Polyolefin elastomers with isotactic propylene crystallinity.

Polyolefin clastomers, based on ethylene and its copolymers with alpha olefins such as propylene or octene, combine the desirable attributes of processibility, economy and resistance to environmental degradation. These properties, in conjunction with excellent elastomeric attributes such as large tensile elongations and quick recovery from large strain deformation, have led to widespread introduction in a host of fabricated forms and products. However, for these copolymers, the elastomeric properties are only accessible for high molecular weight polymers after vulcanization. In the absence of both of these conditions, the properties rapidly degrade. This requirement imposes both chemical and engineering limitations on the elastomers. First, they require a convenient vulcanization site on the polymer chain, typically introduced by the incorporation of a crosslinkable monomer during polymerization. Second, the high molecular weight requires the use of specialized procedures and machines for the compounding and fabrication of articles containing these elastomers. The structural and handling requirements of these conventional polyolefin elastomers have been instrumental in retarding the use of simple, high volume fabrication processes such as those used in the formation of thermoplastic films and nonwoven fabrics for these polyolefin elastomers.

The solution to these limitations of polyolefin elastomers is in the introduction (ref. 1) in June 2003, to be followed by commercialization by Exxon Mobil Chemical early next year, of the family of Vistamaxx specialty elastomers, hereinafter referred to as P-E polymers. In contrast to most known polyolefin elastomers, which contain at least 50 mole % of ethylene, these elastomers contain at least 70 mole % of propylene. A significant difference between this polymer and the established polyolefin elastomers (e.g., EPR), which are either amorphous or contain small amounts of ethylene crystallinity, is that the new elastomers necessarily contain isotactic propylene (iPP) crystallinity. This crystallinity arises from the aggregation of long sequences of propylene residues within chains where the pendant methyl group is meso to each other, The successful synthesis of these elastomers, which requires both intramolecular control of the tacticity of the insertion of the propylene units, as well as intermolecular control of the composition of the polymer, is only possible through the use of discrete metallocene catalysts. These catalysts, particularly in combination with a solution polymerization process, lead to a detailed control of the polymer characteristics, which define the properties of the polymer. The presence of a limited amount of the isotactic propylene crystallinity implies that the new polymers are thermoplastic elastomers. Structurally, at a molecular level, they display a phase separation of crystalline propylene-rich sequences within the continuum of a majority of the amorphous copolymer. These crystalline segments, which act in concert between the different chains, are virtual vulcanization sites at room temperatures leading to very high viscosities and a comparative lack of creep or deformation. However, at elevated temperatures beyond the melting point of the crystalline sites, this virtual crosslinking is removed and the material is easy to process.

The new polymers are intermediate in composition and crystallinity between the essentially amorphous EPR and the semicrystalline isotactic polypropylene (iPP). This is shown graphically in figure 1, which shows the continuum of polyolefins having both ethylene and propylene crystallinity that are now available.

[FIGURE 1 OMITTED]

The presence of the complementary blocks of elastomers in figure 1 lot both ethylene and propylene crystallinity should not indicate a similarity, beyond the levels of the crystallinity, in the properties of the plastomers and the P-E polymers. Figure 2 (a and b) shows this comparison between the clastomers, along with the totally amorphous EPR and the crystalline polyolefins, such as iPP. This comparison does not include the differences in the rheology, which are an important determinant in the processing and fabrication differences between these two elastomers. In figure 2a, which compares the tensile strength and the tension set from a 100% elongation for a range of P-E polymers, both plastomers and P-E polymers have lower tension set than EPR and iPP. However, for comparative plastomers and the P-E polymers at equivalent tensile strength, the latter have significantly better tension set. In figure 2b, the comparison of polymers at similar flexural modulus indicates an increase in the tensile elongation in the series of EPR, plastomers to P-E polymers. In summary, P-E polymers are tough polyolefins, which are uniquely soft and elastic.

[FIGURE 2 OMITTED]

Composition, tacticity and their distributions

P-E polymers are semicrystalline, elastomeric copolymers composed predominantly of propylene with limited amounts of ethylene. The concentration of ethylene is typically les than 20 wt, %. The placement of the propylene residues is predominantly in a stereoregular isotactic manner. This leads to the crystallinity in the copolymer. This crystallinity is critical. The extent of the crystallinity is attenuated by errors in the placement of the propylene and by the incorporation of ethylene. These two structural features contribute to lower the crystallinity, as measured by the heat of fusion, to less than 40 J/g. Copolymers of propylene and ethylene, which have higher levels of crystallinity, are thermoplastics and are not elastomers. During a uniaxial tensile elongation, these thermoplastic polymers undergo an irreversible deformation (commonly referred to as the yield point) and thus do not retract essentially their original dimensions on removal of the force of distention. Conversely, totally amorphous polymers of propylene and ethylene will creep during elongation and will not retract to their original dimensions. Thus, the elasticity and the elastic recovery of P-E polymers is limited by both an upper and a lower, greater than zero, bounds on the crystallinity. The polymerization catalysts and the exact condition of the polymerization determine the errors in the propylene placements, which are responsible for a part of this diminution in the crystallinity. The balance of the reduction in the crystallinity arises from the insertion of ethylene. Thus, within the family of P-E polymers that are made under similar synthesis conditions, the crystallinity is strongly and inversely correlated to the ethylene content of the polymer. This is shown in detail in figure 3a, which shows the melting point and the heat of fusion for P-E polymers made under similar polymerization conditions, but differing in the ethylene content. In conclusion, we note that at very low crystallinity levels, such as below 10 J/g, the melting point is apparently constant.

[FIGURE 3 OMITTED]

This apparent independence of melting temperature on comonomer content for the P-E polymers with high ethylene content is quite unusual and deserves further attention. DSC experiments were carried out on samples that were allowed to "age" at room temperature. The melting temperature of a polymer reflects the melting of crystals formed during primary and secondary crystallization, This frequently gives rise to multiple endotherms in a DSC heating scan, where the upper peak corresponds to melting of primary crystals and the lower peak corresponds to melting of crystals formed during secondary crystallization (e.g., annealing).

In figure 3b, the lower ethylene content copolymer exhibits a prominent melting peak at ~98[degrees]C, corresponding to primary crystallites, and a weak endotherm at ~50[degrees]C, corresponding to secondary crystallites. The P-E polymer with the higher ethylene content, however, displays a prominent melting peak at ~47[degrees]C, followed by a weak high temperature shoulder at ~60[degrees]C. In fact, P-E polymers showed a peak at ~50[degrees]C. with this peak being the weak peak for ethylene contents below 12 wt. %, and being the prominent peak for ethylene contents above 12 wt. %. This strongly suggests that the peak at ~50[degrees]C is not characteristic of the primary crystallization response of the polymers; rather it is the response of the materials to annealing at room temperature for two weeks. The net result is an apparent independence of the melting temperature on ethylene contents above ~12 wt % when only the peak melting temperature is recorded. It is curious that annealing at room temperature should yield similar melting temperatures over a range of comonomer contents. This insensitivity of the melting temperature to comonomer content could conceivably be due to one or more of the following causes:

* Crystallization of the same sequence length--such that, independent of comonomer content, a minimum sequence length always exists that is capable of crystallization at room temperature, but whose fraction decreases as comonomer content increases.

* Change in crystal structure above a certain comonomer content such that the new crystal structure could accommodate comonomer units in the crystal. It is possible that, on adding increased amounts of comonomer, the crystal structure changes from the commonly observed monoclinic (PP) or orthorhombic (PE) to a hexagonal structure.

Figure 3c shows the glass transition temperatures of P-E polymers as a function of ethylene content. The [T.sub.g] decreases with increasing ethylene content, primarily due to an increase in chain flexibility and loss of pendant methyl residues due to incorporation of ethylene units in the backbone. It is well known that polypropylene has a [T.sub.g] of ~0[degrees]C, and polyethylene a [T.sub.g] < -65[degrees]C. The addition of ethylene to a propylene polymer would therefore be expected to decrease the [T.sub.g], as is observed here. A secondary effect would be the reduction in the level of crystallinity associated with increasing ethylene content, which is expected to reduce the constraints placed upon the amorphous regions in proximity to the crystallites. Thus, an increase in ethylene content will result in a lower [T.sub.g], as well as an increase in magnitude and a decrease in breadth of the glass transition. This is clearly demonstrated in figure 3b. This is an important finding and has implications with respect to the mechanical properties as is discussed later. The decrease in the heat of fusion can be explained on the basis of reduction of crystallinity due to disruption of crystallizable sequences upon addition of defects (ethylene units).

While the previous description of the structural attributes of the P-E polymers indicates some of the limits of the composition and the crystallinity, the key distinguishing feature of these polymers, compared to the earlier attempts, is the essential uniformity of the composition and the crystallinity of the polymer along each chain (intramolecular homogeneity) and between different chains (intermolecular homogeneity). Intramolecularly, the P-E polymers are random copolymers of propylene and ethylene. They are not and are not intended to be block copolymers. This uniform synthesis is possible through the development of novel metallocene polymerization catalysts and the use of uniform solution polymerization procedures, which have a single polymerization environment for all of the polymer chains. The intramolecular random distribution of the monomers in the P-E polymers can be calculated from an evaluation of the r1/r2 reactivity ratio, by 13 CNMR. The ratio is in the vicinity of 1, indicating a uniform distribution of propylene and ethylene residues. Uniform intermolecular distribution of both crystallinity and composition requires a more extensive analysis. It is theoretically possible that in multi sited catalyst systems there could be independent differences in the composition and crystallinity among different polymer chains. These would arise from a difference in the stereo-regularity of insertion of the propylene on different chains, which emanate from different catalyst sites. Thus, two polymer chains which have the same composition of propylene and ethylene could have different crystallinity, because of the different levels of tacticity in the insertion of propylene in these polymer chains (ref. 2). The unequivocal analysis of these effects is two-fold: First, is a detailed temperature-rising thermal fractionation in a solvent to resolve differences in the crystallinity, followed by infra red analysis of the fractions to isolate the differences in composition. In our determinations using discrete steps of 8[degrees]C in the rising temperature thermal fractionation and hexane as the solvent, P-E polymers, over the entire range of crystallinity and compositions, are narrow in the distribution of crystallinity and narrow, within the limits of analysis, in distribution of composition. We have sometimes observed finite breadth in the solvent fractionation; where a single polymer fractionates in two adjacent fractions with the same analytical crystallinity and composition. We believe that this small dispersion arises from the finite breadth of the molecular weight distribution (Mw/Mn near 2.0) for the P-E polymer. It is expected that the results of thermal solvent fractionation are sometimes contamined by the difficulty in solubilizing a higher molecular weight polymer as compared to a similar polymer having a lower molecular weight. We note, in conclusion, that this structural characteristic of uniform intramolecular and intermolecular composition is extremely difficult to achieve in practice. Ziegler Natta catalysts, and even some of the earlier work on propylene polymerization using single sited metallocene catalysts (ref. 3), lead to broad intermolecular differences in composition and crystallinity. These products do not display the characteristic elongation, and more importantly the elastic recovery, of P-E polymers.

Errors in propylene insertion

The properties of the P-E polymers containing propylene crystallinity are secondarily, after composition, dependent on the extent of the creation of stereo and regio insertion errors in the insertion of propylene. Stereo errors are those that are created by the 1,2 insertion of propylene, but with a racemic orientation of the methyl group with relation to the adjacent methyl groups. This arises almost entirely from an insertion of the propylene in the 'wrong face' of the planar [sp.sup.2] olefin structure. The regio errors are those where the 1,2 insertion process is temporarily violated by the formation of a 2,1 insertion error or a 1,3 insertion error. In the former, the olefin is reversed as it enters the chain and leads to a structure where vicinal carbon atoms each contain a methyl group. These are often referred to as the head to head structure. In the other principal type of regio errors, the insertion of the propylene occurs by a 1,3 insertion such that the section of the chain contains only incorporated methylene units. These errors disrupt the crystallization of the isotactic propylene units and lead to polymers with lower levels of crystallinity. These errors, which are created by the multiple stereochemical pathways for the insertion of propylene, are primarily a reflection of the catalyst structure and the polymerization process. While several lines of characterization have analyzed and resolved the differences in the type and the frequency of errors in the polymerization of propylene, for the P-E polymers, mechanical and physical properties are most closely aligned with the absolute residual stereo-regularity of the polymer. This stereo-regularity is also a measure of the ultimate crystallinity of the polymer.

Thus, as shown in figure 4a, the polymerization of P-E polymers at high polymerization temperatures makes these mistakes in the insertion of propylene more frequent. The stereoregularity in the insertion of propylene is measured by a C13 NMR procedure, which analyzes the sets of three adjacent propylene residues (triads) for the relative orientation of the methyl groups. Isotactic propylene crystallinity corresponds to meso orientation, and thus the prevalence of the mm triads indicates an essentially isotactic propylene polymer. Conversely, the diminution in the amount of the triads leads to a large amount of atactic insertion of the propylene monomer. The correlation of the triad meso stereo regularity to the crystallinity of the polymer is true if the only contribution to the error is the stereo errors. This is generally, though not absolutely true, since other errors such as the regio errors of 2,1 insertion and 1,3 insertion, which contribute to the loss of crystallinity, are not accounted for in this calculation of the prevalence of mm triads. In addition, this procedure does not account for the loss of crystallinity of the propylene polymer that is due to addition of other monomers, such as ethylene. However, within the narrow boundaries of this experimental determination, where only a single P-E composition is attained, the crystallinity (figure 4b) of the copolymer is closely and inversely aligned with the increasing amounts of errors in the propylene incorporation. We note that all of the polymers in this determination described in figures 4a and 4b have, to the limits of the experimental determination of the composition, ethylene contents within 1.4 wt. % of each other. The determination of the differences in the fraction of propylene in mm triads is insensitive to this difference in composition for reasons we have described earlier. The magnitude of the perceived difference in crystallinity is much greater than the effect of difference in the composition. The differences in crystallinity in figure 4b are almost entirely (ref. 4) due to errors in the incorporation of propylene at higher polymerization temperatures.

[FIGURE 4 OMITTED]

Elasticity and elastic recovery

Within the bounds of composition and crystallinity that we alluded to, P-E polymers are uniquely elastic and processible. Their performance is further enhanced by their ease of compatibility with isotactic polypropylene. The incorporation of isotactic polypropylene, in small amounts as a separate blend component, leads to enhanced tensile strength, enhanced abrasion and temperature resistance for the copolymer--in this it acts much like a reinforcing filler such as carbon black in compounds of general purpose rubbers. Traditional elastomers, such as natural rubber and EPDM, which are both essentially amorphous and high molecular weight, owe their properties to strong entanglements. These provide the desired elasticity, but also provide a tremendous barrier to processing and fabrication. Rubber processing, compared to the processing of thermoplastics, is thus extremely time consuming and involved. By comparison, thermoplastics, which owe their properties at room temperature to extensive crystallinity which provides a virtual network between the chains, are easy to process at temperatures above the melting point of the polymer, but fail, for reasons indicated earlier, to provide a elastomeric tensile response at ambient temperatures. P-E polymers provide a unique combination of ease of processing, such that conventional thermoplastic processing routines and equipment can be adapted to this polymer, as well as a final fabricated product that is elastic. This combination of properties leads to the easy fabrication of elastic materials such as fibers and films, which traditionally have been made inelastic by the use of thermoplastics. This advance opens the pathway to the introduction of desirable elastic properties to a host of fabrication processes very different from either the conventional rubber processing equipment or the conventional rubber products, such as tires. We wish to emphasize that P-E polymers and their fabricated products are not only soft, but also elastic.

Elasticity is a combination of the ability of piece of a polymer to be uniaxially deformed to several times its original length and the ability to quickly retract back to its original dimension when the distending force is removed. The tensile elongations for P-E polymers are typically greater than 1,000%; this extensibility is a significant difference and much larger than noted for compositionally heterogeneous blends of propylene ethylene copolymers, which have the same average composition as the corresponding P-E polymers. This fundamental property change between materials of ostensibly the same average composition is due to the lack of dispersity in composition and tacticity for the P-E polymers. In these polymers, the entire elastomeric phase is semicrystalline and a single phase; while in the blends, there is segregation into amorphous phases and semicrystalline phases. It is easy to accept that, in the absence of strong interphase interactions in these blend polymers, tensile deformation occurs exclusively in the less crystalline phase and leads to an easy rupture of the analytical specimen. Comparison of the tensile elongation (stress-strain) data for P-E polymers with different ethylene compositions in figure 5a shows a steady progression of decreasing modulus at comparable elongation with increasing ethylene content of the polymer. A more detailed analysis shows that almost throughout this range, the elongation of the P-E polymers stays above 1,000%. Some of the less crystalline polymers have been recorded to have elongation above 1,500%. For the majority of the P-E polymers with crystallinities less than 20 J/g, corresponding to approximately 12% ethylene, the tensile elongation curves do not show any evidence of irreversible deformations during the elongation cycle. We would thus expect that on removal of the stress, these specimens should retract essentially to their original dimensions. It is also noteworthy that all of the P-E polymers, except the ones which have the least amount of crystallinity, show strong evidence for increasing tensile modulus at elongations above about 300%. We believe that this increase in the tensile modulus for polymers with intermediate levels of crystallinity implies a process of strain induced crystallization that appears for these polymers and contributes both to the extreme elongations, as well as the large value of the tensile modulus. Note that the data in figure 5a are presented in engineering units and conventions with no correction for the diminution of the cross-section of the sample during the elongation process. At the extreme elongation that these samples display, the actual stresses borne by unit cross sections of the sample in the distended state are much larger than the nominal values shown in figure 5a.

[FIGURE 5 OMITTED]

We remarked earlier about the lack of evidence for yielding on the elongation data for a majority of the P-E polymers; even though the parent isotactic polypropylene, with enhanced crystallinity, shows a very marked yield at elongations of the order of about 100% and a complete break at about 200%. This change in the behavior of the polymer is due to diminished crystallinity of the polymer through the introduction of errors in the insertion of propylene and the insertion of ethylene. A more quantitative measure of this is obtained by measuring the permanent set in a 200% cyclic elongation hysterisis cycle, measured after a test conducted at 25.4 cm/min. on both the distention and retraction cycle. The permanent set is measured to be the distention of the sample, in the direction of the elongation, expressed as a percentage of the original length of the sample measured 10 minutes after the complete cycle. These data, which are shown in figure 5b, show that: As in the case of the melting behavior, even the permanent set decreases rapidly at lower ethylene contents (<12 wt %), then seems to plateau at higher ethylene contents. The rapid change in set occurs at low ethylene contents (<12 wt. %), while the changes at higher ethylene contents are smaller.

The tensile and permanent set properties of P-E polymers exhibit remarkable mechanical annealing characteristics. This phenomenon is similar to the Mullins effect for elastomer compounds, but occurs even in pure P-E polymers, in addition to their compounds. This effect has two distinct components. First, the repeated cyclic elongation (to the same distention) results in an observed strain softening in the second and subsequent cycles. This effect has been observed for cyclic elongations in the range of 100% to 400%, but there is very little evidence to suggest that this does not persist throughout the entire elongation spectrum. The most marked strain softening is observed in the first one or two cycles. Further cycling does not lead to any observable changes in the tensile elongation characteristics. These data are shown in figure 6a, which shows the repeated cycling between 100% and 400%, with an increase in the elongation by 100% in each cycle. In each portion of each loop, corresponding to the repeat of the distention achieved in the previous cycle, there is a marked drop in the stress required for the elongation. This stress rises rapidly as the extent of elongation extends in each cycle to elongations not attained in the previous cycle. The noteworthy data are the very little additional strain softening after the second and subsequent extensions. The second feature of the mechanical annealing process, shown in figure 6b, is that the hysterisis set, which is the extent of the transient deformation in the sample during an extension and contraction cycle, rapidly diminishes between the first and second cycles to the same elongation. This transient deformation is the deformation of the sample in the direction of the elongation when the force in the retractive portion of the cycle reaches zero. We have preliminary data that further repeated elongation does not lead to any significant increase, if at all, in the transient hysteretic deformation. These two data, in combination, indicate that for the P-E polymers, the initial mechanical distension leads to a small amount of change in the polymer morphology or crystal structure. This change persists when the extensional forces are removed, and leads to a highly elastic polymer that has low extensional tensile modulus and almost complete recovery from elongation. Thus, the mechanical properties of P-E polymers are improved significantly by the mechanical work, and a true measure of repeatable properties should be conducted on samples that have undergone at least one extensional cycle. We speculate that the irreversible changes that lead to this improvement may be reorientation or agglomeration of crystallites, although the experimental verification of these hypotheses is still awaited.

[FIGURE 6 OMITTED]

Crystal structure

Figure 7a shows the wide-angle x-ray scattering (WAXS) patterns of compression-molded samples of some of the P-E polymers. Two aspects are clear from the WAXS patterns. First, the degree of crystallinity decreases with increasing ethylene content. Secondly, a decrease in the [alpha] polymorph content and a corresponding increase in the [gamma] polymorph content occurs with increasing ethylene content, as shown by disappearance of the peak at ~18[degrees] and appearance of a peak at 19.5[degrees], indicated by the dotted lines in the figure.

[FIGURE 7 OMITTED]

Figure 7b shows the small angle x-ray scattering (SAXS) data for selected P-E polymers. A peak in the SAXS pattern indicates coherent packing of lamellar structures, while absence of a peak indicates either loss of coherent stacking, or loss of lamellar structures. It can be seen in the figure that a peak in the scattering pattern is present for ethylene contents < ~12 wt. % and is absent for higher ethylene contents, This is evidence for a loss of coherent packing of lamellae for P-E polymers with ethylene ~12 wt. % and, as is shown by microcopy later, a loss of lamellar structure at still higher ethylene content (~14 wt. %). A transition from stacked-lamellar to isolated lamellar to fringed-micellar morphology takes place with increasing ethylene content.

Morphology and correlation to mechanical properties

The TEM micrographs of P-E samples with 6.8 wt. % and 14.2 wt. % [C.sub.2] are shown in figures 8a and 8b respectively. A stacked-lamellar morphology is evident for the sample with 6.8 wt. % [C.sub.2]. The sample with 14.2 wt. % [C.sub.2], however, shows no lamellar structure, but exhibits a fine-grained crystalline texture that can best be described as fringed-micellar. A sample with ~12 wt. % [C.sub.2] showed a morphology (micrograph not shown) with individual lamellae or small stacks of lamellae, with no long range correlation. The sample with 14.2 wt. % [C.sub.2] exhibits the best elastic recovery. This is an important correlation, in that it appears that the best elastic recovery characteristics are linked to the existence of a fringed-micellar morphology.

[FIGURE 8 OMITTED]

The stress-strain curves for P-E polymers with two different ethylene contents are shown in figure 9. The mechanical response of THE P-E polymer with [C.sub.2] = 5.7 wt. % is characteristic of most semicrystalline polymers. A high modulus region is followed by a prominent yield, then a draw region, and finally a strain-hardening region. The P-E polymer with [C.sub.2] = 14.2 wt. % behaves like an elastomer, which exhibits low initial modulus, no well-defined yield, no draw region and strain hardening across the elongation range. Recall that the polymer with [C.sub.2] = 5.7 wt. % showed a stacked-lamellar morphology, while the polymer with [C.sub.2] = 14.2 wt. % showed a fringed-micellar morphology. An interesting point to note is that, for the P-E polymers, the loss of prominent yield behavior takes place at the same [C.sub.2] content (~12 wt. %) at which the morphology changes from spherulitic to non-spherulitic.

[FIGURE 9 OMITTED]

The role of errors in propylene insertion on the properties of P-E polymers

Mechanical and thermal properties

One of the remaining questions in the delineation of the profiles for the mechanical properties of P-E polymers is an understanding of the effect of the errors in the insertion of propylene. We are explicit in the understanding that mechanical property changes reflect the changes in the morphology and the crystal structure with increasing amounts of ethylene in the polymer. However, in our discussion hitherto, we have not separated the contribution of monomer insertion errors in the reduction of the crystallinity of P-E polymers. An attempt to analyze this distinction required us to synthesize propylene ethylene copolymers with the same analytical attributes as the P-E polymers, such as composition and molecular weight, except under polymerization conditions and using a polymerization catalyst that minimizes the amount of errors in the insertion of propylene. These polymerization catalysts, under the conditions of our polymerization but in the absence of ethylene, would produce an isotactic polypropylcne with a melting point of about 152[degrees]C. The polymerization catalysts for the PE polymers produce isotactic polypropylene having a melting point in the range of 125[degrees]C. While these differences are qualitative, they do point out the absolute extent of errors in the insertion of propylene for these two different catalyst systems. These catalysts also show mm triad frequencies in the range of 97 to 78% over the composition range of 5.2% ethylene to 17.2 wt. % ethylene. This is somewhat larger than the corresponding range for the P-E polymer, but we caution that this determination only quantifies that portion of the errors in the insertion of the propylene which are due to stereo errors and neglects those due to regio errors.

A comparison of the mechanical properties of the two classes of propylene ethylene copolymers shows the effect of the loss of propylene insertion errors on the mechanical properties of the polymers. Figure 10a shows the permanent set of the polymers made with error-free propylene insertion catalysts. Comparison to the corresponding data in figure 5b for the P-E polymer shows that the latter have a lower permanent set at all compositions than the comparative polymers. In essence, the corresponding data sets of the two types of polymers are related by having a displacement to higher ethylene contents for the polymers made with a more stereo regular catalyst. The more stereo regular polymers in an attempt to compensate for the loss of the crystallinity from insertion errors introduce more of the amorphous portions of the chain by incorporation of an additional amount of ethylene. We find it surprising that even with this offset between the ethylene contents of the two types of polymers, the absolute lowest values of the permanent set are lower for the P-E polymer than for the polymers made with a more isotactic catalyst. The permanent set is a measure of the elasticity of the polymer and measures the ability of the polymer to retract to its original dimensions on removal of the expansive force. This feature, which has important ramifications in the applicability, development and the synthesis of these polymers, needs further investigation.

[FIGURE 10 OMITTED]

The variation of initial modulus as a function of ethylene content is shown in figure 10b for both types of polymers. Both the P-E polymers and stereoregular polymer products exhibit similar behavior. Initial modulus falls rapidly with increasing ethylene content, reaching an approximately constant value when sufficient ethylene is present to inhibit all but small levels of crystallinity. This type of behavior is typical of most semicrystalline polymers at temperatures above their [T.sub.g]. However, the stereoregular propylene ethylene polymers have a higher modulus at equivalent [C.sub.2] content, due to the higher degree of crystallinity.

Optical microscopy

The change in morphology as a function of ethylene content for a series of stereo-regular copolymers is revealed by the polarized optical micrographs shown in figure 10c. As the ethylene content increases, the morphology changes from spherulitic to a grainy birefringent to a non-birefringent structure indicating loss of a well-defined superstructure. In these stereo-regular samples, the transition away from a spherulitic morphology occurs between 11.5 and 14.7 wt. % ethylene. In the P-E polymers, the transition occurs before 12.4 wt. % ethylene.

These data, in conjunction with the SAXS results discussed earlier indicate that, at low ethylene contents, the polymers exhibit a lamellar morphology that is arranged in a spherulitic superstructure. At higher ethylene contents, the lamellar morphology is not arranged in a larger superstructure; instead, the lamellae are present in local stacks with no long-range correlation. The ethylene content at which this transition occurs is consistent with the data obtained from SAXS. At still higher ethylene contents, the crystalline regions occur as isolated micellar crystallites, as shown by the TEM results discussed previously. Interestingly, the transition from spherulitic to lamellar morphology occurs over the same range over which the melting temperature becomes independent of ethylene content.

Conclusions

The P-E polymers combine an unusual and unexpected degree of elongation and elastic recovery without the need for crosslinking. These properties mimic those of crosslinked elastomers, while the processing and fabrication should be similar to that of the conventional polyolefins, such as polypropylene. Based on the structural information obtained from various techniques, a coherent picture of the morphology of P-E polymers with increasing ethylene content has been generated. The details of the structure and morphology have been used to explain the observed physical-mechanical properties

P-E polymer catalysts introduce substantial errors in the insertion of propylene, leading to much diminished crystallinity at equivalent ethylene content than more stereoregular polypropylene polymerization catalysts. The stereoregular polymers are shown to be different from the P-E polymers at equivalent ethylene content due to differences in defect content in the PP sequences. This appears to be an important parameter in the determination of the elastic properties of the polymer. The most striking changes in the morphology correlate to the appearance of elastic properties, since the morphology of the polypropylene crystallites changes from spherulitic to individual lamellar to fringed-micellar with increasing [C.sub.2] content, as observed by TEM, optical microscopy and SAXS. Wide-angle x-ray scattering showed that increasing [C.sub.2] content led to a decrease in [alpha]-crystallinity and an increase in [gamma]-crystallinity. Although the degree of crystallinity obtained by DSC and WAXS decreased monotonically with [C.sub.2] content, mechanical properties like modulus and permanent set decreased rapidly up to the morphological transition and thereafter are weakly dependent (or independent) of [C.sub.2] content. The mechanical properties of interest have been explained on the basis of the morphology of the polymer. Differences in properties above and below a specific [C.sub.2] content have been attributed to the morphological transition. For P-E samples and stereoregular samples, the transitions occur at ~12 wt. % and ~15 wt. % [C.sub.2] respectively.

In a spherulitic or stacked-lamellar system, the morphological superstructure occupies all volume, thus forming a framework of crystalline regions, inside which the amorphous regions are present. The system thus behaves like a rigid system. When stress is applied to such a system, all crystallites experience stress and it is not possible to deform or destroy one crystallite in isolation. The net result of this cooperative deformation is the occurrence of localized yielding within a cross-sectional slice of the sample. Once yielding has occurred within a given slice, it destabilizes the neighboring slices and deformation occurs sequentially along the length of the sample, a process observed as necking. During this process, new crystallites form that stabilize the deformed structure, thereby preventing elastic recovery.

In a fringed-micellar or individual lamellar system, the crystallites have no long-range correlation with each other, and thus do not form a framework. In this case, the morphology principally consists of a non-crystalline matrix, the chains of which are connected by physical crosslinks comprising isolated crystallites and entanglements. In such samples, the crystallites are free to deform independently of their neighbors. In the case of P-E polymers, the [T.sub.g] is at least 40[degrees]C below room temperature. As a result, the non-crystalline phase is soft and mobile, thereby making available an entropic mechanism for recovery. Upon application of a stress, low levels of deformation can occur simultaneously throughout the length of the sample, resulting in affine deformation. The crystallites act as physical crosslinks, endowing the sample with good elastic recovery. In addition to crystallites, entanglements can also act as crosslinks on the timescale of the experiment and are principally active at higher strains, contributing in strain hardening behavior.

References

(1.) Vistamaxx specialty elastomers were announced on Jane 24, 2003, at the National Plastics Exposition, Chicago, IL.

(2.) L. Resconi and F. Piemontesi, U.S. patent 5747621 (5/5/1998) assigned to Montell Technology.

(3.) (a) R.J. Waymouth, Science 267, 217 (1995). (b) R.J. Waymouth, et al, Macromolecules 31, 6,908 (1998).

(4.) S. Bruckner, S. V. Meille, V. Petraccone and B. Pirozzi, Prog. Polym. Sci., 16, 361-404, (1991).

(5.) According to the description of European Application 0374695A2 (Davis, S.C.).