Effect of the compatibilizer on the physical properties of biaxially deformed in situ composites.

INTRODUCTION

For high performance composite systems, fibrous fillers (usually glass fiber or carbon fiber) have been used. However, the processing and fabrication of fiber composites present some technical difficulties, such as wear (abrasion on screw and barrel walls of processing machinery), increased viscosity, difficulties in compounding, and breakage of the solid fibers. which reduces the aspect ratio and consequently reduces the performance of the end product. For these reasons, self-reinforced composites have been sought to avoid the drawbacks of the fiber-reinforced system (1-9).

Self-reinforced composites, which are produced by blending thermotropic liquid crystalline polymers (TLCPs) with less expensive thermoplastics, offer the full advantage of using many of the desirable TLCP characteristics. Deformation of the TLCP domains dispersed in a matrix polymer into fibril shapes during processing produces a so-called in situ composite because of in situ shaping during processing (1-5). Because such in situ composites can solve some problems that arise during the processing of conventional fiber-reinforced composites, they have attracted a great deal of interest. However, they also have some drawbacks. Among the problems of in situ composites, one of the worst is poor performance in the transverse direction. Most previous studies have been concerned with deformations of TLCP domains in a uniaxial elongation process (4-9). A drawback of this process is the high degree of anisotropy, which provides exceptional properties in the flow direction, but poor properties in the transverse directi on (7-12). Other ordinary processes, such as injection molding and extrusion, also produce products whose properties in the flow direction may be outstanding because of a deformed TLCP phase, but the anisotropy usually present is such that the transverse properties are grossly inferior (13, 14).

One method of reducing this anisotropy is to use biaxial deformation. Provided the biaxial orientation can be produced without excessive deterioration of the properties in the flow direction, the polymers will have a wider range of applications. Another problem common to most TLCP blends is that most thermoplastics are incompatible with TLCPs. This incompatibility between the matrix polymers and the reinforcing TLCPs leads to poor interfacial adhesion, which leads to a reinforcing effect less than that expected from the law of mixtures (6). Compatibility between the matrix polymers and the reinforcing TLCPs has been sought and directed at determining the optimal reinforcing effects of TLCP domains to solve the incompatibility problem in in situ composites (6-19). One method of compatibilizing an immiscible system is to use a third component as a compatibilizer or a coupling agent.

Recently, some researchers have applied the film blowing process to obtain a more or less well-balanced in situ composite (20-23). However, poor interfacial adhesion problem was still a problem. In our previous study, we investigated the effect of the compatibilizer on the blown films morphology (24). The objective has been to apply the biaxial deformation process (film blowing process) for investigating the structure development in the dispersed TLCP phase and the effect of a compatibilizer on the biaxial deformation. The present paper examines the compatibilizer's role in the biaxial deformation process and the change of physical properties of the in situ blown films. Correlations of the physical property change with the structure development (hence the morphology) are examined.

EXPERIMENTAL

The TLCP used was a copolyester amide of 6-hydroxy-2-naphthoic acid (60%). terephthalic acid (20%), and aminophenol (20%), commercially known as Vectra [R] B950 (VB) manufactured by Hoechst Celanese. This material has been used and characterized by many researchers (6-9, 11, 25-34). It was supplied in the form of pellets. Poly(ether imide) (PEI), commercially known as Ultem [R] 1000, an amorphous polymer made by G.E. Plastics, was used as the matrix. The compatibilizer used in this study was a poly(ester imide) (PEsI). Details of its synthesis scheme and miscibility with VB and PEI were filly described in our previous study (7). Scheme 1 shows the chemical structures of these polymers.

The PEI and VB pellets were dried in a vacuum oven at 120[degrees]C for at least 24 hours before use, and steps were taken during processing to minimize exposure to atmospheric moisture. The TLCP content was kept at 10 wt%. The amount of PEsI was kept to less than 1.5 wt% (7). Blending was carried out in a 42 mm Brabender twin-screw extruder (AEV651) at a fixed rotation speed of 20 rpm. At the end of the extruder, a connector and an annular die (Brabender No. 8) were attached for film blowing. The annulus had a slit thickness [h.sub.0] = 1 mm and an inner radius [a.sub.0] = 26 mm. The apparatus was also equipped with an adjustable film tower with a guide, nip and take-off rolls, and a Brabender torque winder onto which the sheet was subsequently wound. The extrusion temperatures of the feeding zone/transporting zone/melting zone/die were set as 240/330/330/335[degrees]C, respectively. The connector was also wrapped with a heating band. The temperature of the connector and the annular die were set as 330[degre es]C. The die temperature could not be varied to any large extent because of the solidification in the narrow annular gap at lower temperatures and because of insufficient melt strength at higher temperature. The blown parison was encircled with a poly(tetrafluoroethylene) bag around It to keep the hot atmosphere, which could prevent early solidification. The major adjustable parameter was the pressure of the nitrogen blown into the bubble. The blow-up ratio (ratio of the diameter of the final bubble to that of the annulus) was controlled in the range of 1 to 3. Though both the thermal history and the deformation history have a significant effect on the final structure in and the final properties of the blown film, we could not vary them over a wide range because the processing window was quite narrow for the same reason as the die temperature. By the same token, cooling air was not used in this study. For removal of the anisotropy, use of a counter-rotating annular die is desirable (23), but our equipment ha d only a non-rotating die. This should be pursued in the future.

A dynamic mechanical thermal analysis (DMTA) of the blends was carried out with a dynamic mechanical thermal analyzer (Polymer Laboratories Model 2) at a frequency of 1 Hz. The dynamic tensile mode was used. Differential scanning calorimetry (DSC) studies of the thermal property characteristics were performed on a DuPont 910 DSC controlled by a 9900 thermal analyzer. Every thermogram was repeated at least twice to verify the reproducibility of the measurements.

Rheological properties were measured using a UDS200 (Physica, Germany) rheometer on which 25mm-diameter cone and plate were mounted. The frequency range was set at 0.1-500 rad/sec and the applied strain was 10%. Before the measurement, the samples were prepared using a compression molder at 320[degrees]C. Measurements were done under nitrogen atmosphere.

Testing of the mechanical properties of the blown ternary-blend films (5 mm X 40 mm in size) was performed using an Instron Universal Testing Machine (model 4204) at a constant temperature of 23[degrees]C. A gauge length of 30 mm and a crosshead speed of 10 mm/min were used. All the reported results are averages of at least 10 measurements. Impact testing was done for a 5 cm x 5 cm samples using a pneumatic driving, instrumented impact tester (Radmana ITR2000) at a constant temperature and humidity. The impact strength of the blown film was measured following the ISO7765 method (35). The specimen support was a hollow cylinder. The axis of the cylinder coincided with the line of fall of the striker and a soft shock-absorbing disc was placed inside the cylinder to rest on its base. The specimen was tightly clamped onto the support. All the reported results are averages of at least seven measurements.

Observations of the composite film morphologies were performed on a Hitachi S-2500 model. The samples were fractured in liquid nitrogen and were coated with gold to enhance the phase contrast.

RESULTS AND DISCUSSION

The glass transition temperature ([T.sub.g]) of PEI is known to be 218[degrees]C and that of VB is 142[degrees]C. (13) Because of the small amount of VB, the [T.sub.g] shifting was not so vivid in DSC thermograms. In our previous study, thermograms by DMTA measured in bending mode showed definite [T.sub.g] movement of PEsI toward that of other polymers (PEI and VB) (7). Figure 1 shows that there appears a slight movement of tan peak (from 220[degrees]C to 210[degrees]C) corresponding to [T.sub.g] of PEI in DMTA isotherms of the ternary blended film measured in the tensile mode, but the movement of VB's [T.sub.g] is a little bit obscure (from ca. 142[degrees]C to 148[degrees]C). In our previous study (7, 24), we showed the effect of the compatibilizer (PEsI) on the morphology of PEsI/VB blends under uniaxial and biaxial deformation. Though the effect of the compatibilizer was vividly observed, we then lacked detailed information on the interaction between PEsT and PEI, and VB. Recently we carried out molecular dynamics simulation (36). According to the simulation results, miscibility of PEsI with PEI and VB comes from its molecular structure. PEsI chains have a disordered structure in themselves (interaction energy per repeating unit was -10.5 Kcal/mol) while they are more stable with PEI (interaction energy per repeating unit was -30.4 Kcal/mol) and with VB (interaction energy per repeating unit was -41.5 Kcal/mol). The repeating unit of PEsI is in an almost ordered state with VB. Though it has a less overlapping structure with PET, PEsI is still in a more ordered state with PEI molecules than themselves. The binary (PEsI/PEI or PEsI/VB) blend system is quite interesting since PEsI chains favor mixing with PEI or with VB rather than homopolymers. More details about the miscibility of these blend pairs and their structures will be reported elsewhere (36).

Knowledge of the rheological behavior of polymer blends is essential for understanding the interaction between the matrix and dispersed phases. It is also pivotal for the understanding of the deformation of the dispersed phase. It is well known that the viscosity of a TLCP is lower than that of many engineering plastics (6-13). Figure 2 shows the dynamic viscosity of PEI, binary and ternary blends at 290[degrees]C and 320[degrees]C. Several facts are worthy of note. The general behavior of PEI viscosity follows that of a typical thermoplastic polymer, i.e., Newtonian behavior at low shear rate and shear-thinning at high shear rate. Shear thinning effect appears more vividly at low temperatures because of high viscosity. The binary blend has a lower viscosity than PEI. It is well known that the viscosity of TLCP blends decreases with the TLCP amount because of the processing aid role played by the TLCP phase (1, 6). This is different from usual immiscible polymer blends, which show higher zero shear viscosity values than that of the matrix. For ternary blends, the viscosity is even lower than that of the binary blend. However, the ternary blend including 0.6 wt% PEsI shows even lower viscosity values than those of the ternary blend of 1.3 wt% PEsI. This is ascribable to the optimum morphology in the ternary blend of 0.6 wt% PEsI and to more interaction between the dispersed phase and the matrix (24). Han et al. (37) experimentally showed that plots of the logarithm of zero shear viscosity versus blend composition for miscible blends under isothermal conditions exhibit negative deviations from linearity for binary blends having large negative values of the interaction parameter and positive deviations from linearity for blends having extremely small positive values of the interaction parameter. In another paper, we reported that a ternary blend of 0.6 wt% PEsI has the most uniform and finely dispersed VB phase (10 wt%), whereas the ternary blend including 1.3 wt% PEsI showed coalescence and agglomeration of the dis persed VB phase (24). This is because of excess compatibilizer, which leads to a larger VB phase in the matrix (7, 24).

Other rheological properties (the dynamic storage modulus (G') and the dynamic loss modulus (G")) show similar behavior (Fig. 3). The ternary blend including 0.6 wt% PEsI shows the lowest elasticity. Log G' vs. log [omega] and log G" vs. log [omega] for the PEI and blends lie in the terminal region and plateau region. Because the VB phase is in the nematic state and there is no structural change of the VB, the viscoelastic responses of the blends are much the same. Though not shown here due to the space limit, the crossover point between G' and G" appears at higher frequency with the temperature, which means the terminal zone starts at higher frequency for high temperature due to more molecule movements. Addition of VB to a thermoplastic matrix, PEI, reduces both the storage modulus, G', and the loss modulus, G". It is worth noting that both the storage and loss moduli of ternary blends are lower than those of the binary blend, and the optimum ternary blend containing 0.6 wt% PEsI has the lowest values. This is ascribable to more uniform dispersion of VB phase in the PEI matrix for that ternary blend. This is also related to lower viscosity of the binary and ternary blends than PEI.

In order to observe the effect of temperature on the rheological properties of the blend, a logarithmic plot of the dynamic storage modulus (G') vs. the dynamic loss modulus (G"), a modified Cole-Cole plot, is considered (Fig. 4). If the blend is totally compatible, the curves will overlap to form one line (37). Also, from the time-temperature superposition principle, it is well known that no shift factor is required for homopolymers to obtain a master curve for quantities not including any units of time (38). This means that a plot of one such quantity versus another will be temperature-independent. For example, plots of log G' vs. log G", with each point corresponding to a different frequency, are temperature invariant. This type of plot shows the relation between rheological properties and the molecular structure, such as molecular weight distribution and molecular weight, excluding the effect of processing variables. Han and coworkers have demonstrated that the curves log G' vs. log G" plots are very sens itive to a variation in the morphological state of multi-component/multi-phase polymer systems (37-39). It can be seen in Fig. 4 that log G' vs. log G" plot for PEI becomes independent of temperature and the slope is close to 2, which is expected for a homogeneous polymer. However, the slopes of other curves are apparently less than 2, which indicates that they are not homogeneous. Most blend data closely fall on to a single curve for different temperatures (290[degrees]C and 320[degrees]C) except near the terminal zone. This indicates that the morphology of the blends do not vary with the temperature in this temperature range. In the terminal zone, curves of blends show a slight curvature. This is ascribable to the heterogeneous phase inclusion (VB). It appears more vividly at high temperature (340[degrees]C) where more phase separation is obvious. The slope of the ternary blend of 0.6 wt% PEsI in the terminal zone at 320[degrees]C appeared closer to that of the Ultem, which indicates more uniform distributi on of VB phase. The nematic crystalline formation temperature of VB is ca 282[degrees]C. AT 290[degrees]C, the VB is in the nematic phase. Since the nematic-isotropic transition temperature of the VB phase is above the thermal degradation temperature, the VB phases in the blends are all in the nematic state. Hence, in spite of morphological difference, their microstructures are the same. which is the reason why the blend curves display similar behaviors.

The morphological difference between binary and ternary definitely affects their physical properties. Since details of the blown film's morphology were already reported in our previous paper (24), we briefly report the morphological observation for clarity of the discussion. Figure 5 shows the fractured surfaces of the binary and the ternary blends both in the machine direction (flow direction) and the normal direction (thickness direction). The micrograph of the binary blend film displays poor adhesion between the two phases, which leads to an open ring around the VB phase and to holes formed by pulling out the VB domains during the fracturing process. The surface of uncompatibilized binary blend, which is fractured normal to the flow direction, shows relatively large VB domains, indicating a poor dispersion (40, 41). In contrast, the VB phase in the ternary blends is more evenly distributed and finer in size than that of the binary blend. The fracture is seen to occur more within the VB phase in the compat ibilized blend, and not many open rings exist around the VB domain, reflecting a better binding between the two phases because of the compatibilizer's existence at the interface. All of the dispersed VB domains are deformed into long stripes. The VB phase in blown films can be seen to be mostly elongated in the flow direction due to unequal biaxial deformation, which indicates that more deformation occurred in the flow direction owing to strong pulling at take-up unit, and also to be flattened in the plane of the film. This is an evident difference between uniaxial deformation (extrudate drawing), equal planar biaxial deformation, and unequal biaxial deformation (film blowing).

The SEM micrographs of blown-film samples broken parallel to the flow direction reveal that the layers formed by the dispersed stripes are thinner for the compatibilized blends due to a more uniform and fine distribution of the VB phase. In the binary (uncompatibilized) blend, the long VB stripes are bundled together. Also, the stripe surfaces look clean and smooth along the flow direction, which indicates poor adhesion between the VB phase and the matrix. In contrast, the compatibilized blown-film containing 0.6 wt% PEsI exhibits finer and thinner stripes with rough and rugged surfaces. This definitely indicates good adhesion between the VB stripes and the matrix phase. However, the VB domains coalesce when excess PEsI is added. Thick bundles of stripes appear as excess PEsI (1.3 wt%) is added, but their surfaces were still rough due to strong interactions.

The tensile strength and the tensile modulus of the film in the flow direction are shown in Fig. 6. The tensile strength and modulus are considerably increased with the addition of 10 wt% VB. Also, the tensile strength and the tensile modulus measured along the flow direction increase with the blow-up ratio (ratio of the diameter of the final bubble to that of the annulus). Since the accurate values of the pulling force were not measured, quantitative analysis for the film blowing process was not done. The pulling force was controlled by the roller speed of the take-up unit. A higher blow-up ratio requires a higher pulling force that leads to more orientation of the VB phase in the flow direction (20-23, 42). In this sense biaxial deformation in this film blowing process Is unbalanced deformation, of stronger stretching in the flow direction. Tensile strength and modulus increase with the addition of TLCP because of in situ formation of TLCP stripes in the blown films. In addition, a systematic increase in th e machine direction modulus and strength is observed with the addition of the compatibilizer because of the formation of finer, more uniformly distributed stripes. Addition of the compatibilizer (PEsI) also promotes improved adhesion at the interface, leading to better stress transfer to the dispersed phase (43). The tensile strength is considerably increased when an optimum amount of PEsI (0.6 wt%) is added, but decreases when excess PEsI is added (7, 24). The tensile modulus of the compatibilized blown-films shows a similar behavior and exhibits a peak when the proper amount of compatibilizer is added. In spite of experimental errors, this trend is manifest for all different blow-up ratios.

The effect of the compatibilizer appears more vividly in the mechanical properties of the blown films in the hoop direction. Figure 7 shows the tensile strength and modulus of the blown films in the circumferential (hoop) direction. The effect of the compatibilizer is quite dramatic. The tensile strength decreases with the addition of VB to PEI due to their immiscibility, but addition of the compatibilizer changes that significantly. In fact, the tensile strength along the hoop direction of the in situ composite film (binary blend film) is less than half that of pure Ultem film. This kind of anisotropy is consistent with widely reported results for in situ composites. TLCP chain segments align preferentially along the flow or machine direction because of their rigid backbone structures and the ease with which they orient themselves (6, 7). As we already mentioned, biaxial deformation in this film blowing process is not equal in flow direction and in hoop direction. Strong deformation in the flow direction due to high pulling force dominates. Therefore, tensile properties in the flow direction are dominated by the reinforcing effect of the TLCP stripes, but those in the hoop direction are affected more by the adhesion at the interface because poor adhesion between the matrix and the dispersed VB phase acts as a defect. This can be verified by comparing changes in tensile strength and tensile modulus. Though the tensile modulus of the binary blend is higher than that of an Ultem film, its tensile strength is definitely lower because of poor interfacial adhesion. Addition of PEsI improves this; it enhances the tensile strength because of better adhesion at the interface. When 0.6 wt% PEsI is added, the tensile strength of the ternary blended film markedly increased, the largest value approaching that of Ultem film. Even when an excess amount of the compatibilizer is added, the tensile strength is much higher than that of the binary blend although the moduli of the two are similar with each other. The blow-up ratio e ffect is not so dramatic because the deformation was limited, but a higher blow-up ratio clearly associated with more deformation in both directions, hence higher modulus and higher tensile strength.

The impact strength of blown films with a blow-up ratio of 2 is shown in Fig. 8. The impact strength of the binary blend is much lower than that of the Ultern film which is amorphous and tough. The impact strength of the compatibilized blend containing 0.6 wt% PEsI is almost twice that of the binary film. Also, the ternary-blend film containing an excess amount of PEsI (1.3 wt%) shows higher impact strength than the binary system, which is similar to the behavior of in situ composite fibers (7). In a non-compatibilized system, propagating stress passes around the dispersed VB phase boundary because the VB and the matrix (Ultem) are immiscible and the interfacial adhesion is poor. In the compatibilized ternary-blend film, the stress acting on the matrix is transmitted to the VB phase across the interface. Excess energy is consumed by plastic deformation of the VB phase. Good adhesion provides a marked improvement in the impact strength. On the other hand, a film with excessive compatibilizer shows less impact strength due to both coalescence and a less uniform dispersion. This trend is similar to that of the tensile properties in the hoop direction.

As we have already seen in the previous study (24), addition of the compatibilizer increases the width of the dispersed phase in the hoop direction. As a result of both this widening and the good adhesion at the interface, the mechanical properties in the transverse direction are significantly improved, and some blends show almost balanced properties in both directions. However, their general properties are not fully maximized. Even though the addition of a compatibilizer induced remarkable changes in the properties of blown films, films with properties well balanced in both the flow and the hoop directions have yet to be produced. This study suggests a possible way to reach that goal. A recent study shows that more balanced TLCP-containing films can be obtained by employing a counter-rotating mandrel die in the film blowing process (22, 23). Using both a counter-rotating mandrel and the addition of a compatibilizer may remedy the anisotropy problem of in situ composite films of TLCP containing films, and mer its further study.

CONCLUSIONS

The compatibilizer's decisive role in the physical properties of biaxially elongated in situ composites (blown films) of TLCP ternary blends was investigated in this study. The miscibility of PEsI comes out from the molecular ordering structure of PEsI with PEI and with VB. Due to the molecular conformation, PEsI is more stable when mixed with PEI or VB than its homopolymers. Noticeable thermal property change was not observed in the blends. Rheological properties evidently show that the blends are heterogeneous, but become compatible with the matrix by adding the compatibilizer. Viscosity of the ternary blend is lower than that of the binary blend while that of the optimum ternary blend of 0.6 wt% PEsI is even lower than the ternary blend of 1.3 wt% PEsI (excessive compatibilizer added system).

Coalescence during the blending process occurs when an excessive amount of the compatibilizer is used, resulting in a larger particle size. The distributions of the particle sizes in the dispersed phase were generally broad for the ternary blend of excessive compatibilizer. The morphology evidently shows that ca. 0.6 wt% PEsI provides the best morphology when 10 wt% VB phase is included. Mechanical properties show the maximum at this composition.

The mechanical properties are significantly improved by adding the compatibilizer. This is ascribed to the compatibilizing action of PEsI, which leads to better adhesion, better stress transfer, and a more uniform dispersion of the VB phase. Improvement in the hoop-direction tensile properties is quite considerable for the compatibilized blown-films, especially when the proper amount of the compatibilizer is added. The impact strength data also supports the importance of the compatibilizer, the maximum strength being almost twice that of the binary blend when the proper amount of PEsI is added. Thus, from a processing point of view, it is highly desirable to have the proper (optimum) amount of the compatibilizer for the best dispersion and the most uniform deformation of the dispersed phase, hence the best properties. The anisotropy seen in TLCP-containing blend films may be possibly overcome by adopting a film blowing process using a counter-rotating mandrel die (22, 23) along with the addition of the compat ibilizer; this has yet to be investigated.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

ACKNOWLEDGMENT

Financial support by Ministry of Commerce, Industry and Energy (2M11320) is greatly appreciated. Special thanks go to Mr. Sehyun Kim for his help with experiments.

REFERENCES

(1.) F. R. La Mantia, ed., Thermotropic Liquid Crystal Polymer Blends, Technomic Publishing, Lancaster, Pa. (1993) .

(2.) A. I. Isayev and T. Limtasiri, in International Encyclopedia of Composites, Vol. III, S. M. Lee, ed., VCH Publishers, New York (1990) .

(3.) D. J. Williams, Adv. Polym. Tech., 10, 173 (1990).

(4.) V. Handlos and D. G. Baird, J. Macromol. Sci., Rev., C 35 (2), 183 (1995).

(5.) C. Carfagna, E. Amendola, and M. R. Nobile, International Encyclopedia of Composites, Vol. II, S. M. Lee, ed., VCH Publishers, New York (1990) .

(6.) Y. Seo, S. M. Hong, and K. U. Kim, Macromolecules, 30, 2978 (1997).

(7.) Y. Seo, S. M. Hong, S. S. Hwang, T. S. Park, K. U. Kim. S. Lee, and J. W. Lee. Polymer, 36, 515 and 525 (1995).

(8.) Y. Seo, B. Y. Kim, S. Kwak, K. U. Kim, and J. Kim, Polymer, 40, 4441 (1999).

(9.) Y. Seo and K. U. Kim, Polym. Eng. Sci., 38, 596 (1998).

(10.) W. Lee and A. T. DiBenedetto, Polymer, 34, 684 (1993).

(11.) H. J. O'Donnell and D. G. Baird, Polymer. 36, 3113 (1995).

(12.) A. Datta, H. H. Chen, and D. G. Baird, Polymer, 34, 759 (1993).

(13.) Y. Seo, J. Appl. Polym. Sci., 70, 1589 (1998).

(14.) R. K. Krishwaswamy and D. G. Baird, Polymer, 40, 701 (1999).

(15.) D. Dutta, R. A. Weiss, and J. He, Polymer, 37, 429 (1996).

(16.) M. M. Miller, J. M. G. Cowie, D. L. Brydon, and R. R. Mather, Polymer, 38, 1565 (1997).

(17.) M. Singer, G. P. Simon, R. Varley, and M. R. Nobile, Polym. Eng. Sci., 36, 1038 (1996).

(18.) K. Wei, W. Hwang, and H. Tyan, Polymer, 37, 2087 (1996).

(19.) Y. Seo, B. Y. Kim, and K. U. Kim, Polymer, 40, 4483 (1999).

(20.) K. G. Blizard, T. S. Wilson, and D. G. Baird, Intern. Polym. Proc., V. 53 (1990).

(21.) W. Chinsirikul, T. C. Hsu, and I. R. Harrison. Polym. Eng. Sci., 36, 2708 (1996).

(22.) T. C. Hsu, A. M. Lichkus, and I. R. Harrison, Polym. Eng. Sci., 33, 860 (1993).

(23.) R. W. Lusignea, in Thermotropic Liquid Crystal Polymers, T. S. Chung. ed., Technolic, Lancaster, Pa. (2001).

(24.) Y. Seo and J. Kim, Polymer, 42, 5029 (2001).

(25.) A. Boersman, J. van Turnout, and M. Wubbenhorst, Macromolecules, 31, 7453 and 7461 (1998).

(26.) T. De Neve, P. Navard, and M. Klemen, J. Rheology, 37, 515 (1993).

(27.) A. Valenza, F. P. La Mantia, M. Paci, and P. L. Magagnini, Intern. Polym. Process., 6, 247 (1991).

(28.) G. W. Calundann and M. Jaffe, Proc. of the Robert A. Welch Foundation Conferences on Chemical Research, 26, 247 (1983).

(29.) T. S. Chung, Polym. Eng. Sci., 26, 901 (1986).

(30.) F. P. La Mantia and A. Valenza, Polym. Eng. Sci., 29, 625 (1989).

(31.) F. P. La Mantia, A. Valenza, M. Paci, and P. L. Magagnini, Polym. Eng. Sci., 30, 7 (1990).

(32.) G. Crevecour and G. Groeninckx, Polym. Eng. Sci., 30, 532 (1990).

(33.) H. J. O'Donnell and D. G. Baird, Intern. Polym. Process., 11, 257 (1996).

(34.) Q. Lin. J. Jho, and A. F. Yee, Polym. Eng. Sci., 33, 789 (1993).

(35.) R. P. Brown, Handbook of Plastics Test Methods, 3rd Edition, The Plastics and Rubber Institute (1988).

(36.) I. Lee, B. Lee, and Y. Seo, to be submitted to Macromolecules.

(37.) S. S. Kim and C. D. Han, Polymer, 35, 93 (1994).

(38.) C. D. Han and J. K. Kim, Polymer, 34, 2533 (1993).

(39.) S. S. Kim and C. D. Han, Macromolecules, 26, 6633 (1993).

(40.) W. Ahn and K. H. Ha, Korea Polym. J., 7, 130 (1999).

(41.) D. J. Lee, T. J. Lee. and H. D Kim., Korea Polym. J., 7, 58 (1999).

(42.) Y. Seo and E. H. Wissler, Polym. Eng. Sci., 29, 722 (1989).

(43.) Y. Seo. H. J. Kim, B. Kim, S. M. Hong, S. S. Hwang, and K. U. Kim, Korea Polym. J., 9, 238 (2001).

YONGSOK SEO *

JINHO KIM +

* To whom correspondence should be addressed. E-mail: [e-mail]

+ Current address: Dept. of Material Science, University of Florida. FL 32611, U.S.A.