Long chain branched impact copolymer of polypropylene: microstructure and rheology.
The remarkable growth in polypropylene (PP) over the recent years is spurred by its attractive property profile, which includes excellent chemical resistance, high stiffness, high melting point, low density and low cost . However PP has low melt strength, low elasticity, and a narrow processing window because of its linear macromolecular architecture and semicrystalline nature [2-4]. As a result, upon heating PP to its melting temperature, it undergoes sharp transition from a semi crystalline solid to a melt that has no appreciable rubbery plateau. Hence, PP cannot be easily used in melt processing operations such as deep draw thermoforming, upward film blowing and extrusion coating, which involve free surfaces undergoing extensional deformation. However, PP can be made amenable to these processing operations by improving its melt strength thereby opening new opportunities and markets. Therefore, in recent years, there has been a growing interest in developing high melt strength PP grades [5-10] suited for applications where conventional PP typically fails.
The melt strength of PP can be increased by at least four different ways: increasing the weight average molecular weight of PP, broadening the molecular weight distribution by incorporating high and low molecular weight chain fractions, blending PP with polymers such as low density polyethylene (LDPE), and introducing long chain branching (LCB) on the backbone of PP. The first option adversely affects processability; the second option is not easy to implement in conventional reactor technologies for PP; and the third option, although cost-effective, has the problem that LDPE and PP are immiscible and therefore adversely affect properties such as transparency. By far, the last option namely introducing LCB on PP chains is the most attractive option not only because long chain branching is known to enhance extensional resistance of an entangled melt but also because of the relative ease of modification via a reactive extrusion process.
Reactive extrusion of homopolymer polypropylene with peroxides in presence or absence of multifunctional acrylate monomers has been widely reported to produce LCB containing PP [11-19]. During this process two side reactions namely, [beta] scission of PP macroradicals and homopolymerisation of acrylate monomers take place [15, 20]. To control these side reactions, various coagents like tetraethyl thiuram disulfide (TETDS) , heteroatomic ring derivatives , dithiocarbamates [22-24] have been used in recent years. Most of these studies reported dynamic and transient rheological properties , molecular weight distribution, as well as thermal properties of the modified PP homopolymers. Pivokonsky et al.  reported reactive modification of metallocene homopolymer PP (mPP). They used molecular constitutive equations like the extended Pom-Pom (XPP) to quantify the level and type of long chain branching in the modified mPPs.
Relatively few studies have been reported on the reactive modifications of copolymers of PP. Augier et al.  have studied modification of 77/23 mol% semiblock copolymer of propylene/ethylene in the presence of a peroxide, maleic anhydride (MAH), and butyl 3-(2-furyl) propionate (BFA) as a coagent to control the [beta] scission. Specifically, MAH or BFA with different molar ratios were used as functionalizing reagents either separately or as mixture (MAH/BFA). They compared the number of grafted groups and molecular weights of all the functionalized samples and proposed a reaction mechanism relating the macromolecular chain modifications to the process conditions. Augier et al.  studied MAH/BFA grafted samples of PP block copolymer with the help of rheology, differential scanning calorimetry (DSC) and gel permeation chromatography (GPC) to further understand the relations between the macromolecular structure and reactive extrusion conditions. Zulli et al.  discussed quantitative correlations between molar mass distribution, feed conditions, functionalization degree, and LCB, and their effects on the linear viscoelastic response of PP block copolymers functionalized by butyl 3-(2-furyl) propionate (BFA) and/ or maleic anhydride (MAH).
With its glass transition temperature around 0[degrees]C, the PP homopolymer possesses only moderate low temperature impact properties. Significant improvements in impact properties can be achieved by copolymerizing propylene with ethylene to obtain an impact copolymer polypropylene (ICP). A typical ICP is a heterophasic polymer consisting ethylene propylene rubber (EPR) particles dispersed in an isotactic PP matrix phase. Additionally, small amounts of ethylene-propylene segmented copolymers of different sequences and segment lengths are typically present at the interphase of the EPR and PP phases [30-32]. Similar to homopolymer PP, the ICP also has low melt strength which limits its use in processes like thermoforming, film blowing, extrusion coating and foaming. On the other hand, ICPs are ideally suited for such applications as refrigerator liners and automotive parts, provided they can be thermoformed or foamed. Hence, there is a need to increase the melt strength of PP impact copolymers as well.
While extensive research has been carried out on long chain branched PP homopolymer as mentioned above, only a few reports are available on the introduction of long chain branching in impact copolymer prepared by reactive extrusion. None of these papers have reported the microstructure of the modified impact copolymer. Since the ICP is a heterophasic polymer, it is likely that the peroxide will not only modify the matrix PP phase but also diffuse inside the EPR particles and modify them. A schematic of the possible structure of ICP after reactive extrusion with peroxide is shown in Fig. 1.
A detailed study of the resulting microstructure of the modified ICP is important since it will govern the properties and hence the applicability of the ICP. For example, long chain branching of the matrix PP chains can improve the processability of the ICP whereas excessive crosslinking and microgel formation in the EPR phase can result in reduced impact strength of the ICP. Therefore the main objective of the present study was to prepare modified ICPs by reactive extrusion with peroxide and undertake detailed characterization of the modified ICP in order to understand the microstructure of its phases. To do this, each ICP sample was separated into its subcomponent phases and molecular as well as rheological characterization of the separated phases was carried out. Rheological data was fitted with molecular constitutive equations to obtain model parameters that were interpreted in combination with molecular characterization data to establish a molecular structure-rheology correlation.
A commercial impact copolymer (virgin ICP, MFI 1.5 g/10 min) containing 5 to 7% of ethylene in the total polymer ([E.sub.t]), 40 to 44% of ethylene in the EPR phase ([E.sub.c]) and 13 to 16% EPR content in the total copolymer ([F.sub.c]) was selected for the present study. Two modified ICP polymers, ICP-A (MFI 0.84 g/10 min) and ICP-B (MFI 0.44 g/10 min) were prepared from the commercial ICP by reactive extrusion. The lower MFI values of the modified ICP-A and ICP-B relative to the virgin ICP are indicative of their extent of modifications. Additionally, a commercial high melt strength PP impact copolymer (Dow Chemicals, USA, MFI 0.6 g/10 min) was also studied. This resin will be referred to as ICP-C. The manufacturing method of this resin is not known.
Virgin ICP, in the form of reactor fluff was compounded with 5000 ppm of hexadecoxycarbonyloxy hexadecyl carbonate (Perkadox 24L, Akzo Nobel, Netherlands), 500 ppm of primary antioxidant Irganox-1010 (BASF, SE), and 600 ppm of secondary antioxidant Hostanox P-EPQ (Clariant International, Switzerland). Concentrations of peroxide (Perkadox 24L) and primary antioxidant Irganox-1010 used in this study were similar to those reported in the literature . Reactive extrusion was carried out in co-rotating twin screw compounders configured for mainly convective flow. The strategies for mixing the peroxide with the PP were different for ICP-A and ICP-B, and this resulted in the different extents of reactive modifications in these two polymers. Specifically, virgin ICP and additives were pre-mixed in a high-speed homogenizer followed by compounding in a twin-screw extruder to prepare ICP-A. A 25 mm co-rotating twin-screw compounder ZSK25WLE (Coperion, Germany) with L/D = 25 and a conveying screw configuration was used for making ICP-A. Temperature in the various zones of twin screw extruder was varied from 210[degrees]C to 240[degrees]C. The modified ICP-B was prepared by pre-mixing only the additives in a high-speed homogenizer and then co-feeding them with the virgin copolymer from separate feed ports into a 300 mm conveying twin-screw compounder of L/D = 22. Neither multifunctional acrylates nor co-agents were used in the reactive extrusion process. Temperature profile used as similar to ICP-A.
Gel Content Measurement
Samples for gel content measurements were prepared according to ASTM D2765 . While the ASTM standard suggests use of screen mesh for determining gel content, we tested screen meshes and filter papers of various pore sizes and found that a Whatman filter paper of 8 micron average pore size was best suited for gel content measurements. The screen meshes were found to allow fine microgel particles to escape thereby giving non-reproducible and incorrect gel content. On the other hand, the filter paper was found to trap microgels more efficiently and gave more reproducible values of gel content.
One gram of finely ground sample enclosed in a filter paper (Whatman, 8 micron average pore size) was immersed in 500 ml of xylene (Merck, extra pure) containing 1% by weight of Irganox-1010 (Ciba). Refluxing was carried out for 8 to 10 h at 150[degrees]C in xylene. Gel content was calculated based on the difference in the weights of filter paper before and after drying.
Separation of Rubber Phase
The polymer solution, left after filtration of microgels, was cooled to ambient temperature and was thereby separated into its subcomponent phases: a xylene insoluble crystalline PP phase and a xylene soluble amorphous ethylene propylene rubber (EPR) phase. The PP phase was precipitated out and separated by centrifugation followed by washing with fresh xylene. The EPR phase was then separated from the xylene solution by precipitation in acetone. Both the PP phase and the EPR phase were dried in vacuum oven to remove residual solvent.
Gardener Impact Tests
Gardener impact tests were carried out according to ASTM D5420  using a Gardner Impact tester (CRYOTEST, Nivtech Instruments, India). Minimum 30 specimens (injection molded discs of diameter 101.6 mm and thickness 3.20 mm) were tested for determination of Mean Failure Height (MFH) and subsequent calculation of Mean Failure Energy (MFE). Samples were conditioned at test temperature (- 30[degrees]C) for at least three hours before the test.
Differential scanning calorimeter (Model: Q100, TA instruments) was used for thermal characterization of all samples under inert nitrogen atmosphere. Three to four milligrams of sample was crimped in an aluminium pan and the same was heated from 25[degrees]C to 220[degrees]C at the rate of 10[degrees]C/min to eliminate initial thermal history. This was followed by cooling to 25[degrees]C at the same rate to obtain the crystallization temperature. The melting temperature was determined from the second heating ramp performed at a rate of 10[degrees]C/min.
High Temperature Get Permeation Chromatography
A high temperature gel permeation chromatography (HT-GPC) equipped with three detectors: refractive index, light scattering, and viscosity, was utilized to determine the molecular weight distribution and branching distribution. The light scattering detector used information from right-angle light scattering (RALS) and 7[degrees] low-angle light scattering (LALS). The viscosity detector comprised a four capillary bridge (Model: Viskotek 430, Malvern Instruments). The samples were dissolved in decalin (Anhydrous, [greater than or equal to] 99%, Sigma Aldrich) at 150[degrees]C for 5 to 6 h at concentrations of approximately 5 mg/ml prior to analysis. About 0.05% by weight of butylated hydroxytoluene (BHT) (FG, Kosher, Sigma Aldrich) was used as an antioxidant. The HT-GPC was operated at 160[degrees]C with trichlorobenzene (Reagent plus, [greater than or equal to] 99%, Sigma Aldrich) as the eluting phase using three TSK-gel columns (GMHhr-H(S) HT).
Dynamic shear rheological measurements were performed on a strain controlled ARES rheometer (TA Instruments) under nitrogen atmosphere to prevent degradation of samples. Parallel plate geometry with a diameter of 25 mm and gap of 1 mm was used for shear rheology. Disk specimens of 1 mm thickness and 25 mm diameter were molded using a 15 ton laboratory heat press (Technosearch Instruments, India) at 200[degrees]C. Time sweep experiments were first performed at a frequency of 10 rad/s for 1 h at 230[degrees]C to check melt stability. It was observed that all the samples were reasonably stable for an hour. The PP samples were then subjected to small amplitude oscillatory shear (SAOS) at temperatures ranging from 170 to 230[degrees]C over a frequency range of 0.016 to 100 rad/second at a strain amplitude of 10%, which was in the linear viscoelastic regime for all samples. The dynamic modulli, G' and G", and the complex viscosity, [[eta].sup.*], were measured as functions of the frequency, to. The dynamic rheological data was shifted horizontally along the frequency axis to the reference temperature of 170[degrees]C using the time-temperature superposition (TTS) principle to obtain master-curves. Discrete relaxation time spectrum was calculated from TTS mastercurve with the help of TA Orchestrator software of ARES rheometer.
Shear start-up experiments were carried out using MCR 301 rheometer (Anton Paar GmbH) at 170[degrees]C under nitrogen atmosphere to prevent degradation of samples. Tests were carried out at five different shear rates ranging from 0.01 [s.sup.-1] to 1 [s.sup.-1]. Cone and plate geometry with a diameter of 25 mm and gap of 49 micron was used for all the tests.
Transient uniaxial extensional rheology measurements were carried out using the Sentmanat Extensional Rheology (SER) fixture on the ARES rheometer. Measurements were performed at Hencky strain rates of 0.1, 0.3, 1.0, 3.0, and 10.0 [s.sup.-1] at a temperature of 170[degrees]C. Corrections were made for force base line drift and start time error.
Melt Flow Index (MFI), Die Swell, and Melt Strength
MFI and extrudate swell were measured on Melt Indexer 10 (Davenport, Ametek Instruments, UK). MFI was measured according to ASTM D1238  (230[degrees]C, 2.16 kg) with standard capillary die of 2.09 mm diameter and 8 mm length. Die swell was measured as percentage increase in diameter of strand with respect to diameter of capillary die at room temperature after solidification of strand. The diameter of strand was measured with micrometer.
Melt strength measurements were carried out at 200[degrees]C using a Rheotens 71.97 (Goettfert, Germany) fixture combined with a single screw extruder. Strands were extruded from a die of 2 mm diameter X 30 mm length at the rate of (0.5 kg/h) and accelerated at 30 mm/[s.sup.2]. Spinline length was 100 mm. The maximum force at which a melt strand broke was noted as its melt strength.
The transient shear and elongational rheology data for the modified ICP was fitted with the single equation extended Pom-Pom (XPP)  constitutive equation given as:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (3)
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)
The various symbols in Eqs. 1-4 are as follows: [??] represents the upper convected derivative of the stress tensor [tau], the subscript i indicates the ith relaxation mode, and for this mode [G.sub.i] is the relaxation modulus, [[lambda].sub.bi] is the orientation relaxation time of the backbone (taken as the mode relaxation time in the relaxation spectrum obtained from linear rheology), [[lambda].sub.si] is the backbone stretch relaxation time, [[LAMBDA].sub.i] is the backbone tube stretch defined as the length of the backbone tube divided by its length at equilibrium, and [q.sub.i] is the number of arms attached to the backbone of the ith mode pom-pom. Also, [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] and [alpha] is a scalar parameter. The net deviatoric stress is given by
The Rolie-Poly (RP) constitutive equation was used to fit the transient extensional and shear rheology data for the unmodified (linear chain) copolymer 1CP. The Rolie-Poly constitutive equation  containing the stretching term (RP-S) is given by:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (5)
In the above equation, [[lambda].sub.di] and [[lambda].sub.ri] are the reptation and Rouse relaxation times respectively of the ith relaxation mode, [beta] is the convective constraint release (CCR) coefficient, d is a fitting parameter and the rest of the terms carry their usual meanings. Note that for [beta] = 0, [delta] does not have to be specified. It is noted here that the stretch in the RP-S constitutive equation can become unbounded; this can be corrected using a suitable form of the RPS model that accounts for finite extensibility of chains [39, 40].
RESULTS AND DISCUSSION
Characterization of Microgels
During reactive extrusion of polypropylene with peroxide, a number of reactions can take place. Although peroxides are known to cause chain scission of PP at high temperatures, it has been reported in the literature  that reactive extrusion of PP with peroxides such as Perkadox 24L can result in branching, grafting and crosslinking reactions in addition to degradation. Excessive branching reactions occurring in a localized volume can lead to crosslinking and thereby formation of microgels. Since microgels can adversely affect several product properties such as impact resistance, optical clarity and pigment dispersibility, it is necessary to quantify the extent of microgel formation.
The results of microgel content measurements are shown in Table 1. It may be noted that these refer to the weight of microgel particles that are above an average size of 8 pm; smaller microgel particles may not be accounted for in this analysis. The results shown in Table 1 indicate that while the unmodified copolymer ICP did not contain any microgels, the modified copolymers ICP-A and ICP-B contained substantial quantities of microgel particles, and the ICP-C sample did not contain measurable quantities of microgels. The ICP-B polymer had a higher amount of gels than ICP-A. Table 1 also shows results of Gardner drop tests at - 30[degrees]C for the unmodified and modified ICPs. It can be seen that the ICP-A and ICP-B have lower mean failure heights and mean failure energy than the ICP, with ICP-B having the least resistance to impact. As might be expected, the drop impact correlates well with the gel content of the three polymers; the higher the gel content the lower is the drop impact.
Further it was observed that relative to the unmodified ICP sample, the weight fraction of the PP phase in gel-free pan of the modified samples increased slightly and the weight fraction of the EPR phase decreased correspondingly. This suggests that gelation might be taking place to a relatively larger extent in the EPR phase. However, thermal analysis of the filtered gel phase (Fig. 2) shows that the observed melting temperature Tm and the crystallization temperature [T.sub.c] correspond with those for homopolymer PP. Thus when the gel content measurements and the DSC data are considered together, it may be inferred that the gel phase is likely to contain both PP and EPR indicating that gelation took place in both phases during the reactive extrusion.
Characterization of ICPs
Table 2 shows MFI, die swell, and melt strength results for unmodified ICP and modified ICPs. It is observed that relative to the unmodified ICP, the modified ICPs have lower MFI and higher die swell. The reason for this can be attributed to an increase in melt elasticity of the modified ICPs due to presence of long chain branches. ICP-A and ICP-B samples show marginal increase in the melt strength, while 1CP-C shows considerable increase in the same. Increase in melt strength values also indicates presence of long chain branching in modified samples.
Table 2 displays melting and crystallization temperatures of all ICP samples, modified and unmodified. It is observed that the melting temperatures ([T.sub.m]) of the modified samples are similar to that of the unmodified sample. However, the crystallization temperatures ([T.sub.c]) of the modified samples were higher as compared to the [T.sub.c] of the unmodified sample. The reason for the increased [T.sub.c] of the modified samples has been attributed to the increased nucleation rate caused by the presence of branches and/or microgels in the modified samples [15, 41].
Table 2 also shows comparison of average molecular weights of all ICPs along with their polydispersity indices. It is to be noted that the polymer solutions are filtered through 0.5 [micro]m pre-filter before entering the light scattering detector. This implies that the samples injected in the GPC are free of microgels greater than 500 nm size. It is observed that the weight average molecular weights of the ICP-A and ICP-B are higher than the unmodified ICP. This agrees with their observed decrease in MFI values. Interestingly, the MFI of ICP-B is considerably lower than the MFI of ICP-A, yet the Mw of ICP-B is only marginally lower than the [M.sub.w] of ICP-A. It may be inferred that the decrease in the MFI of ICP-B is a result of larger microgel content in it as seen in Table 1. Similar trend can be seen for melt strength as shown in Table 2; ICP-A which has higher [M.sub.w] and lower gel content has higher melt strength than ICP-B. The ICP-C has comparatively the highest [M.sub.w] of the three modified ICPs.
Figure 3a shows the Mark-Houwink-Sakurada (MHS) plots of log (IV) versus log ([M.sub.w]) (IV=K [Mw.sup.a]) for the unmodified and modified ICP samples. The MHS plots are obtained from the signals of all three detectors in the HT-GPC: the Mw is derived from the light scattering and RI detectors, while the IV is derived from the viscosity and RI detectors. The MHS plots for all samples overlap in the low molecular weight region, but differ in the high molecular weight region. It is observed that for the unmodified ICP, the M-H-K exponent "a" in the high molecular weight region was 0.78, whereas for modified samples ICP-A and ICP-C, it was about 0.74 (see the inset in Fig. 3a). The reason for this decrease in "a" can be attributed to the presence of long chain branches in the high molecular weight region. Similar values of the M-H-K exponent were reported earlier for homopolymer PP and modified homopolymer PP by Lagendijk et al. .
For the ICP-B sample, the M-H-K exponent was 0.78, i.e. the same as for the unmodified ICP. Thus the ICP-B had practically no LCB content. When considered together with the gel content results discussed earlier (Table 1) it can be inferred that the reaction with peroxide in case of ICP-B resulted in excessive branching causing the formation of microgels rather than random long chain branches on the PP chains. On the other hand, ICP-A contains relatively less microgel content and a measurable content of randomly placed long chain branching on the longer PP chains. This also explains why the melt strength of ICP-B is lower than that of ICP-A (Table 2).
Characterization of PP Phase of ICPs
Table 2 shows DSC melting and crystallization temperatures for PP phase of unmodified and modified ICPs. It can be seen that melting transition for PP phase of unmodified and modified ICP is similar, whereas temperature of crystallization increases for PP phase of all modified ICPs. The result is similar to that observed for ICPs. Here also, the reason for the increased [T.sub.c] of the PP phase of modified ICPs can be attributed to the increased nucleation rate caused by the presence of branches and/or microgels in the same [15, 41].
Table 2 shows comparison of average molecular weights of the PP phase of all ICPs along with polydispersity indices. It is observed that the weight average molecular weights of the PP phase of ICP-A and ICP-B were higher than the unmodified ICP. The [M.sub.w] for the PP phase of ICP-B is lower than ICP-A. Of the PP phases of the three modified ICPs, the PP phase of ICP-C has the highest [M.sub.w]. Similar trends were observed for the [M.sub.w] values of ICPs as shown in Table 2.
Figure 3b shows the Mark-Houwink-Sakurada (MHS) plots of log (IV) versus log (Mw) for the PP phase of unmodified and modified ICP samples. The M-H-K exponents for the ICP, ICPB, ICP-A, and ICP-C were respectively, 0.77, 0.76, 0.74, and 0.73. The lower values of the exponent for the ICP-A and ICP-C indicate presence of LCB on the high molecular weight PP fraction in these samples. Similarly, the fact that the exponent for ICP-B is similar to that of the ICP implies a lower extent of LCB in the former. Similar result is observed for ICPs as shown in Fig. 3a. Thus, the PP phase of the three modified ICPs have long chain branches with the PP phase of ICP-A having similar branching extent as the PP phase of ICP-C but more branches than the PP phase of ICP-B.
Characterization of Rubber (EPR) Phase of ICP
DSC thermograms for the EPR phase of all ICPs show that the EPR phase is mostly amorphous with glass transition temperature ([T.sub.g]) of about - 50[degrees]C. Table 2 shows average molecular weights of EPR phase of all ICPs. The weight average molecular weight of the EPR phase of ICP-A was found to be slightly lower than the [M.sub.w] of EPR phase of the unmodified ICP, whereas the [M.sub.w] of the EPR phase of ICP-B was found to be considerably higher. This data, when considered along with the gel content data (Table 1) suggests that in the case of ICP-A reactive modification resulted in the formation of randomly placed branches in its EPR phase, while that for ICP-B resulted in the formation of crosslinks and microgels in its EPR phase.
Figure 3c shows the MHS plot for the EPR phase. It can be seen that while the M-H-K exponent for ICP and ICP-B were similar, the ICP-A and ICP-C showed smaller exponents of 0.57 and 0.6, respectively. The reduction in the M-H-K exponent for the EPR phase of ICP-A and ICP-C suggest the presence of randomly placed long chain branches in the high molecular weight region, with the EPR phase of ICP-A showing presence of more branching. From the earlier discussions it can be inferred that too much branching in the EPR phase of ICP-B created microgels, which were filtered out during GPC.
The mastercurve obtained at the reference temperature of 170[degrees]C from time temperature superposition (TTS) of the small amplitude oscillatory shear data for the unmodified ICP is shown in Fig. 4a while the mastercurve for the separated PP phase of the same is shown in Fig. 4b. In the low frequency regime, the frequency dependence of the storage and loss modulus curves of the total copolymer and its separated PP phase are G'~ [[omega].sup.m] (m < 2) and G" ~ [omega]. The value of m < 2 is expected for polydisperse polymers and indicates that the terminal regime may be expected at still lower frequency range. The Arrhenius model fitted the temperature dependence of the shift factors adequately for both the samples with the coefficient of regression of about 0.99. The flow activation energy values are in the expected range for polypropylene  as shown in Table 2.
TTS mastercurve for the modified copolymer ICP-A is shown in Fig. 4c and the same for the separated PP phase of ICP-A is shown in Fig. 4d at the same reference temperature as for the unmodified ICP. It can be observed from Fig. 4c and d that only horizontal shifting is insufficient for superposition. The reason for the same can be attributed to thermorheological complexity of the modified ICP-A. The thermorheological complexity is evident in the van Gurp Palmen (vGP) plot of phase angle ([delta]) versus complex modulus (G*) at different temperatures [43, 44]. Comparison of vGP plot for unmodified ICP and modified ICP-A is shown in Fig. 5. While the vGP plot for the unmodified ICP shows excellent superposition of linear visco-elastic data, the vGP plot of the modified ICP-A does not show superposition in the low G*-low [delta] region. This happens because the long chain branched chains in the modified ICP-A have a different temperature dependence of relaxation time than the linear chains. As a result, horizontal shifting alone is insufficient to obtain time temperature superposed mastcrcurves for ICP-A. Similar results have been reported for mLDPEs [45, 46] and LCB-PPs .
It can also be seen that the frequency dependence of the storage modulus data at low frequencies was considerably reduced relative to that observed for the unmodified ICP. Indeed the G' shows a tendency to flatten at low frequencies for ICP-A and correspondingly the loss tangent showed a peak. These features arise because of the heterogeneous nature of the material resulting from the long chain branching distribution as well as the presence of microgels. The removal of microgels during separation of the individual phases of ICP-A results in a somewhat reduced scatter in the mastercurve of the separated PP phase as well as disappearance of the low frequency plateau in storage modulus. The nearly equal frequency dependence of storage and loss modulus curves at low frequencies results in a plateau in loss tangent. Similar rheological features have been reported earlier for long chain branched metallocene HDPE (mHDPE-LCB)  and model comb polyethylenes . The values of loss tangent for the ICP-A and its PP phase are observed to be lower than for the unmodified ICP and its PP phase at the same frequency. Similarly, the flow activation energies for the ICP-A and its PP phase are observed to be higher than for the unmodified ICP and its PP phase (Table 2). These results are strong indications of the presence of long chain branching in the modified ICP-A polymer, which is in agreement with the HT-GPC results.
Similar results were observed for TTS mastercurves of the modified copolymer ICP-B and its separated PP phase as shown in Fig. 4e and f and for ICP-C and its separated PP phase in Fig. 4g and h. The ICP-B showed a flattening of G' at low frequency (and a corresponding peak in loss tangent), while this feature was replaced with similar frequency dependence for G' and G" (and a corresponding plateau in loss tangent) for its PP phase. This result suggests that the flattening of the G' in the ICP-B is caused by its large microgel content. Neither the ICP-C nor its PP phase showed flattening of G' at low frequency, which agrees with the fact that there is negligible microgel content in ICP-C.
Table 2 shows a compilation of the characteristic relaxation time and flow activation energies of the unmodified and modified 1CP samples and their PP phases. The increase in the characteristic relaxation time and the activation energy for the modified samples relative to the unmodified samples are evident from the tabulated values.
Time temperature superposition mastercurves for the EPR phases of all samples are shown in Fig. 6a-d. For the EPR phase of unmodified ICP the storage and loss modulli data displayed viscous dominated response in the low frequency regime [G' ~ [[omega].sup.m] (m < 2)> G" ~ [[omega].sup.n] (n < 1)]. The lower than expected values of the exponents m and n suggest that the terminal regime is likely to be observed at much lower frequencies than those probed here. Flow activation energy of the EPR phase of ICP was found to be 36 kJ/mole. For the EPR phase of 1CP-A the viscoelastic modulli showed nearly equal frequency dependence, reduced values of loss tangent and higher flow activation energy ([E.sub.a] = 52 kJ/mole), which indicates presence of long chain branches in this phase. As for the EPR phase of 1CP-B, the crossover frequency is reduced to much lower value relative to the EPR phase of the unmodified ICP or the ICP-A. This is in agreement with the high A/w observed in the HT-GPC data (Table 2). The flow activation energy of the EPR phase of ICP-B ([E.sub.a] = 46 kJ/mole) was lower than that for ICP-A. The mastercurve of the EPR phase of ICP-C did not show strong indications of long chain branching; the viscoelastic modulli did not show similar frequency dependence and the flow activation energy was lower ([E.sub.a] = 43 kJ/mole) than for ICP-A. These results are also in agreement with the HT-GPC data.
Figure 7a displays transient rheology data obtained from uniaxial extension and step-shear experiments for the unmodified copolymer ICP. The copolymer showed thinning in shear start-up flow experiments. The linear viscoelastic elongational start-up flow curve was calculated from the discrete relaxation time spectrum and is shown in Fig. 7a. The elongational viscosities measured at all applied extensional rates superpose on the linear viscoelastic start-up curve. No strain-hardening tendency was observed in uniaxial extensional experiments over the range of applied strain rates. This can be expected for a linear polymer as has been reported for linear polyolefins such as PP [25, 50] and HDPE . These results, when considered together with linear rheology (Fig. 4a) and HTGPC results (Fig. 3a), confirm that LCBs are absent in unmodified copolymer ICP. The data could be successfully fitted with the RP-S model for linear chains described earlier in the section on Constitutive Models. The model parameters were chosen such that the start-up shear viscosity data and the start-up extensional viscosity data are fitted simultaneously. The goodness of fit was decided qualitatively. Table 3 shows the RP-S model parameters for the ICP resin at 170[degrees]C.
Figure 7b shows uniaxial and shear transient rheology data together with XPP model fits for the ICP-A sample. Strong strain hardening behavior in uniaxial extension accompanied by shear thinning in step-shear was observed. At all extension rates studied, the elongational viscosity rises above the linear viscoelastic elongational start-up flow curve. The strain hardening was pronounced at the highest strain rate of 10 [s.sup.-1] but became weaker with decreasing strain rate. However, even at the lowest applied strain rates of 0.1 and 0.3 [s.sup.-1] a significant strain hardening can be observed. Once again, this confirms qualitatively the presence of long chain branches in this sample. Table 4 displays the XPP model parameters for ICP-A. In the XPP model, the parameter [GAMMA] was chosen according to the guidelines given by Verbecten et al. . The fitting of extensional viscosity data was found to be sensitive to the number of branches <7,, while the fitting of the step shear data was sensitive to the ratio of backbone relaxation time to branch relaxation time ([GAMMA]). If the discreet relaxation spectrum ([[lambda].sub.bi] - [G.sub.i]) is considered to represent molecular weight distribution, then the values of [q.sub.i] in Table 4 suggest the presence of long chain branches on higher molecular weight fraction of the copolymer. These results are in agreement with the HT-GPC result (Fig. 3a). Such a comparison between rheology and HT-GPC would not have been possible from phenomenological constitutive models which cannot provide any information about the distribution of long chain branching in the modified samples. Similar results correlating the influence of molecular architecture on elongational rheology was reported by Kurzbeck et al. .
Extensional rheology measurements were also performed on ICP-B and ICP-C samples and the results are shown in Fig. 7c and d. ICP-B displays lesser amount of extensional strain hardening as compared to that observed for the ICP-A and ICP-C samples. Also XPP model predicts fewer number of branches for the ICP-B (Table 5). While the HTGPC results for this copolymer (Fig. 3a) indicated almost no branching, the extensional rheology suggests the presence of at least a few long chain branches. The XPP model predicts similar distribution of long chain branches for ICP-C as for ICP-A (Table 6).
In the present work heterophasic impact copolymer of polypropylene (ICP) was modified by reactive extrusion in presence of peroxide. In comparison with the unmodified ICP, the modified ICP samples displayed higher flow activation energy, longer characteristic relaxation time, pronounced strain hardening and shear thinning respectively in extensional and shear rheology, increased crystallization temperature and lower M-H-K exponents. All of these indicated the presence of long chain branches in the modified ICP. Extended Pom-Pom constitutive model fits to transient rheology data of modified ICPs indicated the presence of long branches on slower relaxation modes that corresponded to the high molecular weight fraction of these samples. This result was in agreement with independent measurements of long chain branching in the modified ICPs from gel permeation chromatography.
An important objective of this work was to undertake detailed molecular characterization of the unmodified and modified ICPs in order to establish structure-property linkages. Towards this goal, the constituent polypropylene (PP) matrix phases and ethylene-propylene rubber (EPR) rubber phases of the unmodified and modified ICPs were separated and characterized for molecular weight distribution, branching content, microgel content and rheology.
The PP phase of the unmodified ICP did not show presence of long branches in HT-GPC results. Shear and extensional rheology data showed linear and non-linear viscoelastic properties that are typical of a linear-chain polymer melt. The thermal transitions were similar to a conventional homopolymer PP. In contrast, the PP phase of the modified ICPs showed presence of long branches on the high molecular weight fraction in HT-GPC data. Thermal analysis showed an increase in crystallization temperature caused by enhanced nucleation rates due to the branching. The frequency dependences of linear viscoelastic storage and loss modulli were similar, leading to a plateau in loss tangent at low frequencies. Considerable strain hardening was observed in elongation rheology. These rheological features confirm the presence of long chain branching in the PP phase of the modified ICPs.
The EPR phases of the unmodified as well as the modified ICPs were amorphous. From the Mark-Houwink-Sakurada plots obtained using HT-GPC it was observed that the EPR phase of unmodified ICP had linear chains, whereas the EPR phase of modified ICPs had long chain branching on the high molecular weight fraction. The EPR phase of unmodified ICP showed typical melt like relaxation spectrum while that of modified ICPs showed typical gel like relaxation spectrum. This suggested the presence of microgels in the EPR phase. Indeed, microgels were observed in the PP phase as well as the EPR phase of modified ICPs. The presence of microgels is most likely responsible for the lowering of impact strength of the modified ICPs.
Thus the detailed microstructural characterization of reactively modified ICPs has enabled us to correlate molecular structure with properties such as viscoelasticity and impact. This study can be of use in developing various grades of long chain branched ICPs suited for melt processing operations such as extrusion coating, thermoforming etc.
The authors thank Professor P. Sunthar, Department of Chemical Engineering Indian Institute of Technology, Bombay, for useful discussions.
[1.] N. Pasquini, Polypropylene Handbook, 2nd ed., Hanser Publications, Carl Hanser Verlag, Munich (2005).
[2.] H.C. Lau, S. Bhattacharya, and G.J. Field, Polym. Eng. Sci., 38(11), 1915 (1998).
[3.] H.C. Lau, S.N. Bhattacharya, and G.J. Field, Polym. Eng. Sci., 40(7), 1564 (2000).
[4.] J.L. Throne, Thermoforming, Carl Hanser Verlag, Munich (1987).
[5.] A.D. Gotsis, B.L.F. Zeevenhoven, and A.H. Hogt, Polym. Eng. Sci., 44(5), 973 (2004).
[6.] R.A. Steinkamp and T.J. Grail, U.S. Patent 3,862,265 (1975).
[7.] V. Braga and R. Ghisellini, European Patent Specification 0,450,342,B 1 (1991).
[8.] J. Saito, S. Kawazoe, and S. Kikukawa, U.S. Patent 5,416,169 (1995).
[9.] A.H. Hogt and H. Westmijze, WO 97/49759 (1997).
[10.] A.H. Hogt, B. Fisher, and G.K. Spijkerman, WO 99/27007 (1999).
[11.] F. Su and H. Huang, J. Appl. Polym. Sci., 113(4), 2126 (2009).
[12.] F. Su and H. Huang, Adv. Polym. Techno!., 28(1), 16 (2009).
[13.] F. Su and H. Huang, J. Appl. Polym. Sci., 116(5), 2557 (2010).
[14.] F. Su and H. Huang, Polym. Eng. Sci., 50(2), 342 (2010).
[15.] X. Wang, C. Tzoganakis, and G.L. Rempel, J. Appl. Polym. Sci., 61(8), 1395 (1996).
[16.] G.J. Nam, J.H. Yoo, and J.W. Lee, J. Appl. Polym. Sci., 96, no. 5, 1793 (2005).
[17.] J. Tian, W. Yu, and C. Zhou, Polymer, 47(23) 7962 (2006).
[18.] B.K. Kim and K.J. Kim, Adv. Polym. Technol. 12(3), 263, 1993.
[19.] J. Parent, a Bodsworth, S. Sengupta, M. Kontopoulou, B. Chaudhary, D. Poche, and S. Cousteaux, Polymer, 50(1), 85 (2009).
[20.] D. Graebling, Macromolecules, 35(12), 4602 (2002).
[21.] D. Wan, L. Ma, Z. Zhang, H. Xing, L. Wang, Z. Jiang, G. Zhang, and T. Tang, Polym. Degrad. Slab., 97(1), 40 (2012).
[22.] Z. Zhang, H. Xing, I. Qiu, Z. Jiang, H. Yu, X. Du, Y. Wang, L. Ma, and T. Tang, Polymer, 51(7), 1593 (2010).
[23.] Z. Zhang, D. Wan, H. Xing, Z. Zhang, H. Tan, L. Wang, J. Zheng, Y. An, and T. Tang, Polymer, 53(1), 121 (2012).
[24.] H. Xing, Z. Jiang, Z. Zhang, J. Qiu, Y. Wang, L. Ma, and T. Tang, Polymer, 53(4), 947 (2012).
[25.] R. Hingmann and B.L. Marczinke, J. Rheol. 38(3), 573 (1994).
[26.] R. Pivokonsky, M. Zatloukal, P. Filip, and C. Tzoganakis, J. Non-Newtonian Fluid Mech. 156(1), 1 (2009).
[27.] S. Augier, S. Coiai, T. Gragnoli, E. Passaglia, J. Pradel. and J. Flat, Polymer, 47(15), 5243 (2006).
[28.] S. Augier, S. Coiai, E. Passaglia, F. Ciardelli, F. Zulli, and M. Giordano, Polymer Int., 59(11), 1499 (2010).
[29.] F. Zulli, L. Andreozzi, E. Passaglia, S. Augier, and M. Giordano,. J. Appl. Polym. Sci., 127(2), 1423 (2013).
[30.] P.L. Fernando and J.G. Williams, Polym. Eng. Sci., 21(15), 1003 (1981).
[31.] M. Bramuzzo, Polym. Eng. Sci., 29(16), 1077 (1989).
[32.] I. Narisawa, Polym. Eng. Sci., 27(1), 41 (1987).
[33.] R.P. Lagendijk, A.H. Hogt, A. Buijtenhuijs, and A.D. Gotsis, Polymer, 42(25), 10035 (2001).
[34.] ASTM D 2756-95, "Standard test methods for determination of gel content and swell ratio of crosslinked ethylene plastics".
[35.] ASTM D5420-10, "Standard Test Method for Impact Resistance of Flat, Rigid Plastic Specimen by Means of a Striker Impacted by a Falling Weight (Gardner Impact)1".
[36.] ASTM D 1238-13, "Standard Test Method for Melt Flow Rates of Thermoplastics by Extrusion Plastometer".
[37.] W.M.H. Verbeeten, G.W.M. Peters, and F.P.T. Baaijens, J. Rheol. 45(4), 823 (2001).
[38.] A.E. Likhtman and R.S. Graham, J. Non-Newtonian Fluid Mech., 114(1), 1 (2003).
[39.] K.K. Kabanemi and J.F. Hctu, Rheol. Acta, 48(7), 801 (2009).
[40.] P.C. Roozemond, R.J.A. Steenbakkers, and G.W.M. Peters, Macromol. Theory Simul., 20(2), 93 (2011).
[41.] A.J. DeNicola Jr., J.A.Smith, and M. Felloni, U.S. Patent 5,414,027 (1995).
[42.] A.D. Gotsis and B.L.F. Zeevenhoven, and C. Tsenoglou, J. Rlieolo., 48, 895 (2004).
[43.] M.van Gurp and J. Palmen, Rheol. Hull., 67, 5 (1998).
[44.] S. Trinkle, P. Walter, and C. Friedrich, Rheol Acta, 41, 103 (2002).
[45.] F.J. Stadler, J. Kaschta, and H. Munstedt, Macromolecules, 41, 1328 (2008).
[46.] U. Kessner, J. Kaschta, F.J. Stadler, C.S. Le Duff, X. Drooghaag, and H. Munstedt, Macromolecules, 43, 7341 (2010).
[47.] J.A. Langston, R.H. Colby, T.C. M. Chung, F. Shimizu, T. Suzuki, and M. Aoki, Macromolecules, 40(8), 2712 (2007).
[48.] P.M. Wood-Adams and J.M. Dealy, Macromolecules, 33(20), 7481 (2000).
[49.] D.J. Lohse, S.T. Milner, L.J. Fetters, M. Xenidou, and M.K. Lyon, Macromolecules, 35(8), 3066 (2002).
[50.] S. Kurzbeck, F. Oster, H. Munstedt, T.Q. Nguyen, and R. Gensler, J. Rheol., 43(2), 359 (1999).
[51.] H. Munstedt and M. Laun, Rheol. Acta, 20, 211 (1981).
Kalyani Chikhalikar, (1,2) Anushree Deshpande, (1) Harshawardhan Pol, (1) Deepa Dhoble, (1) Saroj Jha, (1) Kishor Jadhav, (3) Sunil Mahajan, (3) Zubair Ahmad, (3) Surendra Kulkarni, (3) Surendra Gupta, (3) Ashish Lele (1)
(1) Polymer Science and Engineering Division, CSIR-National Chemical Laboratory, Pune 411008, India
(2) Department of Chemical Engineering, Indian Institute of Technology-Bombay, Powai 400706, India
(3) Reliance Technology Group-Polymers, Reliance Industries Limited, Mumbai 400071, India
Correspondence to: Ashish Lele; e-mail: email@example.com
Published online in Wiley Online Library (wileyonlinelibrary.com).
TABLE 1. Gel content and Gardner drop impact of ICPs. Gel content PP phase EPR phase Mean failure Mean failure Sample (wt%) (a) (wt%) (a) (wt%) height (cm) energy (J) ICP 0 83 17 40 14 ICP-A 10 85 15 33.6 12 ICP-B 19 88 12 22.6 8 ICP-C 0 88 12 No failure at max height (100 cm) (a) Based on weight of gel free sample. TABLE 2. MFI, die swell, melt strength, thermal properties, average molecular weights and rheological properties of ICPs. [T.sub.m] [T.sub.c] Melt ([degrees]C) ([degrees]C) MFI Die swell strength Sample (g/10 min) (%) (mN) ICP PP ICP PP ICP 1.5 9 128 163 163 112 119 ICP-A 0.84 52 162 165 163 127 123 ICP-B 0.44 73 142 166 167 127 125 ICP-C 0.6 80 350 165 165 128 125 [M.sub.w] (Da) [M.sub.n] (Da) x [10.sup.5] x [10.sup.5] PDI Sample ICP PP EPR ICP PP EPR ICP PP EPR ICP 3.71 3.29 3.67 0.83 0.83 1.79 4.5 3.9 2 ICP-A 4.08 4.48 3.39 0.87 1.05 1.07 4.7 4.2 3.1 ICP-B 3.93 4.2 7.59 1.15 0.93 3.24 3.4 4.5 2.3 ICP-C 4.67 4.67 3.28 1.29 1.23 1.55 3.6 3.8 2.1 Char [lambda] at FAE (kJ/ 170[degrees]C (s) mol) Sample ICP PP ICP PP ICP 0.17 0.17 37 41 ICP-A 0.37 0.37 48 45 ICP-B 0.19 0.19 40 46 ICP-C 0.29 0.58 39 43 TABLE 3. RP-S parameters for extensional and start up flow data of ICP at 170[degrees]C pa. [[lambda].sub.bi] [[lambda].sub.r,i] Mode (i) [G.sub.i] (Pa) (s) (s) 1 59,359 0.01 0.001 2 23,096 0.04 0.001 3 17,277 0.16 0.01 4 7.412 0.65 0.01 5 2,495 3 0.1 6 571 II 0.1 7 88 42 0.15 8 23 170 0.2 TABLE 4. XPP model parameters for extensional and start up flow data of ICP-A at I70[degrees]C. [GAMMA] = [[lambda] [G.sub.i] [[lambda].sub.bi]/ Mode (i) .sub.bi] (s) (Pa) [[lambda].sub.si] 1 0.01 63.430 5 2 0.044 28,011 5 3 0.2 18,992 4 4 0.9 8,234 3 5 4 2,749 3 6 18 900 2 7 80 187 2 8 357 335 1 [[lambda] [alpha] = Mode (i) .sub.si] (s) [q.sub.i] 0.3/[q.sub.i] v = 2/[q.sub.i] 1 0.002 1 0.3 2 2 0.009 1 0.3 2 3 0.045 1 0.3 2 4 0.3 1 0.3 2 5 1.33 1 0.3 2 6 9 1 0.3 2 7 40 2 0.15 1 8 357 2 0.15 1 TABLE 5. XPP model parameters for extensional and step shear data of 1CP-B at 170[degrees]C. [GAMMA] = [[lambda] [G.sub.i] [[lambda].sub.bi]/ [[lambda] Mode (i) .sub.i] (s) (Pa) [[lambda].sub.si] .sub.si] 1 0.01 90,678 5 0.002 2 0.043 37,681 5 0.0086 3 0.18 25.323 5 0.0376 4 0.81 10,898 4 0.204 5 3.54 3,407 3 1.18 6 15 1,244 2 7.68 7 66 123 1 66 8 289 514 1 289 [alpha] = 0.3/ Mode (i) [q.sub.i] [q.sub.i] v = 2/[q.sub.i] 1 1 0.3 2 2 1 0.3 2 3 1 0.3 2 4 1 0.3 2 5 1 0.3 2 6 1 0.3 2 7 1 0.3 2 8 1 0.3 2 TABLE 6. XPP model parameters for extensional and step shear data of ICP-C at 170[degrees]C [GAMMA] = [[lambda] [G.sub.i] [[lambda].sub.bi]/ [[lambda] Mode (i) .sub.bi] (s) (Pa) [[lambda].sub.si] .sub.si] (s) 1 0.01 113,725 5 0.002 2 0.042 49,880 5 0.0084 3 0.18 34,571 5 0.036 4 0.77 15,480 5 0.15 5 3.25 5,381 4 0.81 6 14 1,964 4 3.45 7 58 368 1 58 8 248 660 1 248 [alpha] = O.3/ Mode (i) [q.sub.i] [q.sub.i] v = 2/[q.sub.i] 1 1 0.3 2 2 1 0.3 2 3 1 0.3 2 4 1 0.3 2 5 1 0.3 2 6 1 0.3 2 7 2 0.15 1 8 2 0.15 1
|Printer friendly Cite/link Email Feedback|
|Author:||Chikhalikar, Kalyani; Deshpande, Anushree; Pol, Harshawardhan; Dhoble, Deepa; Jha, Saroj; Jadhav, Ki|
|Publication:||Polymer Engineering and Science|
|Date:||Jul 1, 2015|
|Previous Article:||Effect of water on the solubility and diffusivity of hydrocarbons in polyethylene.|
|Next Article:||Block and random copolyamides of poly(m-xylylene adipamide) and poly(hexamethylene isophthalamide-co-terephthalamide): methods of preparation and...|