Influence of the HDPE molecular weight and content on the morphology and properties of the impact polypropylene copolymer/HDPE blends

Abstract

Two HDPE resins with different molecular weights were blended with one impact polypropylene copolymer with varied weight contents. The influence of molecular weight and content of PE on the morphology and properties of the blends was studied. It is found that by changing the molecular weight and weight content of PE, the morphologies of the blends show great differences. Spherical particles with different structure details and obvious orientation structures were found. Properties, such as impact strength and elongation at break, depend not only the composition but also on the morphology. This study may shed light on the structure–property relationship study in polyolefin impact blends.

---

Introduction

Polypropylene (PP), known as a thermoplastic with an excellent cost-to-performance ratio, is widely used in a variety of areas.¹ However, drawbacks, such as brittleness at low temperature or high deformation speed of homo PP, limit its application. Impact modifiers, such as ethylene-propylene rubbers (EPRs), ethylene-propylene-diene copolymer (EPDM) and the other olefin copolymers, are utilized to improve the toughness by compounding or in-reactor polymerization technology. The heterophasic materials of this type are often called impact polypropylene copolymers (IPCs), which are extensively used in many areas such as automotive and housewares.

HDPE resins are often introduced in IPCs to purposely improve the properties by compounding with IPCs directly or by a multi-reactor polymerization process.^2–10 Correspondingly, the relationship between the structure and properties of this type of multiphase materials has been intensively studied. Addition of HDPE was reported to influence the impact strength,^{2,3,6,7,9,11,12} minimize the magnitude of the modulus decrease,^2,3 and suppress the stress-whitening.^5,7

Mirabella studied IPC/HDPE blends with HDPE weight percent 0–40%.¹¹ A monotonic decrease in the flexural modulus was observed. The notched Izod impact strength (NIS) showed a maximum at HDPE content of 10 wt% and then NIS decreased slightly with increasing HDPE content up to 40 wt%. The morphology of the compression molded samples was observed by atomic force microscopy (AFM) and the shift from a two-phase to a three-phase system was assigned to be responsible for the decrease of NIS at HDPE content great than 10 wt%.

Jang et al. reported a similar phenomenon in IPC/HDPE blends at HDPE content of 0–20 wt%.⁷ NIS increased at low HDPE content (5–10 wt%) and decreased at HDPE content higher than 10 wt%. The authors explained these results by the separation of HDPE from EPR domains as a separated dispersed phase. In Jang et al.'s study, the morphology of the cross-section of the injection-molded tensile bar was observed by TEM; however, the morphology of the surface perpendicular to the cross-section was not shown.

Stehling et al. studied the structure and properties of PP/rubber/HDPE blends.² The influence of mixing sequence and HDPE content on the mechanical properties was reported. It was found that the impact strength increased when HDPE content increased from 6 to 12 wt%. The samples were microtomed, xylene-etched and observed by SEM. Although the impact strength was tested using injection-molded samples, only the SEM micrographs of the compression-molded samples were shown.

Tchomakov et al. found that the morphology depends on PE viscosity and the mixing sequence for PP/PE/ethylene-propylene-diene (EPDM) blends and the mechanical properties change correspondingly.⁶ The dependence of NIS on the PE content was found to be similar to the results of the work by Mirabella and Jang et al. NIS shows a maximum at PE content of 10 wt% for both of the PE resins with different melt indexes (MI, 4.8 and 0.5 g/10 min). A solution of KMnO₄ and acid was used to reveal the detailed structure of the blends, and PP, rubber and PE can be clearly identified, but similar to the situation mentioned above, only the surface perpendicular to the flow direction in the injection-molded samples was analysed.

All the works listed above studied the effects of the addition of HDPE on the impact strength of PP/rubber/HDPE blends and the morphology has been noticed to be a very important factor. However, there are still some drawbacks. First, the specimens used in the impact strength testing are usually injection-molded and thus the morphology of the injection-molded specimens, rather than compression-molded ones, should be observed. Second, the injection-molding is known to induce an anisotropic structure.^13,14 Therefore, the microtome slicing should be conducted both perpendicular and parallel to the flow direction of the melt.

The role of HDPE in influencing the mechanical properties of the PP/rubber/HDPE blends is also widely noticed. It is generally accepted that the morphology change induced by the addition of HDPE is a very important factor. A core–shell like structure has been reported by many researchers.^2–4,6,7,11 The morphological detail of the particles composed of rubber and HDPE is varied, but in most cases, rubber acts as a “shell” and encapsulates HDPE. Based on the core–shell structure, Gahleitner et al. stated that the addition of HDPE increases the “total elastomer volume” of a heterophasic copolymer and thus improves NIS at room temperature to some extent.³

In this study, to elucidate the influence of HDPE on the morphology and mechanical properties, two types of HDPE resins with different molecular weights were used for blending with the IPC. The HDPE resins were blended with an IPC by HDPE content (w_PE) of up to 40% based on total mixture weight. The specimens used in both the mechanical property and morphology tests were injection-molded bars, which makes the relationship between structure and property more reliable. To study this type of structure, the morphology of the plane both parallel and perpendicular to the flow direction of the melt was observed. The morphology of the resulting blends was found to depend greatly on the molecular weight of the PE resins and the PE content, and the mechanical properties changed correspondingly. Some results are different from the reported work. The relationship between structure and properties is discussed.

Experimental section

Materials

The impact polypropylene copolymer (IPC) resin is a commercial product supplied by SINOPEC. IPC is composed of isotactic polypropylene and ethylene-propylene rubber (EPR). The polyethylene (PE) samples used here are both commercial HDPE resins supplied by SINOPEC and denoted by PE-A and PE-B, respectively. As shown in Table 1, the two PE resins have a great difference in molecular weights.

Table 1 Characteristics of the samples

	Mw × 10^3	Mcr (g mol^−1)	Eη(∞) (kJ mol^−1)	η0^c at 200 °C (N s m^−2)
a Calculated by C2 in EPR data and the corresponding data for PP and PE.b Uses the data in the handbook.^15c Calculated by eqn (4)–(6) and the data in the handbook.^15
IPC	301	—	—	—
Soluble fraction	614	5616^a	38^a	2.90 × 10^6
Insoluble fraction	227	7000^b	43^b	6.25 × 10^4
PE-A	222	3500^b	27^b	2.43 × 10^5
PE-B	87	3500^b	27^b	1.00 × 10^4

---

Xylene soluble fraction

CrystEX made by the Polymer Char company was used to obtain the xylene soluble (mainly EPR in the case of this study) fraction (ω_sol) and the weight fraction of ethylene comonomer in the soluble fraction (C2 in EPR). CrystEX is equipped with a special IR sensor for ethylene content determination. The solvent is xylene. The values of ω_sol and C2 in EPR thus obtained are 29.25% and 39.53%, respectively.

Molecular weight and molecular weight distribution

Gel permeating chromatography (GPC, PL-GPC220, Polymer Laboratories, Inc., Great Britain) and an IRS detector (Polymer Char Inc., Spain) are used to measure the molecular weight and molecular weight distribution of the samples. The chromatography columns are three Plgel 10 µm MIXED-B columns connected in series; the solvent and fluid phase is 1,2,4-trichlorobenzene (comprising 0.3 g/1000 mL antioxidant 2,6-dibutyl p-cresol), the column temperature is 150 °C, and the flow rate is 1.0 mL min⁻¹. We used a method set up by our laboratory to analyse the 1,2,4-trichlorobenzene soluble fraction in the IPC sample, which involves dissolving, filtration and GPC experiments. Molecular weight and molecular weight distribution can be readily obtained as well as the relative content of the soluble fraction. The parameters of the insoluble fraction can be obtained by subtraction of the result of the soluble fraction from that of the IPC. Weight-average molecular weights (M_w) of the samples are shown in Table 1.

Blend preparation

IPC and HDPE pellets were first mixed by a mechanical mixer. Then, the mixed pellets were melted, blended, extruded and pelletized by ZSK-25 twin-screw extruder with a barrel temperature of 204–211 °C. A compounded antioxidant of 0.3% by weight was added to suppress thermal decomposition during melt processing. The antioxidant is a mixture of Irganox 1010 and Irganox 168 with a ratio of 1 : 1 by weight. The pellets made thereof were dried at 80 °C for at least 4 h before injection-molding to desirable specimens. The composition and the code of the blends are shown in Table 2, wherein the PE content denotes the weight content based on the total weight of IPC and HDPE.

Table 2 Basic formulations of the IPC/HDPE blends

Sample	HDPE type wt%	HDPE content^a wt%	HDPE content^b wt%	log(η0)^c N s m^−2	λ	Dn µm	Dw µm
a HDPE content in the whole blends of PP/EPR/PE.b HDPE content taken EPR/PE as a whole.c Calculated by eqn (7) and the data in Table 1.d The length of the minor axis of the ellipsoidal particles of A5.e The width of the oriented strips.
IPC	—	0	0	—	46.24	1.17	1.31
A1	PE-A	5.0	0.15	6.30	31.68	1.37	1.58
A2	PE-A	11.5	0.31	6.13	21.56	1.42	1.62
A3	PE-A	20.6	0.47	5.95	14.41	1.39	1.72
A4	PE-A	30.0	0.59	5.82	10.59	1.75	1.99
A5	PE-A	40.0	0.70	5.71	8.25	1.19^d	1.49
B1	PE-B	5.0	0.15	6.09	19.49	1.61	1.83
B2	PE-B	11.5	0.31	5.70	8.09	0.91	1.07
B3	PE-B	20.6	0.47	5.30	3.22	0.75^e	0.94
B4	PE-B	30.0	0.59	5.00	1.59	0.75^e	1.02
B5	PE-B	40.0	0.70	4.75	0.90	0.88^e	1.53

---

Measurement of mechanical properties

The specimens used for mechanical tests were injection-molded using the MA1200/370 injection-molding machine. The barrel temperatures of the injection-molding machine were 200 °C. The notched Izod impact strength (NIS) of the specimens was measured according to ISO180 at 20 °C and −20 °C. The tensile test was measured according to ISO527. For each sample, several parallel measurements were carried out and the average values were recorded.

Morphology

Morphology studies of the phase structure were done using a scanning electron microscope Hitachi S-4800 for etched samples and FEI XL30 for the fractured surface of samples after the impact test. The specimens for SEM observation were cut from injection-molded bars for Izod impact tests and the observation area was the core of the bar. Due to the anisotropic characteristics of the injection-molded bar, it is necessary to mark different directions according to the flow direction and the geometry of the mold. The definition of the directions of the injection-molded bar is designated in Fig. 1. The specimens were cryomicrotomed in the presence of liquid nitrogen, etched by permanganic reagent according to Olley and Bassett's method,¹⁶ sputtered with gold and finally subjected to SEM observation. The morphology of the planes both parallel (XZ plane, as shown in Fig. 1) and perpendicular (YZ plane, as shown in Fig. 1) to the flow direction of the melt (X direction, denoted as FD) were revealed by microtome sections. Thus, the flow-induced morphology can be readily observed.

Fig. 1 Definitions of the directions: X is the flow direction denoted as FD, Y is the transverse direction denoted as TD and Z is the normal direction denoted as ND.

The size of the dispersed phase was evaluated by the Nano Measurer software programmed by Ms Xu Jie in the Department of Surface Chemistry and Catalysis laboratory in the Fudan University. For spherical particles, number (D_n) and weight (D_w) average particle diameters were measured and calculated by eqn (1).¹⁷ For spherical particles, diameter was measured directly; for the oriented dispersed phase, the width, i.e. the shorter dimension, was measured, which is also denoted as D_n and D_w.

Results

Morphology

Permanganic etching has been proved to be effective in the characterization of semi-crystalline polyolefins, when combined with proper specimen preparation methods such as cryo-microtome and observation methods such as SEM and TEM.^6,16,18–21 It is found that the etchant is capable of discriminating different components in a multi-phase system as complicated as the in-reactor alloy. Fine structure at the lamellar level can be revealed as the etchant selectively removes defective regions.¹⁸

With the aid of the permanganic etching method, the morphology of the dispersed particles composed of EPR and HDPE can be clearly observed, as shown schematically in Fig. 2. In Fig. 2, the PE part is represented by light grey and the EPR dark grey. Basically, all the particles observed in this study own an overall core–shell like structure, wherein EPR acts as the shell and PE the core. Based on the shape and amount of the PE, five types of particles were observed, which were denoted as type I–V.

Fig. 2 Illustration of the composite particle composed of EPR and PE. The darker part represents EPR and the lighter part crystalline PE.

The characteristic of type I particle is that this type of particles contain one PE inclusion in the EPR domain and EPR is the major part. In type II particles, there are several PE inclusions in one EPR domain and EPR is still the major part. In type III particles, the core consisting of PE is large and acts as the majority part, and the EPR layer encapsulating the PE is very thin. Type IV particles contain several PE inclusions and the inclusions are the majority part. Type V particles also have an overall core–shell structure and the large PE core is encapsulated by the EPR shell. The shape of the PE core is somewhat irregular. Some EPR particles are included in the PE domain and some EPR penetrate into the PE domain through the EPR/PE interphase, as shown in Fig. 2.

Morphology of IPC

The morphology of the IPC sample used in this study is typical of an in-reactor alloy, as shown in Fig. 3(a) and (b). Fig. 3(a) and (b) show the morphology of the YZ and XZ planes (see Fig. 1) of the injection-molded sample. In the SEM images, the dispersed particles can be clearly seen. In many particles, PE-type inclusions with lighter color can be identified. As shown in Fig. 3(a) and (b), when observed along the X (YZ plane) and Y (XZ plane) directions, the EPR rubber particles dispersed in the PP matrix both appear to be nearly round and no apparent difference was observed. This indicates that there is no anisotropic structure in the observed section of the IPC injection-molded bar and the dispersed EPR phase owns a nearly spherical shape. Inside some EPR particles, there are one or more “inclusions”, which are assigned to PE-type crystalline fractions^3,4 and can be represented by type I and II particles, as shown in Fig. 2, respectively.

Fig. 3 SEM micrographs of IPC. (a) YZ plane and (b) XZ plane, as shown in Fig. 1, Z is horizontal.

The SEM images of the XZ plane for the blends with different types and amounts of PE are shown in Fig. 4 and 5 and images with larger magnification for A5 and B1 are also shown in Fig. 6(a) and (b), respectively. The SEM images of the YZ plane for A3 and B3 are shown in Fig. 7(a) and (b), respectively. In the SEM images, components can be identified by color contrast combined with the interface between the different phases. PE appears lighter, and PP and EPR darker. Furthermore, the lamellae of the crystalline PE can be clearly revealed, as shown in Fig. 6(a).

Fig. 4 SEM micrographs of the XZ plane of the A1–A5 sample. (a)–(e) Correspond to A1–A5 samples, respectively. Z is horizontal.

Fig. 5 SEM micrographs of the XZ plane of the B1–B5 sample. (a)–(e) Correspond to B1–B5 samples, respectively. Z is horizontal.

Fig. 6 SEM micrographs of the XZ plane of A5 (a) and B1 (b) samples, respectively. Z is horizontal.

Fig. 7 SEM micrographs of the YZ plane of A3 (a) and B3 (b), respectively. Z is horizontal.

The SEM micrographs of the XZ plane for blends A1–A5 are shown in Fig. 4(a)–(e) and those of the YZ plane are shown in Fig. 7(a) for A3 and S1(a) and (b) (see ESI†) for A2 and A3. It is found that the dispersed particles are all nearly round in shape for both XZ and YZ planes. This situation is the same as that in IPC, which indicates that the dispersed phases in A2 and A3 are spherical particles and it also can be concluded from Fig. 4(a) and (d) that the dispersed phases in A1 and A4 are also spherical. The orientation in A5 can be clearly shown by XZ plane morphology, as shown in Fig. 4(e).

As shown in Fig. 4 and SEM micrographs with higher magnification (Fig. 7(a)), the structure of the dispersed phase for A1–A5 can be identified. The core–shell structure can be found in all five samples but the detailed structure of the composite particles is greatly influenced by w_PE. For A1 sample with low w_PE (5 wt%), as shown in Fig. 4(a), type I and II particles are found, which is similar to IPC. Due to the addition of PE, more PE is included in EPR particles and thus type III morphology is also found in A1. When w_PE increases to 11.5%, as shown in Fig. 4(b), an obvious change in morphology occurs. EPR particles, type I and type II particles are rarely seen. There are mainly type III and type IV particles. When w_PE increases to 20.6% and 30%, as shown in Fig. 4(c) and (d), type III, IV and V particles dominate. The type V particles may evolve from type IV during the coalescence of the PE particles in the EPR domain. When w_PE increases, the separated PE particles existing in the same dispersed composite particles begin to coalesce and the EPR phase should be pushed out. However, some EPR particles are “frozen” in the PE domain due to a kinetic reason, for which EPR has not enough time to be pushed out of the PE domain before solidification of the system. For A5, in which w_PE is as high as 40 wt%, the PE content is nearly the same as the PP content. However, the PE/EPR composite particles still exist and act as dispersed phases, which are also core–shell type. Round particles and particles oriented along FD can both be found in A5, as shown in Fig. 4(e), which is different from A1–A4 to some extent. An aspect ratio as high as 7 is found in the oriented particles. Accordingly, orientation of the PE crystallites takes place in such type of particles, as shown in Fig. 6(a) at higher magnification, wherein the existence and orientation of the lamellae of crystalline PE can be clearly seen. The normal layer of lamellae has a preferred orientation along the FD direction.

As shown in Table 2, D_n of IPC and A1–A4 increases from 1.17 to 1.75 µm with the addition of PE. This indicates that when w_PE increases, more PE is encapsulated in the dispersed phase, which leads to an increase in the volume fraction of the dispersed phase and complexity of the particles. Consequently, the particle size increases and the distance between the particles decreases. However, due to the compatible effect of EPR in PP/EPR/HDPE blends,^3,4 the size and shape of the particles do not change too much and the aggregation of the particles is also limited. When w_PE increases further to 40 wt% (A5), a relatively greater change in the morphology takes place. Some particles are no longer spherical and a preferred orientation along FD is found; however, the PE/EPR composite particles still act as a dispersed phase.

For blends B1–B5, which contain low molecular weight PE, the situation becomes quite different, as shown in Fig. 5(a)–(e). In the SEM micrographs of the XZ plane of B1 and B2, as shown in Fig. 5(a) and (b), the dispersed phase shows a round shape. For B3–B5, a highly oriented morphology can be identified. Combined with the SEM micrographs of the YZ plane for B2 and B3, as shown in Fig. S1(c) (see in ESI†), 7(b) and S1(d) (see in ESI†), it is clear that the dispersed phase in B2 is mainly spherical particles as IPC and A1–A4 and the dispersed phase in B3 is strip-like. The morphology of B1 can be concluded to be similar to that of B2, and B4 and B5 to that of B3, i.e., the morphology of B1 and B2 is essentially the same as that of A1, in which spherical particles with the EPR shell and PE core can be clearly identified, as shown in Fig. 6(b). The orientation of the dispersed phase along the FD direction becomes very obvious when w_PE equals and/or exceeds 20.6%, as shown in Fig. 5(c)–(e). For B3 and B4, “long strips” composed of PE and EPR were observed, which had a width as low as 0.2–1.6 µm and a length as high as tens of microns. When w_PE is as high as 40% (B5), the orientation of the PP domain and the PE/EPR domain is very obvious.

Then, the dimension of the PE/EPR phase in B1–B5 is to be considered. The data are listed in Table 2. For B1 and B2, in which spherical particles exist, different particle sizes are found. For B1, D_n is 1.61 µm, which is larger than that of IPC due to the increase of the overall volume fraction of the dispersed phase. However, D_n becomes 0.91 µm for B2 at w_PE = 11.5 wt%, which is smaller than that of IPC, A1–A5 and B1. This may be due to the decrease of the viscosity of the dispersed phase, which will be discussed in the following part of this study. The width of the strips, which can represent the diameter of the cross-section of the strips (also denoted as D_n here), in B3–B5 is measured and is listed in Table 2. Due to the elongation effect, D_n values for B3–B5 are apparently smaller than those in the counter parts with the same PE content in A3–A5. The D_n value is larger and the spacing between the strips seems lower in B5 than those in B3 and B4 because of the higher w_PE in B5, in which w_PE is similar to PP content.

The morphology of the YZ plane of A3 and B3 is compared, as shown in Fig. 7(a) and (b), respectively. There are 91 particles in Fig. 7(a) and 166 particles in Fig. 7(b) and D_n is 1.16 and 0.65 µm, respectively. Assuming the particles seen in the SEM micrographs are all perfectly round, the area occupied by dispersed phase for A3 and B3 is 96.1 and 55.1 µm², respectively. Considering the volume fraction of the dispersed phase is the same for A3 and B3, it can be concluded that more material is distributed in the X direction because of the long strip morphology in B3. However, the count of the particles is higher in B3, so the particles in B3 appear to be denser. This further confirms that when the dispersed phase particles are spherical, the apparent size of the particles in the YZ plane is apparently larger than that in samples with an oriented dispersed phase and the number of particles is lower. Such type of morphology difference appears to have a great influence on the toughness and elongation behaviour of the materials, which will be discussed in the following part in this paper.

From the results shown above, it is found that the morphology of the blends is greatly affected by the type and molecular weight of the PE added. The morphology in turn affects the mechanical properties, which will be discussed in the following section.

It is also important to note that the study of the samples prepared by methods that may induce orientation, such as injection-molding and extrusion, should be conducted with care. Taking the results in this paper for example, misleading may occur when only the morphology of the YZ plane is to be considered, which is the common situation because the cryo-fracture of the injection-molded bar is easy to realize. For B3–B5 samples in this study, the morphology of the YZ plane, as shown in Fig. S1(d) (ESI†) and 7(b), is not enough to reveal the real morphology. Thus, observation of the YZ plane is necessary and the existence of an oriented morphology can be more effectively observed in the YZ plane view.

Mechanical performance

Notched Izod impact strength (NIS) of IPC and blends at room temperature and −20 °C is plotted against PE content in Fig. 8. For IPC and blends A1–A5, at room temperature, NIS increases with increasing PE content at 0–20.6 wt% and levels off with further increase of w_PE up to 40 wt%. The situation is similar for NIS at −20 °C.

Fig. 8 Notched Izod impact strength of IPC and the IPC/PE blends. (a) Room temperature, (b) −20 °C. (Error bar is added).

For blends B1–B5, at room temperature, NIS of B1 slightly increases compared with that of IPC and afterwards, NIS of B2–B5 decreases upon the increase of PE content. When w_PE is greater than 20.6%, NIS of the blend is lower than that of IPC. The trend of NIS at −20 °C is similar to that at room temperature. NIS of B1 is greater than that of IPC and then NIS goes downwards. When w_PE is greater than 20.6 wt%, NIS is lower than that of IPC, but NIS of B5 increases slightly with respect to B3 and B4 and is still lower than that of the IPC.

The elongation at breaks (ε_b) was also found to have a tight relationship with morphology. As shown in Fig. 9, elongation at the break of A1 (245%) is nearly four times of that of IPC (63%) and ε_b decreases with further increase of w_PE. ε_b reaches a minimum of 25% at w_PE = 20.6% and then slightly increases to 31% (A4) and 53% (A5). ε_b of B1 and B2 is nearly the same as that of IPC and ε_b increases from 33 (B2) to 260% (B3) at w_PE of 20.6%. Further increase of ε_b takes place when w_PE increases from 20% to 40%. ε_b as high as about 460% was found at w_PE = 40%. It is obvious that the orientation morphology in B3–B5 plays an important role in increasing ε_b.

Fig. 9 Elongation at break of IPC and IPC/PE blends. (Error bar is added).

Discussion

Influence of PE content and molecular weight on morphology

It can be concluded from the observation above that the molecular weight and amount of PE both have a significant influence on the morphology of the injection-molded specimens. Decreasing the molecular weight of PE and increasing the PE content all contribute to the formation of the orientation structure. For B1–B5, orientation begins to occur at w_PE = 11.5% and becomes very obvious at w_PE ≥ 20.6%. For A1–A5 samples, which contain higher molecular weight PE, oriented morphology was observed only at w_PE = 40% and the orientation degree is relatively less pronounced compared with that in B3–B5.

Then, it is shown to be very important to discuss the factors influencing morphology. The phase structure of the PP/EPR/PE heterophasic system is determined mainly by three factors: viscosity ratio between the dispersed and the matrix phase (λ), compatibility between components and conditions of compounding and processing.³ In the system studied in this study, PP and EPR are the same (the same IPC) and the difference of PE is only molecular weight. The compounding and processing is also the same for all the samples. Therefore, the viscosity ratio seems to be the key factor. The importance of the viscosity ratio in determining the morphology of IPCs has been reported by numerous researchers.^3,22–26 The viscosity ratio (λ) and the capillary number (κ) are defined below:²²

λ = η_d/η_m (2)

κ = η_mẎ/(v/D) (3)

where η_d and η_m are the viscosity of dispersed and continuous phases, respectively, and Ẏ, v, and D are the shear rate, interfacial tension, and droplet diameter, respectively. The deformation of the dispersed phase is reported to be determined by λ and κ. Critical values are reported for different polymer blends such PP/EPR,^22,27 PP/PE/EPR²² and PE/PS²⁸ blends. λ being close to or less than unity seems to facilitate the formation of fibril morphology of the dispersion phase^22,27–30 (corresponding to the oriented morphology mentioned in this study).

Due to the difference in experiment conditions and samples selected, λ can be expressed by different forms in the literatures. It was reported to be obtained by measuring Brabender torque values,³¹ the ratio of zero shear viscosity (η₀)³² and the ratio of intrinsic viscosity.^25,26

In this study, we use IPC samples and thus it is difficult to measure the melt viscosity of the individual EPR and PP phases. However, the molecular weights of EPR (soluble fraction), PP and PE can be readily measured by GPC, so η₀ of the components at the injection-molded temperature T (here it is 200 °C) is calculated according to eqn (4)–(6) below:¹⁵

T_g is the glass transition temperature, E_η(∞) is the value of activation energy for viscous flow (E_η) for T ≫ T_g (glass transition temperature), which can be calculated by the additive function,¹⁵ and η_cr is the viscosity at critical molecular weight. M_w is the weight average molecular weight and M_cr is the critical molecular weight. n is determined by M_w of the samples. In the case in this study, n is 3.4 because M_w of PP, PE and EPR used here are all greater than the critical values (M_cr).¹⁵ T_g value of PP and PE can be found in the handbook¹⁵ and T_g of EPR is obtained by DMA (TA RSA-III), which is −52.5 °C (see Fig. S2 in the ESI†). The parameters for EPR such as M_cr and E_η(∞) are estimated by the values of PP and PE combining with the weight fraction of ethylene in EPR (C2 in EPR). The η₀(200 °C) values of the soluble fraction (mainly EPR) and insoluble fraction (mainly PP) thus calculated are listed in Table 1.

The molecular weight of the dispersion phase consists of PE and EPR, thus the melt viscosity of the PE/EPR blends, as a whole, is the key factor influencing the morphology. The viscosity of the blends obeys the logarithm mixing rule:^24,33,34

log η = ω₁ log η₁ + ω₂ log η₂ (7)

where η is the viscosity of the blend, η₁ and η₂ are the viscosities of the two components measured at the same temperature, respectively, and ω₁ and ω₂ are the volume fraction of the components. The η₀(200 °C) values of EPR/PE as a whole are listed in Table 2 and the viscosity ratio values between EPR/PE phase and PP phase are also given in Table 2 and plotted in Fig. 10. For IPC, the viscosity ratio is obtained by η₀(200 °C) value of EPR and PP.

Fig. 10 Viscosity ratio for IPC, IPC/PE-A and IPC/PE-B blends.

From Fig. 10, it is obvious that the difference in molecular weight and blending content of PE leads to a change in viscosity ratio between the dispersed phase consisting of PE and EPR (or EPR for IPC) and matrix (PP). In this study, the molecular weights of the two PE samples are both lower than that of EPR (soluble fraction of the IPC as shown in Table 1), thus the addition of PE lowers the viscosity of the PE/EPR composite phase. However, the degree of the decrease in the viscosity of the dispersed phase varied for different PE molecular weight and content. When PE is mixed with EPR by higher fraction or PE with lower molecular weight (PE-B) is used, the decrease of viscosity ratio is much more severe. According to the results reported by the literatures,^22,28,29 the lower the viscosity ratio (λ ≤ 1), the easier the oriented fibre-type dispersed phase can occur under shear flow. For B1–B5 blends, according to SEM results, the orientation of the dispersed phase occurs at w_PE = 20.6%, where the viscosity ratio is 3.22. At PE content = 40%, viscosity ratio is as low as 0.90. For A1–A5 blends, the viscosity ratio ranges from 8.25 to 31.68, which are much higher than those of B1–B5. Thus, it is much difficult for the A1–A5 blends to form the oriented dispersed phase.

From the analysis above, it can be concluded that the formation of the oriented structure can be attributed to the decrease of viscosity of the dispersed phase due to the increase of PE content and/or the lower PE viscosity. It should be noticed that orientation can be observed in A5, but it is not as obvious as that in B3–B5. The viscosity ratio is 8.25 for A5, which is higher than that of B3. It may be inferred that formation of the oriented structure is easier when the PE content is high enough, in which case the weight contents of PP and PE are close to each other.

It should also be emphasised that during the processing in the twin-screw extruder and injection-molding, the polymer melt undergoes a complex flow, in which multi modes of deformation may exist such as shearing and extension. Therefore, the viscosity in the real processing condition is complex but it is beyond the scope of this paper.

Influence of morphology on mechanical properties

The addition of PE is believed to increase the overall volume of the elastomer particles and thus lead to an increase in NIS.^3,7,12 It has been reported that there exists a maximum at a w_PE of about 10 wt% in the NIS versus PE content curve of IPC/HDPE blends at room temperature.^7,11 Gahleitner et al.³ reported that NIS of a random-heterophasic copolymer/HDPE increases with increasing PE content at PE content of 0–15 wt% and changes little from 20 to 25 wt% at 23 °C, but at −20 °C, NIS goes down monotonically. In this study, for the two series of blends with different PE, the change of NIS with respect to PE content depends greatly on the type of PE added both at room temperature and −20 °C and distinct trends were found. It can be found that NIS has an obvious relationship with the morphology.

For blends A1–A4, when the PE content increases from 5 to 30 wt%, the dispersed phase maintains a nearly spherical shape, and even at 40 wt% (A5), the orientation of the dispersion particles is still not very obvious. This indicates that if the dispersion particles remain a nearly spherical shape, the addition of PE plays a role in increasing the “total elastomer volume”³ and thus leads to an increase in NIS at both room temperature and −20 °C. However, when PE content exceeds 20.6 wt%, the increase of NIS at room temperature and −20 °C is very little. This indicates that the detailed structure of the composite particles (such as type IV and V particles) and relatively low degree of orientation of the particles have no obvious influence on NIS. This also indicates that particle size (D_n) has little effect on NIS for A3 and A4, where D_n changes from 1.39 to 1.75 µm.

For B1–B5, the maximum NIS is found at the PE content of 5% at both temperatures. For blends B1 and B2, the particles are predominantly spherical and thus their NIS is larger than that of IPC at both temperatures. It is at the PE content of 20.6% that NIS begins to become lower than that of IPC at both temperatures and it should be emphasized that the apparent morphology change also takes place at this PE content. This means that it is the oriented strip-like morphology that leads to a decrease in NIS. That is to say, when oriented morphology occurs, further addition of PE will cause a decrease in NIS.

Compared to the NIS of B series (NIS_B), it is found that the NIS of A series samples (NIS_A) are much tougher at PE content greater than 5 wt% at both room temperature and −20 °C. This indicates that the spherical particles are more effective in toughening the material. At PE content of 5 wt%, NIS_A is nearly the same with NIS_B at room temperature and NIS_B is higher at −20 °C because at PE content of 5 wt%, the dispersed phase particles are all spherical for both A and B samples.

The negative effect of strip-like morphology can also be confirmed by the experiment result of crack surface observation for samples after the NIS test. The morphology of the cracking surface (YZ plane) of A5 and B5 after the impact test at −20 °C was observed and the SEM micrographs are shown in Fig. 11(a) for A5 and (b) for B5. The difference in the morphology is obvious; i.e., the surface of A5 is much coarser than of B5. For A5, a lot of granules and holes with a diameter of about 0.5–1.6 µm can be found, but for B5, the amounts of granules and holes are much less than those in A5 and their size is also much smaller (about 0.4 µm). Obviously, A5 with coarser surface owns higher toughness.

Fig. 11 SEM micrographs of the cracking surfaces (YZ plane in Fig. 1) of the A5 (a) and B5 (b) samples after NIS test at −20 °C.

Although the samples with oriented structure are inferior in toughness, the elongation at the breaks is superior for this type of samples. From Fig. 9, it is obvious that for B1–B5, ε_b begins to increase rapidly at w_PE = 11.5 wt%, the w_PE point at which the strip-like morphology occurs and develops with further addition of PE. For A1–A5, in which the spherical particles or slightly oriented particles exist, ε_b increases for A1 compared with that of IPC, and decreases from w_PE 5% to 20.6% and remains nearly unchanged afterwards. ε_b of B3–B5, in which strip-like morphology is observed, are greatly larger than those of A3–A5. Herein, the role of the oriented morphology in the enhancement of ε_b is very pronounced. For A1–A2 and B1–B2, in which the particles are basically all spherical, ε_b of A1–A2 are greater than those of B1–B2. This further proves the important role of the orientation structure in ε_b.

Why does the oriented morphology influence mechanical properties so significantly? Undoubtedly, the anisotropic morphology of the injection-molded samples plays an important role. When the samples are subjected to impact or tensile tests, the force is imposed in the specific direction(s). Thus, it is reasonable to conclude that the morphology of some specific plane(s) corresponding to the specific force is crucial to the mechanical properties. For convenience, this morphology of this plane is named “effective morphology”.

When the dispersed phase particles are spherical, the morphology of the XZ plane and the YZ plane are basically the same, such as IPC shown in Fig. 3(a) and (b) and A3 shown in Fig. 4(c) and 7(a). In this case, the morphology of both the XZ and YZ planes can be taken as “effective morphology”. This is the generally seen morphology for impact copolymers, for which the fraction of the dispersed phase is the dominant factor in determining NIS.

When orientation morphology exists, the morphologies of the XZ and YZ planes are distinct and it is reasonable to conclude that one of them is responsible for the decreased NIS and increased ε_b compared with those without oriented morphology. Taking the comparison of A3 versus B3 and A5 versus B5 as examples, NIS of A3 and A5 is much larger than those of B3 and B5 at both room temperature and −20 °C, respectively, but ε_b of A3 and A5 is much smaller than that of B3. As discussed above, considering the morphologies of the YZ plane of A3 versus B3 and A5 versus B5, as shown in Fig. 7(a), (b) and 11(a), (b), respectively, compared with A3 and A5, the amount of particles is much larger and the particle size is much smaller in B3 and B5. According to toughening theory, in toughened crystalline plastics, the role of the dispersed rubber phase is to initiate the craze and shearing yielding^35,36 and it is reported that smaller particle size leads to higher toughness in impact polypropylene copolymers for both PP/EPR^37–39 and PP/EPR/HDPE² blends. From this point of view, it seems that NIS of B3 should be larger than that of A3 and B5 larger than A5 but the situation is reversed here. From the experiment results in this study, it is shown that the morphology of the XZ plane seems to be the “effective morphology”, which is the oriented strip-like morphology.

The influence of the effective morphology on NIS is illustrated in Fig. 12. For samples in which spherical or nearly spherical particles dominate (A1–A5 and B1 and B2), the fracture occurs in the YZ plane and the detachment of the particles from the matrix occurs in the NIS test, which induces the formation of granules and holes in the fractured surface, as shown in Fig. 12(a) (the A type morphology) and the SEM micrograph Fig. 11(a). For samples that own effective morphology, as shown in Fig. 12(b) (B3–B5), when fracture happens, most strips are broken in the cross-section and the B type morphology is observed (Fig. 12(b)), but A type morphology is fewer, which results in a relatively smoother surface. The coarser fractured surface dissipates more impact energy, and thus samples with the spherical dispersed phase own higher NIS.

Fig. 12 Schematic of the effective morphology (morphology of the XZ plane) of the injection-molded bar.

The occurrence of the oriented strip-like morphology gives rise to ε_b. The influence of orientation morphology on ε_b seems more straightforward. In this study, the tensile force is along X (FD), which is in the same direction as the long axis of the strips composed of PE and EPR. When the samples are subjected to tensile deformation, this type of structure may not so easily initiate the fracture as when the particles are spherical, and the existence of the oriented strip may also facilitate the mobility of the crystallites of iPP. The detailed mechanism is worthy of further study.

It is also meaningful to point out that an interesting characteristic of the oriented structure is that it is an anisotropic system. It is anticipated that, if the direction of impact or tension changes, distinct behaviour might be observed, which is worthy of further study.

Conclusions

From the experiment results and discussion above, it is found that the molecular weight and the content of PE both have great influence on the morphology and properties of the injection-molded products of IPC/PE blends. The morphology can be tuned by adjusting molecular weight and PE content.

An important observation of this study is that the impact strength of the IPC/PE blends depends not only on PE content, but also on morphology. “Effective morphology” is suggested to govern the properties. The dispersed particles are effective in toughening only when the particles are spherical or nearly spherical. The orientation of the dispersed phase has a great negative effect on the impact strength. When orientation exists, upon the increase of PE content, the impact strength decreases monotonically up to the limits in this study.

Although having negative influence on impact strength, the existence of the orientation structure of the dispersed phase greatly enhances the elongation at break values of the samples. There is an increase of above 600% of the elongation at break value when PE content is as high as 40 wt%.

Acknowledgements

The authors gratefully thank the SINOPEC Beijing Research Institute of Chemical Industry for the support of this study. The authors also thank Xiaoguang Cai, Wenjun Wei, Honghong Huang, Dong Wei, Dan Wang, and Yiqing Bai for the valuable discussions and the technical help.

Notes and references

1. C. Grein, K. Bernreitner, A. Hauer, M. Gahleitner and W. Neissl, J. Appl. Polym. Sci., 2003, 87, 1702–1712.
2. F. C. Stehling, T. Huff, C. S. Speed and G. Wissler, J. Appl. Polym. Sci., 1981, 26, 2693–2711.
3. M. Gahleitner, A. Hauer, K. Bernreitner and E. Ingolic, Int. Polym. Process., 2002, 318–324.
4. M. Gahleitner, C. Tranninger and P. Doshev, J. Appl. Polym. Sci., 2013, 130, 3028–3037.
5. P. Galli, T. Simonazzi and D. D. Duca, Acta Polym., 1988, 39, 81–90.
6. K. P. Tchomakov, B. D. Favis, M. A. Huneault, M. F. Champagne and F. Tofan, Can. J. Chem. Eng., 2005, 83, 300–309.
7. H. J. Jang, S.-D. Kim, W. Choi and Y. S. Chun, Macromol. Symp., 2012, 312, 34–42.
8. Y. Lin, V. Yakovleva, H. Chen, A. Hiltner and E. Baer, J. Appl. Polym. Sci., 2009, 113, 1945–1952.
9. Z. S. Petrovic, J. Budinski-Simendi, V. Divjakovi and Z. E. Scaronkrbi, J. Appl. Polym. Sci., 1996, 59, 301–310.
10. R. E. Robertson and D. R. Paul, J. Appl. Polym. Sci., 1973, 17, 2579–2595.
11. M. Mirabella Francis, in Applications of Scanned Probe Microscopy to Polymers, American Chemical Society, 2005, vol. 897, ch. 16, pp. 224–238.
12. B. Qiu, F. Chen, Y. Shangguan, L. Zhang, Y. Lin and Q. Zheng, RSC Adv., 2014, 4, 58999–59008.
13. G.-J. Zhong and Z.-M. Li, Polym. Eng. Sci., 2005, 45, 1655–1665.
14. R. R. Tiwari and D. R. Paul, Polymer, 2011, 52, 4955–4969.
15. D. W. van Krevelen and K. Te Nijenhuis, in Properties of Polymers, Elsevier, Amsterdam, 4th edn, 2009, pp. 525–597.
16. R. H. Olley and D. C. Bassett, Polymer, 1982, 23, 1707–1710.
17. P. Doshev, G. Lohse, S. Henning, M. Krumova, A. Heuvelsland, G. Michler and H.-J. Radusch, J. Appl. Polym. Sci., 2006, 101, 2825–2837.
18. R. H. Olley, A. M. Hodge and D. C. Bassett, J. Polym. Sci., Polym. Phys. Ed., 1979, 17, 627–643.
19. D. R. Norton and A. Keller, Polymer, 1985, 26, 704–716.
20. J. X. Li, W. L. Cheung and C. M. Chan, Polymer, 1999, 40, 2089–2102.
21. J. Park, K. Eom, O. Kwon and S. Woo, Microsc. Microanal., 2001, 7, 276–286.
22. B. K. Kim and I. H. Do, J. Appl. Polym. Sci., 1996, 60, 2207–2218.
23. J. Karger-Kocsis, A. Kalló and V. N. Kuleznev, Polymer, 1984, 25, 279–286.
24. M. Kontopoulou, W. Wang, T. G. Gopakumar and C. Cheung, Polymer, 2003, 44, 7495–7504.
25. H. Tan, L. Li, Z. Chen, Y. Song and Q. Zheng, Polymer, 2005, 46, 3522–3527.
26. G. Grestenberger, G. Potter and C. Grein, eXPRESS Polym. Lett., 2014, 8, 282–292.
27. M. Ono, K. Nakajima and T. Nishi, J. Appl. Polym. Sci., 2008, 107, 2930–2943.
28. K. Min, J. L. White and J. F. Fellers, Polym. Eng. Sci., 1984, 24, 1327–1336.
29. M. V. Tsebrenko, N. M. Rezanova and G. V. Vinogradov, Polym. Eng. Sci., 1980, 20, 1023–1028.
30. B. K. Kim and I. H. Do, J. Appl. Polym. Sci., 1996, 61, 439–447.
31. J. Karger-Kocsis and V. N. Kuleznev, Polymer, 1982, 23, 699–705.
32. L. D'Orazio, C. Mancarella, E. Martuscelli and G. Sticotti, J. Mater. Sci., 1991, 26, 4033–4047.
33. L. D'Orazio, C. Mancarella, E. Martuscelli and F. Polato, Polymer, 1991, 32, 1186–1194.
34. A. L. N. D. Silva, M. C. G. Rocha, F. M. B. Coutinho, R. Bretas and C. Scuracchio, J. Appl. Polym. Sci., 2000, 75, 692–704.
35. S. H. Wu, Polymer, 1985, 26, 1855.
36. C. B. Bucknall, J. Elastomers Plast., 1982, 14, 204–221.
37. B. Z. Jang, D. R. Uhlmann and J. B. Vander Sande, J. Appl. Polym. Sci., 1985, 30, 2485–2504.
38. B. Z. Jang, D. R. Uhmann and J. B. Vander Sande, Polym. Eng. Sci., 1985, 25, 643.
39. A. V. D. Wal, A. J. J. Verheul and R. J. Gaymans, Polymer, 1999, 40, 6057–6065.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra08517a

---