Formation and characterization of cast and biaxially stretched poly(butylene terephthalate) film.
Polybutylene terephthalate (PBT) has long history as an engineering thermoplastic. It was first synthesized by Whinfield (1) about 1949. For the next 20 years it was studied notably by ICI researchers and other companies particularly as a fiber. PBT was first introduced commercially to the marketplace as an injection molding resin about 1969 by Celanese Plastics in the U.S.A. It is still widely used as a molding resin. The structural characteristics of PBT have received considerable study (2-15). Two different crystallines generally designated as [Alpha] and [Beta] forms and a smectic glassy state have been identified. The most stable form is the [Alpha]-form crystalline structure.
There have been few, ff any, investigations of the formation of PBT into film, unlike polyethylene terephthalate (PET), which has been widely used in this manner. The only studies of the influence of processing on crystallinity and morphology of PBT have been by Spruiell and his co-workers (11, 12). In this paper, we present a study of the formation of film from PBT. We describe film casting onto a chilled roll for producing film. We also present a study of formation of biaxially oriented PBT film using a film stretcher. Special attention is given to the structural character and orientation of the film produced.
The polymer used in this study was a polybutylene terephthalate (PBT) obtained from BASF AG (Ultradur KR4036-Q692). It has an intrinsic viscosity (IV) of 1.24 in d1/mg, and a number-average molecular weight, [Mathematical Expression Omitted], of 42,340 is reported by the manufacturer. The PBT pellets were dried at 100 [degrees] C for 24 hrs in a laboratory vacuum oven before being fed into the extruder.
Cast Film. Cast films were prepared with a 25 mm Prodex single-screw extruder equipped with a 203mm-wide T-die and a take-up system, which had a chill roll with a temperature controller. The throughput of the extruder was fixed while the speed of the chilled roll was varied to set the draw ratio, i.e. [U.sub.L]/[U.sub.o] where [U.sub.o] is the film speed at the outlet of the die and [U.sub.L] on the chill roll. The chilled roll temperatures were fixed at 0 [degrees] C and 20 [degrees] C. The chilled roll was kept as close as possible to the die and two air fans were used as an air knife throughout the experiment. Extrusion temperatures were varied from 260 [degrees] C to 300 [degrees] C. A schematic diagram of cast film extrusion process is shown in Fig. 1.
Biaxial Film Stretching. The films prepared from the cast film process described above were variously biaxially stretched in an Iwamoto biaxial film stretcher. The undrawn films were cut to dimensions, 10 cm x 10 cm, and gripped by 40 pneumatic clips in the preheated chamber set at 90 [degrees] C. After 1 min of temperature equilibration, the films were stretched to the desired ratios at an initial stretch rate of 1200%/min. Films were either uniaxially stretched at constant width (UCW mode) or simultaneously biaxially stretched (SBS mode). Selected films were fixed annealed at 200 [degrees] C for 10 min using a square frame in a forced convection oven.
X-Ray Diffraction. Wide angle X-ray diffraction and the construction of pole figures were used to characterize the films. WAXS diffraction patterns were prepared with a Rigaku rotating anode generator equipped with a horizontal diffractometer. A CuK[Alpha] radiation with a nickel foil filter was used as the X-ray source. The specimens were prepared by stacking the films carefully along the MD and TD to dimensions of 20 mm x 20 mm x 1.2 mm. AGE copper target X-ray generator with a quarter circle goniometer was used to construct the pole figures. The specimen had dimensions of 1.2 mm x 15 mm x 1.2 mm. The diffraction intensifies for the 100 and 010 planes were collected around two Eulerian angles at intervals of [Chi] of 5 [degrees] and [Phi] of 10 [degrees] .
WAXS measurements were used to determine White and Spruiell biaxial orientation factors (16) defined as:
[Mathematical Expression Omitted] (1.a)
[Mathematical Expression Omitted] (1.b)
where [[Phi].sub.t,j] is the angle between film direction i and crystallographic axis j.
The triclinic unit cell of Desborough and Hall (8) was used to interpret diffraction data. In order to calculate biaxial orientation factors, the triclinic unit cell was approximated as "pseudo-orthorhombic," similar to the PET unit cell of Yoshihara et al. (17). In this pseudo-orthorhombic unit cell, the c axis is along the chain axis, the a axis is parallel to the phenyl ring, and the b axis is taken to be orthogonal to the a-c plane.
Refractive Indices. Refractive indices in the machine, transverse, and thickness direction of the PBT films were prepared using a Bellingham Stanley Abbe 60/HR refractometer with a polarizing eyepiece. For improved resolution a sodium lamp of wavelength 589.6 nm was used as a light source. The specimens were prepared by cutting the films carefully along the MD to dimensions of 40 mm x 40 mm. The immersion liquid was diiodomethane sulfur ([n.sub.D] = 1.75). This is a technique developed by Okajima and his colleagues (18-20) and later used by Samuels (21) and Cakmak et al. (22).
Thermal Analysis. Differential scanning thermal studies were carried out over the temperature range -10 [degrees] C to 300 [degrees] C with a DuPont DSC 9900 instrument at a heating rate of 20 [degrees] C/min. The sample weight used in the DSC experiments was 8.0 [+ or -] 0.2 mg. The instrument was calibrated using pure indium metal. Crystallinity was calculated through
[X.sub.c] = [Delta][H.sub.exp]/[Delta]H [degrees] (2)
where [Delta][H.sub.exp] = [Delta][H.sub.melting] - [Delta][H.sub.cold crystallization], [Delta]H [degrees] = 142 J/g given by Iller (23) as the heat of fusion of 100% crystalline PBT.
Cast Film. The cast film process could be operated in a stable manner without any difficulty. We observed optical clarity of films and found them to be transparent or hazy. It was observed that a PBT melt extruded at 260 [degrees] C into air was transparent just after it emerged from the die. After seconds in air, it suddenly turned hazy. When a melt, extruded at 260 [degrees] C, was instead cast onto a chilled roll, the films became hazy with cooling. Optically non-homogeneous regions, i.e. regions with varying amount of light scattering, were observed in the films. When the melt temperature was raised to 300 [degrees] C this non-homogeneity disappeared.
Cooling of cast films emerging from the coathanger die was primarily controlled by two processing parameters - temperature and thickness of films. It was found that the lower the temperature of a chilled roll, the more optically clear the films. Films were not uniformly transparent without use of an air knife. In order to increase the cooling rate of the film, the thickness of the film was changed from 300 to 100 [[micro]meter]. The temperature of the chilled roll was fixed at 0 [degrees] C.
It was observed that film thickness was a dominate factor in determining the optical clarity of the films produced. A decrease in thickness increased the transparency of resultant films and decreased the fluctuation of thickness. At film thicknesses above 150 [[micro]meter], the films were hazy and distorted in appearance, and fluctuated in thickness with a standard deviation of 20 to 30%. Some films exhibited local haze. Uniformly transparent films were produced at film thicknesses below 120 [[micro]meter] in this experiment. These films were quite uniform in thickness with standard deviation of around 10%.
Another observation worth noting was that the molten films on the chilled roll tended to slip away from the roll surface during film casting. This tendency became less with decreasing film thickness.
Biaxial Film Stretching. Cast films were stretched in the biaxial stretcher to study the extension behavior of the films. In the UCW mode the films were stretched uniformly without necking up to high stretch ratio ([[Lambda].sub.MD] = 4.5). The resultant films were uniform in thickness and fiat in surface appearance.
It was observed that the transparent and hazy films behaved in a different manner during simultaneous biaxial stretching. The transparent films stretched uniformly without necking up to maximum stretch ratios of [[Lambda].sub.MD] x [[Lambda].sub.TD] = 4 x 3; however, the hazy films frequently broke or necked during stretching. When the transparent films were stretched above [[Lambda].sub.MD] x [[Lambda].sub.TD] = 4 x 3, the films would tear.
Pre-heating time also affected necking. We could not avoid necking occurring along the clips of the biaxial stretcher unless the undrawn films were pre-heated as rapidly within 60 seconds in a chamber set at 90 [degrees] C before initiation of stretching.
Wide Angle X-ray Diffraction
Cast Film. We examined the films prepared under a variety of conditions to determine their crystalline character and orientation. Figure 2 presents WAXS meridional and equatorial scans for cast films measured by a conventional [Theta]/2[Theta] goniometer. It was found that transparent films showed an amorphous halo. Hazy films exhibited crystalline reflections at d spacings of 9.83, 5.14, 4.41, and 3.83 [Angstrom]. These are equivalent to the PBT 001, 010, [Mathematical Expression Omitted], and 100 planes of the [Alpha] phase. The diffraction intensity is independent of the scanned direction. No orientation is seen. [Beta]-form crystals were not observed in cast film products.
Biaxially Stretched Cast Films. Figure 3 presents the results of biaxially stretched films. Uniaxially stretched films exhibit a high degree of anisotropy in diffracted peaks. In meridional scans, reflections were observed at d spacings of 9.83, 2.86, and 2.29 [Angstrom], which are equivalent to the PBT 001, [Mathematical Expression Omitted], and [Mathematical Expression Omitted] planes of the a phase. Equatorial reflections on the meridian merge into a single broad peak with increasing [[Lambda].sub.MD]. The equatorial scans show reflections at 5.14 and 3.83 [Angstrom], which are equivalent to the PBT 010 and 100[Alpha] reflections. An increase in [[Lambda].sub.MD] causes an increase in intensity for both off-meridional and equatorial reflections. Strikingly, it was found that the [Beta] phase of PBT is present in uniaxially stretched films. There is a strong reflection at 2.11 [Angstrom] equivalent to a [Mathematical Expression Omitted] reflection and a detectable reflection at 3.19 [Angstrom] equivalent to a [Mathematical Expression Omitted] reflection.
The behavior of diffracted peaks in biaxially stretched films is different. For [[Lambda].sub.MD] above 3, an increase in [[Lambda].sub.TD] causes a decrease in intensity for all reflections both on the meridian and on the equator; however, for [[Lambda].sub.MD] = 2, increasing [[Lambda].sub.TD] increases the intensity of equatorial reflections, especially the 010 plane, indicating an increase in crystal perfection. At equal biaxial conditions the diffracted peaks exhibit transverse isotropy. We observed [Beta] reflections in the biaxially stretched films as well.
Pole figures have been determined for various crystalline reflections. We will present these in the Discussion section.
Cast Film. The transparent cast films exhibited a refractive index of about 1.57012, and the hazy cast films about 1.57557 for all three principal directions. This indicates that both films are optically isotropic and the hazy films have more crystallinity.
Biaxially Stretched Cast Film. Figure 4a shows the refractive indices of biaxially stretched films. The values next to each symbol represent planar strain, [[Lambda].sub.MD] x [[Lambda].sub.TD]. An increase in [[Lambda].sub.MD] for a given [[Lambda].sub.TD] causes an increase in refractive index, [n.sub.MD] in the machine direction (MD), while [n.sub.TD] in transverse direction (TD) and [n.sub.ND] in the normal direction (ND) decreased. Transverse isotropy does not exist in the uniaxial constant width films.
Increasing biaxial stretch [[Lambda].sub.TD] for a given [[Lambda].sub.MD] causes an increase in [n.sub.TD] and a decrease in [n.sub.MD] and [n.sub.ND]. Equal biaxial films exhibit in-plane isotropy, i.e. [n.sub.MD] [approximately equal to] [n.sub.TD]. For all stretch conditions [n.sub.ND] decreases with increasing planar strain.
Birefringences calculated from the refractive index data show similar character to that of the primary refractive indices. Generally, the birefringence, [Delta][n.sub.13] at [[Lambda].sub.MD] above 3 decreases with increasing [[Lambda].sub.TD]. However, at [[Lambda].sub.MD] = 2 [Delta][n.sub.13] increases with increasing [[Lambda].sub.TD]. For a given [[Lambda].sub.MD], [Delta][n.sub.23] increases rapidly with increasing [[Lambda].sub.TD]. It was found in our experiments that the maximum value of [Delta][n.sub.13] is about 0.1788.
Annealed Film. The refractive indices of annealed films are presented in Fig. 4b. The general behavior is similar to that of unannealed precursors. Annealing causes a further increase in [n.sub.MD] and [n.sub.TD], while it decreases [n.sub.ND] slightly. In consequence, mean refractive indices increase considerably with annealing indicating an increase in density of the films. Increased mean refractive index implies increased crystallinity.
Birefringences move to higher values with annealing. It is especially seen that at [L.sub.MD] = 2 birefringences increase sharply as compared to those of unannealed precursors. The maximum value for [Delta][n.sub.13] was found to be about 0.1885 and for [Delta][n.sub.23] about 0.1589.
Differential Scanning Calorimetry
Cast Film. DSC thermograms of cast films are presented in Fig. 5a. We observed a [T.sub.g] around 45 [degrees] C, a cold crystallization exotherm around 54 [degrees] C, and a melting endotherm around 222 [degrees] C. The transparent films exhibited a crystallization exotherm just after the glass transition region, whereas the hazy films showed only a large inflection at [T.sub.g], which occurs at 5 [degrees] C higher than that of the transparent films. Both types of the films showed large heats of fusion, indicating that the films crystallized during an upward temperature scan in DSC. It shows little difference in the features of melting endotherms between both films.
Biaxially Stretched Cast Film. DSC thermograms of biaxially stretched films are presented in Fig. 5a. For uniaxial conditions, the inflection of [T.sub.g] broadened gradually with increasing [[Lambda].sub.MD] and the crystallization exotherm disappeared. It is seen that an increase in [[Lambda].sub.MD] causes an increase in magnitude of the melting endotherm and a decrease in broadness, whereas the position of endothermic maxima changes little with [[Lambda].sub.MD].
It shows that biaxiality causes considerable changes in the feature of DSC thermograms. Increasing biaxiality causes an increase in broadness of both glass transition region and melting endotherm, especially at equal biaxial conditions. The melting peaks changed little in position with biaxial deformation.
Annealed Film. DSC thermograms of annealed biaxially stretched films are presented in Fig. 5b. It was found that annealed uniaxially stretched films exhibited only one melting peak around 224 [degrees] C, similar to the position to that of the unannealed precursors, while annealed biaxially stretched films showed two distinct melting peaks in which the lower melting peak increases in magnitude with increasing biaxiality. Generally, the higher melting endotherm behaved similarly to that of unannealed precursors at the same stretch ratios. It was observed that annealed undrawn films ([[Lambda].sub.MD] x [[Lambda].sub.MD] = 1 x 1) exhibited the largest lower melting peak among the films.
PBT Processing Behavior
Cast Film Extrusion. We observed in our experiments that cast films became more optically homogeneous with increasing melt temperature. This was possibly due to the incomplete melting of the polymer at the lower melt processing temperature. The polymer used in this experiment entered the extruder as a crystalline solid with 41% crystallinity.
The PBT melt crystallized rapidly with cooling on the chilled roll. The higher rate of cooling resulted in a more optically transparent film. One would expect that good heat transfer is essential to suppress crystallization of a molten film on the chilled roll. Heat transfer in a cast film process is governed by processing factors - contact between a molten film and a chilled roll, degree of supercooling, and heat conduction through the thickness direction of films. We achieved good contact of the film with the chilled roll by using an air knife and high supercooling by decreasing the temperature of the chilled roll. Heat conduction through the film is associated with the heat conductivity of a polymer. Since PBT possesses low heat conductivity, film thickness tends to dominate heat conduction through a film, hence its lower extent of crystallization. The thinner the film, the lower the refractive index and the less crystallinity. These films were transparent.
Biaxial Stretching Process. Transparent cast films were successfully stretched both uniaxially and biaxially. Those films with a standard deviation of 10% or less in thickness exhibited typical stress strain behavior during stretching. In non-uniform films, stress concentrations would occur in the thin region and initiate necking. Since the necked region experienced strain-induced crystallization and hardening, deformation propagated rapidly to neighboring regions until the whole film deformed uniformly.
The hazy cast film exhibited higher refractive index, hence higher crystallinity, than the transparent cast film. This difference in crystalline character between both films may cause different extension behavior of the films in biaxial stretching processes. It would seem that the nonhomogeneities in hazy films act as defects, giving rise to frequent breakage during stretching.
Crystalline Character of PBT as a Function of Processing
Wide Angle X-Ray Scattering. Our studies of as-cast and biaxially stretched films produced under different conditions showed different peaks in WAXS diffractometer scans suggesting both incomplete crystallization and polymorphism. Transparent cast films show an amorphous halo. Hazy cast films reveal various reflections of the [Alpha]-form crystalline even though the level of crystallinity might be low. Owing to the crystal imperfection, the diffracted peaks of the hazy film are somewhat overlapped and not defined very well. This is especially seen in reflections at 2[Theta] above 30 [degrees]. [Beta]form crystallites are not found. It is believed that the transparent cast films are largely in a glassy state, but may have some smectic ordering to explain its melting behavior. The hazy cast films may have a densely nucleated miscrostructure of poor crystals.
Off-meridional reflections of the [Beta] phase are seen in biaxially stretched films. It would appear that high extension in film processes favors a more stable [Beta] phase being retained in the resultant films. An increase of [Mathematical Expression Omitted] reflection in intensity with MD extension indicates an increase in the c-axis orientation of the [Beta] phase.
The equatorial reflections of 010 and 100 were used to characterize the crystalline orientation of the [Alpha] phase for pole figures. These planes of preferred orientation both have the strongest intensity on the equator.
In order to clarify the crystalline behavior of oriented films, WAXS film patterns and scan were taken by tilting the specimen along the azimuthal direction with the ND parallel to the X-ray beam. The results are shown in Fig. 6. The flat film photographs reveal that lateral packing of crystallites is not highly ordered; however, quite well-defined Bragg maxima of the 010 and 011 planes both on and off the equator indicate the presence of three-dimensional order.
It is seen that the diffracted peaks of unannealed biaxially stretched films are not sharp enough for accurate indexing of lattice parameters. Uniaxially stretched films (4.5 x 1) annealed under tension at 150 [degrees] C for 8 hours were therefore prepared to investigate Bragg spacings and the anisotropic behavior of diffracted peaks. Figure 7 presents tridirectional WAXS fiat film photographs and scan patterns with tilting of the specimen. The anisotropic behavior of two preferred reflections close to the equator are consistent with those in the unannealed specimen. The lattice parameters observed in this experiment are given in Table 1. These are in good agreement with the triclinic unit cell of Desbrough and Hall (8). It is of interest to see that the tilting method can give quite accurate locations of Bragg angles of various crystalline reflections in WAXS 20 scan.
[TABULAR DATA FOR TABLE 1 OMITTED]
The displacement of [Mathematical Expression Omitted] reflection close to the meridian is dependent on the chain conformation of the glycol residue (9). The length of c axis was determined as 11.65 [Angstrom] by measuring the spacing of this reflection, indicating that the tetramethylene segment is mostly in the [Alpha] modification. The length of the c axis repeat for the [Beta] modification was 13.0 [Angstrom] (7). The preferred plane on the third layer close to the equator in the end pattern establishes that this plane of preferred orientation is roughly parallel to the plane of film surface. In the case of PET the preferred plane is essentially parallel to the plane of phenyl ring (24, 25). It is seen that the [Mathematical Expression Omitted], [Mathematical Expression Omitted], and [Mathematical Expression Omitted] planes of the [Alpha] phase are within 10 [degrees] of the MD. The [Mathematical Expression Omitted] reflection is the strongest in intensity among them on the meridian in the edge scan pattern. These longitudinal reflections might be used to approximate the c-axis orientation of [Alpha] phase. The 001[Alpha] plane is, however, quite far from the meridian. The scattered intensity of 00 1[Alpha] reflection is found to be strong at tilting angles of around 34 [degrees].
Differential Scanning Calorimeter. DSC results show that transparent and hazy cast films exhibited roughly the same crystallinity, being around 23%, but different features in the DSC thermograms. This may imply that the transparent films possess poorly defined order, and the hazy films have a microstructure of poor crystals. This observation is well consistent with the results of WAXS diffractometer scans and refractive indices in the previous section.
We have calculated crystallinities of the films using Eq 2. The results are in Fig. 8. For [[Lambda].sub.TD] = 1, i.e. uniaxial conditions, the crystallinity increases with increasing [[Lambda].sub.MD], achieving a maximum value around 47%. At [[Lambda].sub.MD] above 3 increasing biaxiality causes a decrease in crystallinity and thus disrupts the packing efficiency of the chains. Cakmak et al. (26) have suggested that in biaxial stretching conditions, the population of chains in the minor anisotropy axis acts as a disruption to the crystallization. At [[Lambda].sub.MD] = 2 the crystallinity increased with [[Lambda].sub.TD], possibly owing to the low levels of deformation stress. At these stretching conditions, the thermal effect may have a considerable influence on the crystallinity development of the film. This can be reconciled with an increase of crystal perfection with [[Lambda].sub.TD] in WAXS diffractometer scans.
The general behavior of annealed films was similar to that of unannealed precursors. Annealing increased crystallinity marginally up to 3%. This strongly suggests that there exists an exothermic restructuring process during the course of DSC scans.
Refractive Index. The refractive index is a measure of the velocity of light in the medium. According to arguments that date to Lorentz, refractive index depends upon the local density. The velocity of light in the crystalline region will be different from that in amorphous region. Thus, the refractive index can be related to the crystallinity of that material. For a two-phase system of crystalline and noncrystalline polymer (21),
[Mathematical Expression Omitted] (3)
where [Mathematical Expression Omitted] is the arithmetic average of refractive indices; the subscripts 1,2,3 are three principle directions; ne and [n.sub.AM] are the characteristic refractive indices of the crystal and the amorphous polymer, respectively; and [V.sub.c] is the volume fraction of crystals.
We correlated the average refractive index of the annealed biaxially stretched films with the percent crystallinity determined from DSC measurements. Annealed films primarily possessed the [Alpha]-phase crystallites. Figure 9 shows a linear relationship between refractive index and the crystallinity of annealed films. The equation for this line is,
[Mathematical Expression Omitted] (4)
We have used this relationship and calculated the crystallinity of our cast and annealed biaxially stretched films from their mean refractive indices. This indicates that the crystallinity of transparent cast films is about 3.5% and hazy cast films about 14%. The crystallinity of biaxially stretched films ranges from 20% to 30%. Annealed films exhibit crystallinities in the range of 30% to 50%, which are significantly higher than those of unannealed precursors. These values may better represent the crystallinities of initial specimens before thermal analysis.
It is possible that there exists a considerable quantity of a stable [Beta] phase in unannealed biaxially stretched films since the density of the [Beta]-form crystal ([[Rho].sub.c] = 1.2830 g/[cm.sup.3]), as opposed to [[Rho].sub.c] = 1.404 g/[cm.sup.3] for the [Alpha]-form crystal, is known to be close to that of the amorphous phase ([[Rho].sub.a] = 1.2812 g/[cm.sup.3]) (6, 11). Lu and Spruiell (11) reported that the content of the [Beta] phase for fibers drawn at 90 [degrees] C was in the range of 40% to 70%.
Finally, it is noteworthy that unannealed films showed up to 15% higher crystallinity in DSC measurements than in [Mathematical Expression Omitted] measurements. Muramatsu and Lando (27) made similar observations, though not explicitly stated, that an unannealed injection molded PBT specimen revealed 10% higher crystallinity in DSC measurements than in density measurements. Apparently, the temperature scans in DSC give rise to a significant increase of crystallinity in unannealed films.
Characterization of Orientation in PBT Films
Refractive Index. Refractive indices and hence birefringence measure the total molecular orientation of a polymer (21). We may relate birefringence to a biaxial orientation factor of the system. If we use a two phase model of crystalline and amorphous regions, the birefringence may be written:
[Mathematical Expression Omitted] (5)
[Mathematical Expression Omitted] (6)
where X is the fractional crystallinity; [Mathematical Expression Omitted] and [Mathematical Expression Omitted] are intrinsic birefringence corresponding to the crystallographic cb and ab axes; and [Mathematical Expression Omitted] is the intrinsic birefringence of the amorphous phase. Obviously, birefringence increases with increasing orientation.
Ohkoshi and Nagura (28) have calculated refractive indices and intrinsic birefringence of PBT using the triclinic unit cell of Hall and Pass (7). The values of n and [[Delta].sup.0] for the [Alpha] phase were given as 1.625 and 0.153, and as 1.622 and 0.215 for the [Beta] phase, respectively. In our experiments, the maximum value of [Delta][n.sub.13] for unannealed biaxial films was found to be about 0.1788, and for annealed films about 0.1885. It is noteworthy that unannealed films exhibited a mixed character of both the [Alpha]- and the [Beta]-form crystallites, while annealed films showed primarily a character of the [Alpha]-form crystallites. The maximum value of [Delta][n.sub.12] is reported as 0.1550 for high speed melt spun fibers (12) and as 0.1830 for an annealed fiber (4), both having the [Alpha]form crystal. There clearly exist large differences in intrinsic birefringence of the [Alpha]-form crystallites between calculated and measured values. Ohkoshi and Nagura (28) have suggested that the difference may come from the contribution of form birefringence and trans conformations in the amorphous phase. Experimentally, it was found that the average refractive index for complete amorphous phase was 1.56832 and for complete crystalline 1.61916. The average refractive index [Mathematical Expression Omitted] for transparent cast films was measured as 1.57012, which is fairly close to n for the amorphous phase.
Pole Figures. We have characterized WAXS pole figures in order to better understand the anisotropic behavior of the films. In the triclinic unit cell of PBT, the off-meridional planes of [Mathematical Expression Omitted], [Mathematical Expression Omitted] and [Mathematical Expression Omitted] are roughly perpendicular to the chain axis. The 100[Alpha] plane is roughly parallel to the plane of the phenyl ring on the chain. The normal to the 010[Alpha] plane makes an angle of 50.1 [degrees] with the chain axis and 29.0 [degrees] with the 100[Alpha] plane.
WAXS pole figures for different lattice planes of biaxially stretched films are presented in Fig. 10. At low [[Lambda].sub.MD] of 2, the poles of the 100[Alpha] plane are primarily concentrated in the ND with some spread toward the TD, while the poles of the 010[Alpha] and off-meridional reflections are distributed in the film plane. The phenyl rings become aligned parallel to the film plane even at [[Lambda].sub.MD] = 2. For [[Lambda].sub.MD] above 3, increasing biaxiality makes the poles of the 100[Alpha] reflection concentrate in the ND, especially for equal biaxial conditions, while the 010[Alpha] poles spread in the MD-TD plane. The poles of off-meridional reflections tend to concentrate primarily in the MD with uniaxial deformation. Increasing biaxiality spread them in the plane of the film, indicating that the chains become oriented in the film plane. It is especially seen that the poles of the [Mathematical Expression Omitted] reflection concentrate primarily both in the MD and in the MD-ND plane with uniaxial deformation. An increase in [[Lambda].sub.MD] for a given [[Lambda].sub.TD] causes an increase in pole intensities for all specimens. Clearly, the pole figures are well consistent with the anisotropic behavior of diffracted peaks in WAXS scattering patterns.
Biaxial Orientation Factors. White-Spruiell biaxial orientation factors were calculated for the [Alpha]-crystalline phase for biaxially stretched films. The results are plotted in Fig. 11. Uniaxially stretched films exhibit characteristics of biaxial orientation. As uniaxial deformation increases, the chains align rapidly to the MD and the phenyl rings parallel to the film surface. The TD orientation [Mathematical Expression Omitted] also increases slightly with increasing uniaxial deformation. Increasing biaxiality causes an increase in population of TD oriented crystals, hence a decrease in orientation of c axis, while it increases quickly the alignment of phenyl rings parallel to the film plane. The planarity of the film increases with increasing [[Lambda].sub.TD]. At equal biaxial conditions [Mathematical Expression Omitted] and [Mathematical Expression Omitted] are the same level. The orientation behavior of the 010[Alpha] reflection is quite different from those of the 100[Alpha] plane and the c axis. The MD orientation [Mathematical Expression Omitted] of the 010[Alpha] reflection increases with increasing biaxiality. This is consistent with the behavior of the 010[Alpha] plane in the pole figures.
PBT melts are found to be readily crystallizable in the film casting process: amorphous films can be produced only when rapid quenching is used. Such cast films may be satisfactorily stretched in a biaxial film stretcher. Slow cooling in this process produced semicrystalline films with only [Alpha]-form crystallities and 14% crystallinity.
Biaxial stretching of cast films perfects the crystalline structure of the films. Biaxially stretched films show polymorphism of both the [Alpha]- and [Beta]-form crystallites with crystallinities of 20%-30%. High extension in this process favors a more stable [Beta] phase being retained in the resultant films. Annealing causes an increase in perfection of crystalline structure and a decrease in amount of the [Beta] phase in the films.
Uniaxial deformation causes polymer chains to align to the MD, while biaxial stretching tends to shift them to the TD. The phenyl rings of the chain backbone are increasingly parallel to the plane of the film with simultaneous biaxial stretching. Annealing further increases the orientation in the films.
1. U.S. Patent 2,465,319 (1949), J. R. Whinfield and J. T. Dickson.
2. A. M. Joly, G. Nemoz, A. Douillard, and G. Vallet, Makromol. Chem, 178, 479 (1975).
3. Z. Mencik, J. Polym Sci. Polym, Phys., 13, 2171 (1975).
4. R. Jakeways, I. M. Ward, and M. A. Wilding, J. Polym. Sci. B, 13, 799 (1975).
5. I. M. Ward and M. A. Wilding, J. Polym. Sci. Polym. Phys., 14, 263 (1976).
6. M. Yokouchi, Y. Sakakibara, Y. Chatani, H. Tadokoro, T. Tanaka, and K. Yoda, Macromolecules, 8, 226 (1976).
7. I. H. Hall and M. G. Pass, Polymer, 17, 807 (1976).
8. I. J. Desborough and I. H. Hall, Polymer, 18, 825 (1977).
9. B. Stambaugh, J. L. Koenig, and J. B. Lando, J. Polym. Sct : Polym. Phys., 17, 1053 (1979).
10. R. S. Stein and A. Misra, J. Polym. Sci., 18, 327 (1980).
11. F. M. Lu and J. E. Spruiell, J. Appl. Polym. Sci., 31, 1595 (1986).
12. S. Chen and J. E. Spruiell, J. Appl. Polym. Sci., 33, 1427 (1987).
13. H. J. Ludwig and P. Eyeper, Polym. Eng. Sci., 28, 143 (1988).
14. B. Moginger, C. Lutx, A. Polsak, and U. Fritz, Kunststoffe German Plastics, 81, 3 (1991).
15. J. Roebuck, R. Jakeways, and I. M. Ward, Polymer, 33, 227 (1992).
16. J. L. White and J. E. Spruiell, Polym. Eng. Sci., 20, 247 (1980); Polym. Eng. Sci., 23, 247 (1983).
17. N. Yoshihara, A. Fukushima, Y. Watanabe, and A. Nakai, et al., Sen-i Gakkaishi, 37, T-387 (1981).
18. S. Okajima and Y. Koizumi, Kogyo Kogaku Zashi, 42, 810 (1939).
19. H. Tanaka, T. Masuko, and S. Okajima, J. Polym. Sci., 1, A-1, 3351 (1969).
20. H. Tanaka, T. Masuko, and S. Okajima, J. Appl. Polym. Sci., 19, 441 (1972).
21. R. J. Samuels, J. Appl. Polym. Sci., 26, 1383 (1981).
22. M. Cakmak, J. L. White, and J. E. Sprutell, Polym. Eng. Sci., 29, 1534 (1989).
23. K. H. Iller, Colloid Polym. Sci., 258, 117 (1984).
24. W. J. Dulmage and A. L. Geddes, J. Polym. Sci., 31, 499 (1958).
25. H. Tadokoro, K. Tatsuka, and S. Murashi, J. Polym. Sci., 58, 413 (1962).
26. M. Cakmak, J. L. White, and J. E. Spruiell, J. Polym. Eng., 6, 291 (1986).
27. S. Muramatsu and J. B. Lando, Polym. Eng. Sci., 35, 1077 (1995).
28. Y. Ohkoshi and M. Nagura, Sen-i Gakkaishi, 49, 11, 601 (1993).
|Printer friendly Cite/link Email Feedback|
|Author:||Song, Kwangjin; White, James L.|
|Publication:||Polymer Engineering and Science|
|Date:||Mar 1, 1998|
|Previous Article:||Thermal expansion coefficients.|
|Next Article:||The influence of nucleating agents on the extrusion and thermoforming of polypropylene.|