Structure-Property Relationships in Isotactic Polypropylene
This thesis is the result of a project initiated by the Dutch Polymer Institute (DPI)1. The general theme involves "molecular characterization and morphology development in polypropylene obtained by metallocene polymerization". Throughout this investigation, the general theme was shaped and filled in by a variety of studies on structure-property relationships in polypropylene. The main properties of concern were toughness, thermal properties, and structural orientation. The relationships between primary chain characteristics and morphological features, melting and crystallization and between morphology and mechanical characteristics were established on the various levels of the morphological hierarchy described in the introduction of this thesis, i.e. molecular characteristics (primary level), crystal structure (secondary level), lamellae (tertiary level), and crystal aggregates and texture (quaternary level). The levels of morphology were influenced in several ways, e.g. by changing the amount of regio/stereo defects, using copolymers, shear, and second phases. The following paragraphs give a more detailed summary of the results. Chapter 2 describes the basic structure of semi-crystalline materials and especially the morphology of isotactic polypropylene. The various levels of polymer morphology are treated successively: primary, secondary, tertiary, and quaternary levels. After the discussion of the morphology of isotactic polypropylene and the relationships with mechanical properties two sections discussed briefly the underlying theory concerning impact measurements and x-ray analysis. Chapter 3: The crystallization characteristics and morphology of isotactic polypropylene concerning the a-, ß-, and ?-phases were studied for well-defined and fully characterized polymer samples with varying amounts of stereo- and regioirregularities. The set of samples enabled us to study separately the influence of the type of defect on the parameters of interest. A combined defect fraction (CDF) was introduced to describe arbitrary samples with both a varying amount of stereoand regio-defects. Crystal growth rates as measured by polarized light microscopy (LM) decrease linearly with the defect fraction and are much stronger influenced in samples exhibiting regio-defects as compared with stereo-defects. We found that the decreasing growth rate of the ß-phase is higher than the a-phase. We found a critical defect fraction, (Xcrit) for which the growth rates of the a- and ß-phase are equal. We found that the upper bifurcation temperature is lowered with increasing X. Shear induced crystallization of the samples was studied and compared with quiescent crystallization. The growth rates of the shear induced cylindrites were the same as the spherulitic growth rates. It was evidenced that due to the lowering of the upper bifurcation temperature no ß-cylindrites could be formed above the critical defect concentration (Xcrit). Analysis of the measurement data was made with the model of Sanchez and Eby. We found a good agreement for the samples with a variation in the number of stereo-defects, however a less satisfactory agreement was found for regio-defects. From the analysis it followed that regiodefects are more strongly discriminated against in the crystalline regions than are the stereo-defects. The excess free energy for incorporating a stereo-defect into the trigonal crystal lattice of the ß-phase is lower as compared with the a-phase, which results in a higher partitioning coefficient for the ß-phase. We found from the analysis that the bulk free energy difference for the completely defect-free (homo) polymer mainly determines the differences in growth rate dependence on the defect fraction for the a- and ß-phases, respectively. The theory correctly predicts the earlier mentioned critical defect fraction for which the growth rates of the a- and ß-phases are equal. An extra factor, which decreases the growth rate, is related to the forced introduction of defects into the crystalline phase. Chapter 4 describes the relationships between the chain architecture of polypropylene and the fusion behavior. The melting characteristics of isotactic polypropylene concerning the a-, ß-, and ?-phases were studied for the same group of samples as mentioned in chapter 3. The melting point is significantly lowered by the introduction of defects in the polymer chain and is much stronger influenced by the occurrence of regio-defects than by the presence of stereo-defects. A similar melting point depression was found for all the three polymorphs (a-, ß-, or ?-phases) independent on the number and kind of defect, and no correlation was found to exist between the partitioning coefficient and the crystallization kinetics. The model of Sanchez and Eby for copolymer melting was used for the analysis of the melting point depression, and was found in good agreement for the samples with a variation in the number of stereo-defects, however a less satisfactory agreement was found for regio-defects. From the analysis it followed that regio-defects are more difficult to incorporate in the crystalline regions than are the stereo-defects. The excess free energy for incorporating a stereo-defect into the trigonal crystal lattice of ß-phase is lower than in the a-phase, which results in a higher partitioning coefficient for the ß-phase. The next chapter (Chapter 5) maps the correlation between crystalline morphology, i.e. the lamellar thickness and impact resistance of three different polypropylene homo- and copolymers, obtained by injection molding1. The crystalline morphology was varied using three different nucleating agents. Linear elastic fracture mechanics was applied for the description of fracture resistance. The results indicate that morphological characteristics strongly influence the mechanical performance of the polymers. According to Young's theory for yielding the resistance to crack initiation, represented by KIc, correlates with the thickness of the lamellae. This suggests that the onset for yield can be described by a dislocation mechanism. The total energy absorbed during impact, GIc, is a complex function of the morphology with no single identifiable morphological characteristic that dominates its value. The heterogeneous, dispersed morphology of the polypropylene block copolymer initiates an additional energy absorption mode compared to the homo-polymer and the random-copolymer. Chapter 6 describes structure-property relationships on the quaternary level of the morphological hierarchy. It is known that friction deposited polytetrafluoroethylene (PTFE) layers are able to nucleate and crystallize isotactic polypropylene (iPP). In order to investigate the influence of PTFE on the crystallization behavior of iPP in bulk, PTFE-particles of different sizes (500 and 7 µm) in various concentrations were blended with iPP and subsequently processed by injection molding. Shear during processing led, in the case of the blend with 'large' particles, to PTFE fibers. Charpy impact tests showed a large increase in the strain energy release rate (GIc) for all PTFE nucleated samples. It was shown that this improvement in toughness was found to depend mainly on the concentration of PTFE, but less on the actual variation in PTFE-morphology. Measurements showed that the process-induced oriented iPP morphology dominates the mechanical properties for the injection-molded samples. The formation of the PTFE fibers in the blends induced, after melting and recrystallization, an unusual morphology consisting of oriented iPP lamellar crystals. In contrast to the frequently observed (trans) crystallization, in which lamellae are directed perpendicular to the fiber direction, we found a strong overall lamellar orientation parallel to the PTFE fiber direction. Chapter 7 can be considered as an outlook for further research. In this chapter a new approach to establish structure-property relationships is explored. The results in this chapter showed that based on a newly designed compression mold, it is possible to separate the influence of process induced structures on the (mechanical) properties. A comparison was made between the properties of samples made by the compression mold and samples made by injection molding. It was found that samples produced via compression molding showed negligible anisotropy in contrast to the injection-molded specimens. E-modulus and impact properties were measured for samples made with both techniques. It was shown that the processing induced morphology for the injection-molded samples exerts a pronounced influence on the mechanical properties. This illustrates the need for well-defined samples in structure-property relationship studies.