. Unlike the PP/PVB15 system, there is a continuously decreasing trend in complex viscosity with the addition of 5 and 10% wt PP-g-MA to these blends. Khonakdar et al.⁴² reported similar results of a drop in viscosity by incorporation of the compatibilizer into immiscible PET/PP blends. Also, Ezzati et al.⁴³ found that compatibilized TPO blends had lower complex viscosity than an uncompatibilized blend. These results may be ascribed to the weaker interactions between the compatibilizer and blend components at the interface. According to the explanations in the DMA and DSC sections, higher amounts of plasticizer existed in the PP matrix and at the interface for the PP/PVB30 blend compared to the PP/PVB15 blend. Therefore, plasticizer molecules could considerably decrease the creation of entanglements between the matrix and PP-g-MA chains. As a result, there are poor interconnectivities between PVB particles in the blend. Another important factor is the lower viscosity of the compatibilizer relative to the matrix. As a whole, it appears that the compatibilizer acts as a lubricant in this situation and causes interlayer slippage.⁴³ As a result, at 5 wt% of compatibilizer, the complex viscosity was lower than that of the uncompatibilized blend. Saturation of the interface occurred at this content. As observed in the morphological observations, at higher loading, the compatibilizer migrated to the matrix and increased the particle size of the PVB phase. So, in the PP/PVB30 blend containing 10 wt% PP-g-MA, additional compatibilizer molecules form micelles in the PP matrix and reduce the complex viscosity more significantly.

Fig. 10 Complex viscosity (top), storage (middle) and loss moduli (bottom) as a function of frequency for PP/PVB 30 blends.

A Cole–Cole plot, which is a graph of imaginary viscosity (η′′) against real viscosity (η′), is a method for the investigation of compatibility and homogeneity of the polymer blends.⁴⁴ Cole–Cole plots for PP/PVB blends are presented in Fig. 11. As is shown, there is a shoulder in the right-hand side of the plot for the uncompatibilized PP/PVB15 blend. This is due to the poor compatibility between PP and PVB phases. The compatibilized PP/PVB15 blends show single semi-circular curves, revealing an improvement in homogeneity and compatibility owing to the role of the PP-g-MA. Since systems with higher compatibility and homogeneity show a more semi-circular shape in Cole–Cole plots,⁴⁵ it appears that the sample with 3 wt% PP-g-MA had the best homogeneity compared with the other blends.

Fig. 11 Cole–Cole plots for the different PP/PVB15 and PP/PVB30 blends.

In the Cole–Cole plots of the PP/PVB30 blends, a deviation from the semi-circular shape and the appearance of the shoulder could be seen because of the presence of a second phase with a different relaxation time.⁴⁶ The shape of the plots signifies that these blends are less homogenous than the PP/PVB15 ones. The SEM images and PD values indicated a broader particle size distribution for the PP/PVB30 blends. In fact, the presence of relatively large particles together with smaller ones decreased the homogeneity of the system.

Another criterion for comparing homogeneity among blends is the frequency at which the G′ and G′′ curves cross each other, known as the dynamic intersection frequency (ω_c). In other words, the lower the intersection frequency, the more homogeneous the blend system.⁴⁷ There is a significant difference among the ω_c values of the PP/PVB15 blends with or without compatibilizer. For the uncompatibilized PP/PVB15 blend, ω_c is around 106 rad s⁻¹, whereas the blends containing compatibilizer show intersection points at frequencies around 59 rad s⁻¹. This indicates a finer particle size and more homogeneity in the compatibilized blends. Furthermore, it signifies an increase in the relaxation time for compatibilized blends.^48,49 Similar results exist for the PP/PVB30 blends. The uncompatibilized PP/PVB30 system shows a ω_c at 106 rad s⁻¹, but the compatibilized samples have ω_c at lower frequencies (59–106 rad s⁻¹). This is a sign of the improved homogeneity upon adding PP-g-MA.

Homogeneity in the polymer blends depends on several parameters, including the viscosity ratio, interactions at the interface, elasticity of the dispersed phase and applied shear rate.⁴⁵ The Cole–Cole plots and dynamic intersection frequencies indicated that the role of the compatibilizer in increasing the compatibility and homogeneity of the prepared blends was more marked for the PP/PVB15 blends. So, these represent more evidence for the stronger interactions of the PP-g-MA with blend components in the interface for the aforementioned blends.

Notched Izod impact strength

The notched Izod impact strengths of PP and the various PP/PVB blends are presented in Table 3. Also, the relative impact strength data are shown in Fig. 12. The relative data were obtained by dividing the impact strength of the blend by that of the PP. The pure PP had an impact strength of 32.94 J m⁻¹ and by introducing 15 wt% PVB into the PP matrix, the amount of fracture energy decreased. This decreasing trend was intensified upon increasing the PVB content in the blend. The PP/PVB30 sample showed a 25% loss in impact strength compared to the PP. The PVB and PP are thermodynamically immiscible due to the significant discrepancy between the polarities of their chemical structures. High interfacial tension between PP and the plasticized PVB leads to significant heterogeneity in the distribution of PVB particles in the matrix. As a result, the applied stress in the impact test couldn’t be transferred to the elastomeric PVB particles appropriately and consequently, the impact strength decreases. These results indicate that without compatibilizer, PVB particles couldn’t be dispersed finely in the PP matrix and therefore PVB doesn’t behave as a toughener in this situation.

Table 3 Izod impact strength data and interparticle distance between neighboring PVB particles in the samples

Sample	Impact strength (J m^−1)	Interparticle distance (μm)
PP	32.94 ± 1.14	—
PP/PVB15	29.42 ± 0.29	1.32
PP/PVB15/3com	32.92 ± 0.89	0.95
PP/PVB15/5com	36.39 ± 0.22	1.18
PP/PVB15/10com	48.01 ± 0.41	0.96
PP/PVB30	24.54 ± 0.93	1.60
PP/PVB30/5com	38.36 ± 0.63	0.47
PP/PVB30/10com	58.87 ± 0.31	0.70

---

Fig. 12 Relative impact strength of the prepared PP/PVB blends.

The addition of PP-g-MA as a compatibilizer to the immiscible PP/PVB blends improved the impact strength. This improvement can be attributed to the role of the compatibilizer in increasing the interfacial adhesion between PP and PVB domains. Therefore, stress is transferred more appropriately and as a result, the impact strength of the PP enhances.

At low contents of PP-g-MA, the increase in impact strength was minimal, but at higher levels, it was significant. When the compatibilizer loading was 10 wt%, the notched Izod impact strengths of the PP/PVB15 and PP/PVB30 blends were 48.01 and 58.87 J m⁻¹, respectively. In other words, about 50 and 80% improvement in impact strength relative to PP was exhibited, respectively. As can be seen, the increase in impact strength is more considerable in the case of the PP/PVB30 blends. By analyzing data in Tables 2 and 3, it could be found that there is no certain relationship between the average particle size and the impact data. The D_n and D_v values for the PP/PVB15/10com sample are 0.92 and 1.65 μm, respectively. The corresponding values for the PP/PVB30/10com blend are 1.36 and 2.68 μm, respectively. However, the latter had a higher toughness. So, the smaller particle size of the PVB phase didn’t mean a higher impact strength. This phenomenon is ascribed to the interparticle distance between PVB particles in the matrix. Wu^50,51 proposed that the surface-to-surface interparticle distance between neighboring dispersed phase particles is the most important determining factor in toughening of polymers. According to this criterion, the interparticle distance (τ) is given by the following equation:

where d is the particle diameter and φ is the volume fraction of the PVB phase. Lower d and higher φ values lead to a lower τ value for the dispersed particles. Table 3 shows the interparticle distance between the PVB particles in the prepared blends. By comparing τ values with the Izod impact strength data, consistency between them could be seen in some cases. As observed, the higher τ value (1.60 μm) in the uncompatibilized PP/PVB30 blend compared with the PP/PVB15 blend (1.32 μm) is the reason for the lower impact strength of the former. Also, at 10 wt% PP-g-MA, the τ values for the PP/PVB30 and PP/PVB15 blends are 0.70 and 0.96 μm, respectively, and the impact strength values for these blends are 58.87 and 48.01 J m⁻¹, respectively. The explanation for this behavior could be presented as follows: when the interparticle distance is high and the PVB particles are far away from each other, the stress fields around each particle don’t interact appropriately. That is why low impact strength is observed. On the contrary, when the PVB particles are close together, the stress fields can overlap and interact extensively and as a result, the capability of the particles to dissipate applied stress enhances.^51,52 It appears that the compatibilizer trapped PP chains in the PP/PVB blends, so the effective volumes of the PVB particles increased.⁵³ On the other hand, the interparticle distance is relatively small. Therefore, a higher impact strength could be obtained.

Comparison of the data in Table 3 shows that in addition to the interparticle distance, another factor must be taken into consideration to explain the impact strength of the blends. In other words, the PP/PVB15/3com and PP/PVB15/10com blends have a similar interparticle distance and different impact strength values. The PP/PVB30/5com blend also has a lower interparticle distance than PP/PVB30/10com, but the latter presented higher impact strength. It appears that another important and determining factor is the role of PP-g-MA in the PP matrix in creating entanglements with PP chains. In the PP/PVB15/5com blend, although the interparticle distance is higher than PP/PVB15/3com, the impact strength is higher. In fact, for the former system, PP-g-MA chains are entangled with PP chains and enhance load transfer in the system during the impact test. By the addition of a higher content of the compatibilizer, more PP-g-MA molecules are present in the matrix and the number of the entanglements increases significantly. Consequently, a network-like structure forms leading to an increase in viscosity (as observed in the rheological properties). Due to this effect, the interparticle distance decreases again and becomes similar to that of the PP/PVB15/3com sample, but the impact strength increases considerably. In fact, the formation of the network-like structure leads to a 50% increase in impact strength compared to the PP/PVB15/3com sample. In the test, this structure enhances the capability of stress transfer in the system. The same behavior exists for the PP/PVB30 blends. Therefore, it could be concluded that the interparticle distance and the role of PP-g-MA in generating the network-like structure are the main parameters that influence the impact strength of the blends.

The impact strength values of the blends are promising for the recycling of waste PVB. In fact, waste PVB could increase toughness of PP and act as an impact modifier for this brittle polymer.

Conclusion

PP blends containing waste PVB film were prepared and their morphological, dynamic mechanical, rheological and impact strength properties were studied. The SEM images, DMA and DSC curves and notched Izod impact test revealed that PP and PVB are immiscible. This is due to the significant difference between their chemical structures. Incorporation of PP-g-MA improved the interfacial adhesion between the blend components and as a result, led to the better and finer dispersion of the PVB particles in the PP matrix. In turn, this caused improvement in the Izod impact strength of the blends. In fact, in the presence of the compatibilizer, the PP/PVB blends had higher toughness compared to pure PP.

DMA analysis indicated that the plasticizer of PVB could migrate to the PP phase during fabrication of the blends and reduce its glass transition temperature. In the DSC measurements, an increase in T_onset and T_c of PP crystallization was seen for the compatibilized bends. This was due to the fact that the PVB particles could act as heterogeneous nucleating agents.

Based on the rheological studies, it was found that the compatibilizer had a higher influence on the compatibility and homogeneity of the PP/PVB15 blends relative to the PP/PVB30 blends. This is attributed to the presence of the plasticizer at the interface and in the PP matrix. In fact, the plasticizer could hinder interactions between the compatibilizer and blend components. Since quantity of the plasticizer in the matrix and at the interface is lower (as detected in the tan δ and DSC curves) for the PP/PVB15 blends, the degree of the interactions is higher in these systems.

The obtained results indicated that waste PVB has capability for use in compatibilized PP/PVB blends. As a method of recycling, it could be useful for solving the issue of huge volumes of waste or scrapped windshields and PVB films.

References

1. A. R. Tripathy, W. Chen, S. N. Kukureka and W. J. MacKnight, Polymer, 2003, 44, 1835–1842.
2. B. Xu, H. A. Toutanji and J. Gilbert, Cem. Concr. Res., 2010, 40, 347–351.
3. M. Hajian, G. A. Koohmareh and M. Rastgoo, J. Appl. Polym. Sci., 2010, 115, 3592–3597.
4. M. Hajian, M. R. Reisi, G. A. Koohmareh and A. R. Z. Jam, Polym. Res., 2012, 19, 1–7.
5. M. D. Fernandez, M. J. Fernandez and P. Hoces, J. Appl. Polym. Sci., 2006, 102, 5007–5017.
6. H. K. Jeong, M. Rooney, D. J. David, W. J. MacKnight, F. E. Karasz and T. Kajiyama, Polymer, 2000, 41, 6003–6013.
7. M. Tupý, D. Měřínská, P. Svoboda, A. Kalendová, A. Klásek and J. Zvoníček, J. Appl. Polym. Sci., 2013, 127, 3474–3484.
8. M. Tupý, D. Meřinská, P. Svoboda and J. Zvoníček, J. Appl. Polym. Sci., 2010, 118, 2100–2108.
9. A. R. Zanjanijam, M. Hajian and G. A. Koohmareh, J. Appl. Polym. Sci., 2014, 131, 40481.
10. A. R. Zanjanijam, M. Hajian and G. A. Koohmareh, J. Macromol. Sci., Part A: Pure Appl.Chem., 2014, 51, 369–377.
11. IOCA, The International Organization of Motor Vehicle Manufacturers, http://www.oica.net/category/production-statistics, accessed November 2015.
12. M. Tupy, P. Mokrejs, D. Merinska, P. Svoboda and J. Zvonicek, J. Appl. Polym. Sci., 2014, 131, 39879.
13. Y. Peng and Y. Sui, Desalination, 2006, 196, 13–21.
14. J. W. Qian, H. L. Chen, L. Zhang, S. H. Qin and M. Wang, J. Appl. Polym. Sci., 2002, 83, 2434–2439.
15. W. Chen, D. J. David, W. J. MacKnight and F. E. Karasz, Polymer, 2001, 42, 8407–8414.
16. W. Chen, D. J. David, W. J. MacKnight and F. E. Karasz, Macromolecules, 2001, 34, 4277–4284.
17. E. Cascone, D. J. David, M. L. Di Lorenzo, F. E. Karasz, W. J. Macknight, E. Martuscelli and M. Raimo, J. Appl. Polym. Sci., 2001, 82, 2934–2946.
18. A. Bendaoud, C. Carrot, J. Charbonnier and C. Pillon, Macromol. Mater. Eng., 2013, 298, 1259–1268.
19. C. Sita, M. Burns, R. Häβler and W. W. Focke, J. Appl. Polym. Sci., 2006, 101, 1751–1755.
20. Y. J. Cha, C. H. Lee and S. Choe, J. Appl. Polym. Sci., 1998, 67, 1531–1540.
21. Y. J. Cha, C. H. Lee and S. Choe, J. Ind. Eng. Chem., 1997, 3, 257–262.
22. C. F. Antunes, M. Van Duin and A. V. Machado, Polym. Test., 2011, 30, 907–915.
23. S. H. Su and H. X. Huang, J. Appl. Polym. Sci., 2009, 112, 3016–3023.
24. Y. Lin, H. Chen, C. M. Chan and J. Wu, Polymer, 2010, 51, 3277–3284.
25. A. Marcinčin, Z. Jurčišinová, E. Borsig, M. Krištofič and T. Marcinčinová, Polym. Adv. Technol., 2001, 12, 461–465.
26. X. Zhao, Y. Huang, M. Kong, Q. Yang and G. Li, RSC Adv., 2014, 4, 59302–59309.
27. M. Entezam, H. A. Khonakdar, A. A. Yousefi, S. H. Jafari, U. Wagenknecht, G. Heinrich and B. Kretzschmar, Macromol. Mater. Eng., 2012, 297, 312–328.
28. B. Yang, R. Liu, J. Huang and H. Sun, Ind. Eng. Chem. Res., 2013, 52, 7425–7431.
29. M. D. Fernandez, M. J. Fernandez and P. Hoces, J. Appl. Polym. Sci., 2006, 102, 5007–5017.
30. E. Corroyer, M. C. Brochier-Salon, D. Chaussy, S. Wery and M. N. Belgacem, Int. J. Polym. Anal. Charact., 2013, 18, 346–357.
31. M. Mehrabzadeh, M. R. Kamal and G. Quintanar, Iran. Polym. J., 2009, 18, 833–842.
32. Z. Okulus, T. Buchwald, M. Szybowicz and A. Voelkel, Mater. Chem. Phys., 2014, 145, 304–312.
33. S. Arayachukiat, M. Siriprumpoonthum, S. Nobukawa and M. Yamaguchi, J. Appl. Polym. Sci., 2014, 131, 40337.
34. T. Tsonev, M. Herzog and S. Nenkova, Open Chem., 2013, 11, 2058–2065.
35. J. Chen, J. W. Chen, H. M. Chen, J. H. Yang, C. Chen and Y. Wang, Composites, Part B, 2013, 54, 422–430.
36. D. Ponnamma, J. George, M. G. Thomas, C. H. Chan, S. Valić, M. Mozetič and S. Thomas, Polym. Eng. Sci., 2015, 55, 1203–1210.
37. Z. Yang and K. Mai, J. Appl. Polym. Sci., 2011, 119, 3566–3573.
38. L. Cui, Y. Zhang and Y. Zhang, J. Polym. Sci., Part B: Polym. Phys., 2006, 44, 3288–3303.
39. C. Wang, W. Dai, Z. Zhang and K. Mai, J. Therm. Anal. Calorim., 2013, 111, 1585–1593.
40. J. George, K. Ramamurthy, K. T. Varughese and S. Thomas, J. Polym. Sci., Part B: Polym. Phys., 2000, 38, 1104–1122.
41. S. H. Jafari, A. Yavari, A. Asadinezhad, H. A. Khonakdar and F. Böhme, Polymer, 2005, 46, 5082–5093.
42. H. A. Khonakdar, S. H. Jafari, S. Mirzadeh, M. R. Kalaee, D. Zare and M. R. Saeb, J. Vinyl Addit. Technol., 2013, 19, 25–30.
43. P. Ezzati, I. Ghasemi, M. Karrabi and H. Azizi, Iran. Polym. J., 2008, 17, 669–679.
44. D. Shi, G. H. Hu, Z. Ke, R. K. Y. Li and J. Yin, Polymer, 2006, 47, 4659–4666.
45. L. A. Utracki, Polymer Blends Handbook, Kluwer Academic Publishers, Netherlands, 2002.
46. J. Cha and J. L. White, Polym. Eng. Sci., 2003, 43, 1830–1840.
47. Y. Chen and H. Li, Polymer, 2005, 46, 7707–7714.
48. Y. T. Sung, M. S. Han, J. C. Hyun, W. N. Kim and H. S. Lee, Polymer, 2003, 44, 1681–1687.
49. P. H. P. Macaubas and N. R. Demarquette, Polymer, 2001, 42, 2543–2554.
50. S. Wu, J. Polym. Sci., 1983, 21, 699–719.
51. S. Wu, Polymer, 1985, 26, 855–863.
52. H. Yang, Q. Zhang, M. Guo, C. Wang, R. Du and Q. Fu, Polymer, 2006, 47, 2106–2115.
53. N. A. Memon and R. Muller, J. Polym. Sci., Part B: Polym. Phys., 1998, 36, 2623–2634.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra04143d

---