AFM, atomic force microscope; ARES, advanced rheometric expansion system; DR, draw ratio; DSC, differential scanning calorimetry; HDPE, high-density polyethylene; L-PP, linear polypropylene; MD, machine direction; MDO, machine direction orientation; MFI, melt flow index; PE, polyethylene; PET, polyethylene terephthalate; POM, polarised optical microscopy; PP, polypropylene; PVC, polyvinylidene chloride; TD, transverse direction.
Cast films of two linear polypropylenes (PP) having different molecular weights and their blends with 20 wt% random copolymer were prepared by extrusion. The produced cast films were then uniaxially hot drawn at T = 110 and 140°C under a draw ratio (DR) of 6 using a machine direction orientation (MDO) unit. The effects of drawing temperature and addition of random copolymer on the properties of the films were investigated. The type of crystals (spherulites or fibrils) was studied using differential scanning calorimetry (DSC). The drawing process generated a highly oriented fibrillar crystalline structure, resulting in an increase in the melting point of the films by about 10°C. The films drawn at 140°C revealed coexisting lamellae and fibrils whereas at 110°C mainly a fibrillar structure was observed. Tensile measurements showed a greater strength for the sample stretched at lower temperature (i.e., 110°C). The effect of drawing temperature and copolymer content on tear resistance was also explored. The medium-molecular weight PP film stretched at 110°C showed greater tear resistance than at 140°C. The addition of the random copolymer into the medium-molecular weight PP reduced the tear resistance significantly whereas the impact for the high-molecular weight was not noticeable. Adding the random copolymer significantly decreased the haze for the films and hence drastically improved the clarity.
Des feuilles coulées de deux polypropylènes (PP) linéaires avec des poids moléculaires différents et leurs mélanges avec 20% d'équivalent en poids de copolymère aléatoire ont été préparées par extrusion. Les feuilles coulées produites ont alors été étirées à chaud de façon uniaxiale à T = 110 et 140°C en vertu d'un rapport d'étirage de 6 à l'aide d'une unité d'orientation de sens machine. On a étudié les effets de la température d'étirage et l'ajout du copolymère aléatoire sur les propriétés des feuilles. Le type de cristaux (sphérulites ou fibrilles) a été étudié à l'aide de la calorimétrie à balayage différentiel (CBD). Le processus d'étirage a produit une structure cristalline fibrillaire fortement orientée, menant à une augmentation du point de fusion des feuilles d'environ 10°C. Les feuilles étirées à 140°C ont révélé des lamelles et fibrilles coexistantes, tandis qu'à 110°C, on a observé principalement une structure fibrillaire. Les mesures de la traction ont indiqué une résistance supérieure pour l'échantillon étiré à une température basse (c.-à-d., 110°C). On a également exploré l'effet de la température d'étirage et du contenu en copolymère sur la résistance au déchirement. La feuille de PP de poids moléculaire moyen étirée à 110°C a démontré une résistance au déchirement supérieur qu'à 140°C. L'ajout du copolymère aléatoire dans le PP de poids moléculaire moyen a réduit la résistance au déchirement de façon considérable, tandis que l'effet pour le poids moléculaire élevé n'était pas visible. L'ajout du copolymère aléatoire a considérablement réduit le trouble des feuilles et, par conséquent, a amélioré radicalement la clarté. Can. J. Chem. Eng. © 2010 Canadian Society for Chemical Engineering
Polypropylene (PP) is one of the most widely used polymers for the production of plastic films. Cast extrusion followed by stretching is one of the main processes to produce packaging films. In this process, the molten polymer flows through a slit die and rolls around a set of drums where it is cooled and formed into a relatively thick film with a dominant spherulitic crystalline structure. The produced films could be stretched uniaxially at a temperature near the melting point, which results in a highly oriented film. Packaging is the main application for uniaxial drawn films where high transparency and low-tear resistance along machine direction (MD) are needed. Some of the important Machine Direction Orientation (MDO) applications are: polyethylene terephthalate (PET) strapping, polyvinylchloride (PVC) food wraps, fibrous high-density polyethylene (HDPE) ribbons for weaving sacks, breathable hygienic films in diaper liners, self-adhesive labels, and polyolefin packaging and lamination (Schut, 2005). The MDO process has attracted attention of producers and researchers since it is less expensive compared to biaxial stretching and the structural evolution can be evaluated much more easily in the former (Schut, 2005; Sadeghi and Carreau, 2008a).
The MDO stretching unit can be operated either in-line or off-line with extrusion and is controlled via variables such as: draw ratio (DR), drawing speed, drawing temperature, and heat-setting conditions. The drawing process usually improves the strength and clarity, but reduces the tear resistance along the MD and elongation at break along the transverse direction (TD; Sadeghi and Carreau, 2008a).
Mechanical and physical properties of films are strongly controlled by the morphology and microstructure formed during the process. Nie et al. (2000) have used a modified atomic force microscopy (AFM) technique to investigate the morphology development during biaxial stretching of PP films. Their results showed a transformation of the spherulites into an oriented fibrillar network structure. The main fibrils were formed in the drawing direction while thinner fibrils were connecting the main and thick elongated fibrils.
In Bafna et al. (2007), the uniaxial stretching of HDPE films created a fibrillar structure with highly oriented fibrils in MD. Koike and Cakmak (2004) studied the drawing of PP films with an initial cross-hatched structure (the cross-hatched structure is a common crystalline structure for PP blown films (see, e.g., Sadeghi et al., 2005)). At the early stages of deformation, the structure is first broken into small pieces and then microfibrils start to appear (Koike and Cakmak, 2004). An intermediate shish–kebab crystalline structure is observed at medium deformations while the transformation from a shish–kebab morphology into a fibrillar structure occurs beyond a strain of 1.2 (Koike and Cakmak, 2004).
Rettenberger et al. (2002) studied the effect of temperature on the tensile response of PP samples. Drawing at a temperature up to 155°C led to a typical ductile behaviour with a yield point, neck propagation, and strain hardening. At higher temperatures, instead of yielding, the deformation was quasi-rubber-like.
Tabatabaei et al. (2009a) studied the MDO drawing of PP films at 120°C. Two melting peaks in the differential scanning calorimetry (DSC) thermograms of the films prepared at DR > 4.8 were observed. They attributed the main peak to the shishs and the small one to the lamellae crystals. In addition, it was reported that the films prepared at high DR contained mainly a fibrillar structure. Their findings also showed that the tear strength along MD was decreased about four times as a result of chain orientation. The films stretched at high DR exhibited a lower amorphous tie chain mobility and hence showed a smaller oxygen permeability compared to the samples prepared at low DR (Tabatabaei et al., 2009a).
We have recently considered the morphology development during MDO stretching of PP (Sadeghi and Carreau, 2008a). The stretching was carried out at a high temperature (140°C) where the crystalline lamellae could be reformed and reoriented. It was shown that a mainly fibrillar crystalline structure could be obtained at high DR such as DR = 6. A very distinctive improvement in some mechanical properties was also observed as the DR increased.
Although a few authors have investigated uniaxial stretching of various resins, no study has considered the effects of the drawing temperature and addition of copolymer on the properties of MDO drawn films. It is well known that the thermal properties (e.g., melting peak and broadness of melting curve) are quite sensitive to the crystalline morphology as well as the size of crystals. In this study, we will use DSC tests to investigate in details the role of the stretching temperature and the addition of a random copolymer on the type of crystals (spherulites and fibrils). We will also study the influence of such parameters on the mechanical responses, tear resistance and particularly haze of the stretched films.
Three commercial PP were used in this study. PP5341E1 (PP5341) with a melt flow index (MFI) of 0.83 g/10 min (under ASTM conditions of 230°C and 2.16 kg) is a high-melt strength grade and PP4712E1 (PP4712) with a MFI of 2.8 g/10 min is a resin proposed for the production of oriented films. Both resins are homopolymers supplied by ExxonMobil. RP210G is a random copolymer from Basell with a MFI of 1.8 g/10 min.
Dynamic rheological measurements were carried out using an ARES rheometer with a parallel plate geometry of 25 mm diameter and a gap equal to 1.5 mm at the temperature of 200°C under nitrogen atmosphere. Moulded discs of 2 mm thick and 25 mm in diameter were prepared using a hydraulic press at 200°C. Prior to frequency sweep tests, time sweep tests at a frequency of 0.314 rad/s and 200°C were performed for 2 h to check the thermal stability of the specimens. No significant degradation (changes <5%) was observed for the duration of the frequency sweep measurements.
The cast films were prepared using an industrial multilayer cast film unit from Davis Standard Company (Pawcatuck, CT) equipped with a 2.8 mm opening and 122 cm width slit die and two cooling drums. The extrusion was carried out at 220°C and the distance between the die exit to the nip roll was 15 cm. In MDO, the produced films were uniaxially stretched at 110 and 140°C using a DR of 6.
For FTIR measurements, infrared spectra were recorded on a Nicolet Magna 860 FTIR instrument from Thermo Electron Corp. (Waltham, MA) using a DTGS detector with a resolution of 4 cm−1 and an accumulation of 128 scans. The beam was polarised by means of a Spectra-Tech zinc selenide wire grid polarizer from Thermo Electron Corp. The crystalline and amorphous orientations were measured based on the method explained in details by Sadeghi et al. (2007).
Thermal properties of specimens were analysed using a TA instrument differential scanning calorimeter (DSC) Q1000 (TA Instruments, New-Castle, DE). The thermal behaviour of films was obtained by heating from 50 to 220°C at a heating rate of 10°C/min. The reported crystallinity results were calculated using a heat of fusion of 209 J/g for a fully crystalline PP (Elias et al., 2000). Crystallisation monitoring was performed using a polarised optical microscopy (OPTIHOT-2) to follow spherulites growth. The films were first heated on a programmable hot stage (Mettler FP82HT) from room temperature to 200°C at a heating rate of 10°C/min and were kept at that temperature for 1 min to eliminate initial thermomechanical history, and then cooled to room temperature at the same rate.
Tensile tests were performed using an Instron (Instron Structural Testing Systems Corporation, Novi, MI) 5500R machine. The procedure used was based on the D638-02a ASTM standard. A standard test method for the tear resistance of plastic films based on ASTM D1922 was also used to obtain the MD tear resistance. According to this standard, the work required in tearing is measured by the loss of energy of an encoder, which records the angular position of the pendulum during the tearing operation.
Haze measurement was performed in accordance with the procedure specified in the ASTM D1003-97, using a Haze Guard Plus™ instrument (Model 4725) made by the BYK-Gardner, Inc. (Columbia, MD).
RESULTS AND DISCUSSION
The frequency sweep tests in the linear viscoelastic region for the PP resins are presented in Figure 1. It is obvious that PP5341 shows a larger complex viscosity compared to PP4712 and RP210G in the very low-frequency range, indicative of a larger zero-shear viscosity for the former.
Additionally, the shear-thinning behaviour starts at lower frequencies for PP5341 due to its higher molecular weight. These are in good agreement with the finding of Tabatabaei et al. (2009b) who showed that as the molecular weight of the L-PPs increases the rheological behaviour becomes more shear-thinning and the transition from the Newtonian plateau to the power-law region occurs at lower frequencies. In Figure 1, the shear viscosity of the blend containing 20 wt% random copolymer is shown to be between those of the neat components, but closer to that of the matrix for the high-molecular weight resin (i.e., PP5341). Tabatabaei et al. (2009b) studied the rheological properties of blends of a long-chain branched PP and different linear PP. Their results showed that the shear and extensional properties of the blends were dominated by the high-molecular weight L-PP.
The orientation of the amorphous and crystalline phases (c-axis with respect to MD) of the samples has been measured by Fourier transform infrared spectroscopy (FTIR), and the results are presented in Table 1. The orientation function was calculated as proposed by Lamberti and Brucato (2003):
where D is the ratio of the absorbance in the machine (parallel) direction to that in the TD. For PP, absorption at the wavenumber of 998 cm−1 is attributed to the crystalline phase (c-axis) and the one of 972 cm−1 is due to the contribution of both crystalline and amorphous phases. From the former absorption, the orientation of the crystalline phase, Fc, can be determined while from the latter, the average orientation function, Favg, is obtained. The orientation of the amorphous phase, Fa, can be calculated according to:
where Xc is the degree of crystallinity.
|PP4712 (20 wt%)-110°C||0.89||0.50|
|PP4712 (20 wt%)-140°C||0.82||0.52|
|PP5341 (20 wt%)-110°C||0.88||0.50|
|PP5341 (20 wt%)-140°C||0.88||0.50|
Sadeghi et al. (2007) have shown that FTIR is a powerful and reliable technique that could be used to determine the orientation functions precisely. In general, the orientation of MDO films is relatively high compared to blown films (Sadeghi et al., 2005, 2007). From Table 1, it is clear that the orientation functions of both the crystalline and amorphous phases are quite high and are only slightly affected by the drawing temperature or the addition of the random copolymer. Therefore, the drawing temperature and adding a random copolymer do not have a significant effect on the crystalline alignment. It is speculated that beyond a certain DR, the molecular structure is highly oriented such that stretching temperature or incorporation of a random copolymer cannot impact its orientation significantly.
Figure 2 presents the thermograms for the drawn films of PP4712 at the drawing temperature of 110 and 140°C. The film stretched at 140°C shows two peaks whereas the one stretched at 110°C reveals a single peak. Stretching at a high temperature provides a sustainable relaxation of the shorter chains (disentangled from elongated state), leading to lamellae Crystallisation based on the fibrils instead of Crystallisation in fibrillar form (Sadeghi et al., 2007). However, the longer chains preserve the elongated form and so create fibrils. A double melting point for the specific shish–kebab structure of a PP has also been observed by Somani et al. (2005). According to them, the shish in PP had a melting temperature of about 5–10°C higher than that of the kebabs and about 15–20°C higher than that for spherulites. Elias et al. (2000) reported also a double melting peak for a PP drawn by 5.5 times. They attributed their results to a connectivity of the chains in the shish or fibrils that resulted in a large crystal thickness and, as a consequence, a higher melting point. The polymer chains in the films stretched at low temperature (i.e., 110°C) have less mobility than the ones stretched at 140°C. It is believed that at 110°C less short chains have the chance to disentangle and diffuse to create lamellae in comparison with stretching at 140°C. Therefore, most of the chains form a fibrillar structure and some thin lamellae are created based on these fibrils. This kind of structure formation is considered as re-Crystallisation and should not be mistaken with shear-induced Crystallisation. The later originates from the melt state and usually occurs when polymer chains are under stress in the flow direction (in molten state at die exit or in a mould in injection mouldings) that causes chain orientation accompanied by Crystallisation. However, in our case, initial structure is a crystalline structure and not a melt. In fact a transformation in crystalline structure takes place due to limited mobility of crystal block in the temperature range of 110–140°C.
As pointed out earlier (see Table 1), the orientation factor is slightly higher for the one stretched at 110°C but both films have high orientation compared to the lamellar crystalline structure of blown films (Sadeghi et al., 2005).
The thermograms for the PP5341 stretched films are shown in Figure 3. The double peak for the film drawn at 140°C is not as distinctive as the one observed for the PP4712 stretched film (see Figure 2). As mentioned, PP5341 is a high-molecular weight resin and a larger number of long chains are expected to present. Therefore, a larger number of entanglements help to preserve the elongated shape of the chains and upon Crystallisation a stronger fibrillar structure is formed. The effect is intensified when this high-molecular weight resin is stretched at a lower temperature of 110°C where chain relaxation is reduced. This allows most of the chains to form a fibrillar structure with the presence of some tiny lamellae crystals.
It should also be noted that the peak attributed to the fibrils is observed at a higher temperature for the films stretched at 140°C compared to the films drawn at 110°C, probably because the thicker fibrils are created at 140°C. The effect of the random copolymer addition is illustrated in Figure 4. As pointed out, the peak at the lower temperature is due to lamellae whereas the peak at the higher temperature is attributed to the fibril structure (see Figure 4a). The addition of 20 wt% random copolymer tends to suppress and flatten the peaks in the melting curves. According to Laihonen et al. (1997), the introduction of ethylene would disrupt the crystalline structure, resulting in a reduction in the Crystallisation kinetics and melting point. From Figure 4a and b, it is clear that the ethylene presence affects both peaks, indicating that adding the copolymer interferes with both fibril and lamellae Crystallisation.
To confirm that the double peak is mainly formed during the stretching process, the results for a second heating DSC ramp (i.e., the initial thermomechanical history has been erased) is shown in Figure 5. Obviously, in the second heating ramp, a single melting point at a much lower temperature (representing a spherulitic structure) is observed, confirming that the drawing process influences the crystal structure dramatically.
To see the effect of the addition of the random copolymer, the crystal morphology of PP4712 and RP120G and their blend was monitored during Crystallisation using polarised optical microcopy (POM), as exhibited in Figure 6. The figure shows the spherulites growth for the neat polymers as well as the PP4712 blend containing 20 wt% random copolymer cooled from the molten state at a rate of 10°C/min.
It is obvious that for PP4712 (homopolymer), the spherulites grow uniformly with well uniform circular shape. However, for the random copolymer (Figure 6b), a very clear disruption in the spherulites growth is observed as the sample is influenced by the ethylene interruption of the crystal growth and only few spherulites with a circular shape could be formed. The number of spherulites increased noticeably and their size decreased by the incorporation of ethylene into the homopolymer, indicating that the ethylene segments act as a nucleating agent. The Crystallisation of PP/PE blends studied by Li et al. (2001) elucidated that for a miscible blend the PP was unable to push the PE fractions into the interspherulitic region leading to a deformation in the spherulites similar to what observed in Figure 6. For the 20 wt% blend, the effect of ethylene interruption is less pronounced with only few non-circular shape spherulites created during the Crystallisation. In fact, the low amount of the ethylene slightly hinders the spherulites growth.
The tensile responses along MD were evaluated and are depicted in Figure 7. It is clear that the tensile strength for the samples stretched at 110°C is somewhat larger than those drawn at 140°C. As discussed earlier, the films stretched at 110°C contain more fibrils and hence show a greater strength in MD than the films stretched at 140°C. Our previous study (Sadeghi and Carreau, 2008a) showed that the formation of a structure with a higher fibril density improves the tensile strength. This effect is less pronounced for PP5341, probably because of an already large amount of fibrils. Figure 7b shows that the addition of 20 wt% random copolymer reduces the tensile strength and increases the elongation at break, confirming the finding of Nitta et al. (2005); and Bedia et al. (2000). No yielding in the mechanical responses along MD is seen, possibly due to a strong fibril structure of the stretched films (Sadeghi and Carreau, 2008a).
The drawing temperature has more impact on the tensile response of the samples in the TD, as illustrated in Figure 8. The samples stretched at 140°C show higher strength and elongation at break than the samples stretched at 110°C. This could be explained by the presence of more lamellae in the structure of the films produced at 140°C compared to the fibrillar structure of those prepared at 110°C. The lamellae connect the fibrils in a much stronger way in the former compared to the weak Van der Waals bonds present in the latter.
It should be mentioned that adding 20 wt% copolymer did not affect the stress–strain response (results not shown).
The tear resistance values in the MD are reported in Table 2. It is well known (Sadeghi and Carreau, 2008a) that the source of failure in tear for MDO films (highly oriented) comes from the delamination or debonding of the fibrils meaning that tear propagates mainly in the fibril structure. The tear process is very rapid and mainly concentrated in the fibril structure. Hence, for the interpretation of the tear data only the structure of the fibrils are quite important and needs to be analysed. The larger tear resistance observed for the sample drawn at the temperature of 110°C is attributed to the stronger fibrillar structure compared to samples drawn at 140°C. The addition of 20 wt% random copolymer decreases the tear resistance significantly for PP4712 stretched at 140°C. That is, likely because of a very weak crystalline network connection between fibrils that becomes much weaker with the introduction of ethylene groups, leading to an easier fibril delamination. The effect is much less pronounced for the high-molecular weight resin (i.e., PP5341) possibly due to a more dense and populated fibril structure, that is, hardly influenced by ethylene. The rheological measurements (Figure 1) suggest that the random copolymer could have a better miscibility with PP4712 than with PP5341. Also, for the PP5341-20 wt% blend, the properties are influenced mostly by the high-molecular weight component. In other words, the ethylene disruption will be more effective for PP4712 since the ethylene can be better incorporated into the structure.
|Sample||Tear (g/µm) × 103|
|PP4712 (20 wt%)-110°C||57|
|PP4712 (20 wt%)-140°C||7.0|
|PP5341 (20 wt%)-110°C||114|
|PP5341 (20 wt%)-140°C||91|
Haze is also an important parameter for packaging applications. The haze values are related to films transparency and clarity. A random copolymer can be blended with a homopolymer to reduce haze in packaging films (Kim and Stephans, 2004). Figure 9 illustrates the haze as a function of copolymer content for the two stretching temperatures. The film of PP4712 shows a much larger haze value at the drawing temperature of 140°C. That is, probably because of a greater portion of lamellae that exist for the films stretched at 140°C. However, for PP5341 the haze variations between the two temperatures are not noticeable. This again could be interpreted by the presence of the high-molecular weight chains that form a compact fibrillar structure with less lamellae. In other words, the large population of long chains creates such a strong fibril network that would not allow much chance to the lamellae to affect the haze values. The haze reduction for a stretched PP film upon the addition of a random copolymer is clearly illustrated in Figure 9, as it has been reported elsewhere (Laihonen et al., 1997). The haze reduction is related to the crystalline structure, particularly the crystal size. As it was extensively discussed for the results of Figures 4 and 6, the random copolymer affects the Crystallisation kinetics and reduces the crystal size.
In this study, we have investigated the effects of the drawing temperature and addition of a random copolymer on the properties of the MDO stretched PP films. The MDO process generated a highly oriented fibrillar crystalline structure as a result of deformation of the initial crystalline structure. Our findings can be summarised as follows: The drawing temperature (i.e., 110°C and 140°C) as well as the addition of 20 wt% random copolymer for the uniaxially stretched PP films in the MDO process at DR of 6 did not influence drastically the molecular orientation in MD. Drawing at a high temperature of 140°C resulted in the formation of a distinctive coexisting fibrillar and lamellar crystalline structures and the effect was more pronounced for the medium-molecular weight PP. Furthermore, stretching at a lower temperature of 110°C generated mainly a fibril structure. MDO films prepared at 140°C showed a better tensile strength along TD than films produced at 110°C. However, the effect of the drawing temperature was less significant along MD. The tear resistance was improved for samples stretched at 110°C compared to films prepared at 140°C. The addition of a random copolymer reduced the tear resistance for the medium-molecular weight PP (i.e., PP4712). However, the effect of the copolymer addition on the tear resistance of the high-molecular weight PP (i.e., PP5341) was negligible. A significant reduction in the haze was observed when 20 wt% random copolymer was blended with the linear PPs.
amorphous orientation function
average orientation function
enthalpy of fusion (J/g)
degree of crystallinity
complex viscosity (Pa.s)
Financial support from NSERC (Natural Science and Engineering Research Council of Canada) is gratefully acknowledged. We also acknowledge the large infrastructure grant received from the Canadian Foundation for Innovation (Governments of Canada and Province of Quebec), which allowed us to build the unique POLYNOV facility. We are also thankful to Messrs. P. Cigana, L. Parent, and P. M. Simard for their technical help. Finally, we are thankful to ExxonMobil and Basell for donating the resins used in this study.