Fiber structure development in high-speed melt spinning of copolyesters: poly(ethylene terephthalate-co- 1,4-cyclohexylene dimethylene terephthalate).
Poly(1,4-cyclohexylene dimethylene terephthalate) (PCT) is a polyester prepared by replacing the ethylene glycol group of poly(ethylene terephthalate) (PET) with 1,4-cyclohexylene dimethanol group. PCT is a highly heat-resistant crystalline polymer having a melting temperature 30 [degrees] C higher than PET (1). Copolyesters of PET and PCT show wide range of crystallizability depending on their copolymer compositions, and an amorphous copolyester, whose crystallization rate is practically zero, can be obtained at a certain composition (2-4). The copolyester of this composition has been widely applied to the commercial production because of its high processibility in various processing technology.
We have previously reported the copolymer composition and sequence distribution (5), and the co-crystallization behavior and melting temperature depression (6) of these copolyesters. From these studies, fundamental characteristics of the polymers in an isotropic state was revealed.
On the other hand, it is widely known that in high-speed melt spinning of PET, crystallization starts to occur above a certain take-up velocity despite rapid quenching (7). This is due to the considerable increase in crystallization rate with the development of molecular orientation. In this study, PET, PCT, and a series of P(ET/CT) copolyesters were subjected to the high-speed melt spinning process. Through the analysis of the structure and properties of as-spun fibers, the changes in the molecular orientation and crystallization behavior of the polymers with copolymer composition were investigated.
Materials and Melt Spinning
Six commercial copolyesters (Eastman Chemical Products Inc.) were used. The chemical structure of these copolyesters are considered to be as follows:
[Mathematical Expression Omitted]
The copolymer composition (in mol%) of the samples analyzed by 1H NMR (5) are listed in Table 1 with their intrinsic viscosity values. The 1H NMR analysis also revealed that the 1,4-cyclohexylene dimethanol group of all the samples is composed of [approximately] 68% of trans- and [approximately] 32% of cis-isomers. The sequence distributions of the copolyesters determined by 13C NMR spectra were statistically random.
Prior to the melt spinning, the polymer pellets were dried under vacuum for 48 h at 100 [degrees] C and then for 24 h at 150 [degrees] C except for the one with copolymer composition ET/CT of 66/34. Since this polymer is not crystallizable, the polymer pellets would stick together if the temperature is raised above its glass transition temperature. Thus the polymer was dried at 65 [degrees] C for 72 h. Although the water content under this drying condition is higher than that for other polymers, molecular weight reduction in the melt-processing is reported to be less significant since lower processing temperature can be chosen (2). Melt spinning conditions are also shown in Table 1. The take-up velocity was increased by 0.5 km/min up to the attainable highest velocity. All the polymers had fairly good spinnability and could be spun up to the take-up velocity of 6.0 km/min or higher.
Wide-Angle X-Ray Diffraction
Wide-angle X-ray diffraction (WAXD) intensity of the aligned as-spun fibers was measured along the equator using a nickel filtered CuK [Alpha] radiation source (Rigaku, Rad-B).
Refractive Index and Birefringence
Refractive indices parallel and perpendicular to the fiber axis ([n.sub.[parallel]] and [n.sub.[perpendicular]]) were measured using an interference microscope (Carl Zeiss Jena) with a polarizing filter. Mixtures of tricresyl phosphate (refractive index n = 1.558), trans-phenyl-acrylaldehyde (n = 1.620) and [Alpha]-bromonaphthalene (n = 1.658) were used as the immersion liquids. Birefringence [Delta] n was calculated as [Delta] n = [n.sub.[parallel]] - [n.sub.[perpendicular]].
Density of as-spun fibers was measured at 25 [degrees] C in a density gradient column of sodium bromide (NaBr)/water system. Bundles of filaments were degassed [TABULAR DATA FOR TABLE 1 OMITTED] in a NaBr aqueous solution before putting into the column, and the density was determined from the position of the samples in the column 24 h after the immersion.
Stress-extension curves of as-spun single filaments were obtained using a tensile testing machine (TOYO SOKKI, TENSILON UTM-4L). The specimen length and elongation rate were 20 mm and 100%/min, respectively. Ten samples were measured for each spinning condition. Initial Young's modulus, elongation at break and tensile strength were obtained from the stress-extension curves.
Effect of annealing on the structure of fibers prepared from the copolyester of ET/CT = 66/34 was investigated. The fibers melt-spun at 1 km/min was drawn at 70 [degrees] C to a draw ratio of 3. The as-spun filaments of various take-up velocities and the drawn filaments were heat-treated in a heating chamber (METTLER, FP90) under a polarizing microscope (OLYMPUS, BHS-P). The filament length was fixed and the heating rate was 1 [degree] C/min. The observed image was recorded by a video cassette recorder and the change of birefringence with temperature was analyzed from the pattern of interference fringes.
RESULTS AND DISCUSSION
Wide-Angle X-Ray Diffraction
WAXD intensity along the equator for high-speed spun fibers of six polyesters are shown in Fig. 1. With an increase in the take-up velocity, an amorphous peak around the Bragg angle of 42 [degrees] became smaller in all the polymers. Since this peak can be assigned to the scattering from the intra-molecular interference, the reduction of peak intensity on the equator can be interpreted as a result of molecular orientation.
Except for the fibers of ET/CT = 66/34, crystalline reflections started to appear above a certain take-up velocity for each copolymer composition. In cases of the polymers of ET/CT = 100/0 and 95/5, crystallization was observed above 4.5 km/min, and three main peaks observed were assigned to (010), [Mathematical Expression Omitted], and (100) reflections of PET crystals (8). On the other hand, crystallization of fibers spun from ET/CT = 0/100, 20/80, and 34/66 copolymers was detected above 4.5, 5.0, and 6.0 km/min, respectively, and crystalline peaks were assigned to (010), [Mathematical Expression Omitted], and (100) of PCT crystals (9).
From these results, it was confirmed that the major copolymer component tended to crystallize preferentially under molecular orientation as in the isotropic state (6). With the addition of ET component to PCT, the onset of crystallization obviously shifted to higher take-up velocities. We could think of two possible reasons for the ET/CT = 66/34 polymer remaining amorphous; 1) the polymer cannot crystallize even under the oriented state, and 2) the attained take-up velocity was simply not high enough for the onset of orientation-induced crystallization. At the moment, however, these two reasons are indistinguishable.
Refractive Index, Birefringence, and Density
Refractive indices parallel and perpendicular to the fiber axis ([n.sub.[parallel]] and [n.sub.[perpendicular]]) measured using an interference microscope, and birefringence calculated from these two refractive indices were plotted against the take-up velocity in Fig. 2. It is obvious from the result of nearly isotropic 0.5 km/min spun fibers that the mean refractive index of the polymers in an amorphous state decreased with an increase in the amount of CT component. With increasing take-up velocity, [n.sub.[parallel]] increased with the take-up velocity, whereas [n.sub.[perpendicular]] decreased. Consequently, birefringence increased with the take-up velocity. Birefringence of all the polymers tended to saturate or reach a maximum at [approximately] 6.0 to 6.5 km/min. The maximum birefringence was lower for the polymers with higher CT composition.
Dependence of density of as-spun fibers on the take-up velocity is shown in Fig. 3. On the whole, the density decreased with an increase in the CT composition. The decreases in the mean refractive index and density with the addition of CT component might be attributable to the bulkiness of the CT component which leads to the coarse packing of copolymer molecules. Density for ET/CT = 66/34 fibers remained almost constant with the change in the take-up velocity, whereas other polymers registered an increase at take-up velocities where the occurrence of orientation-induced crystallization was confirmed in the WAXD measurement. In cases of ET/CT = 34/66, 20/80, and 0/100, however, the increase in density started at take-up velocities 0.5 to 1.0 km/min lower than those where the crystalline reflections were detected in the WAXD measurements. Above 6.5 km/min, density saturated or slightly decreased with the take-up velocity as in the case of birefringence.
The Lorentz-Lorenz equation expresses the relation between refractive index n and density [Rho] of the filament as follows.
[n.sup.2] - 1/[n.sup.2] + 2 = 4[Pi]N[Rho]/3M P (1)
where N, M, and P denote the Avogadro Number, molecular weight, and polarizability, respectively. For a uniaxially anisotropic material such as filaments, the mean refractive index n may be expressed using the refractive indices parallel and perpendicular to the principal axis ([n.sub.[parallel]], [n.sub.[perpendicular]]) as follows (10).
[Mathematical Expression Omitted]
According to Eqs 1 and 2, the values ([n.sup.2] - 1)/([n.sup.2] + 2) calculated using the experimental data were plotted against the density as shown in Fig. 4. The slope of straight lines drawn through the origin decreased with an increase in the CT composition. Specific refractivity (= 4[Pi]N[Rho]/3M) and polarizability of the polymers calculated from the slopes were tabulated in Table 2. For the evaluation of the polarizability, molecular weight of the copolymers were calculated from the mole fraction and molecular weight of ET and CT components.
Table 2. Experimental and Calculated Molar Polarizabilities for PET, PCT and P(ET/CT) Copolyesters.
Composition Molar Polarizability x [10.sup.31] ET/CT ([m.sup.3]) (mol%) experimental calculated
100/0 187.9 187.0 95/5 192.4 192.1 66/34 222.0 221.5 34/66 257.9 254.0 20/80 270.0 268.2 0/100 289.9 288.5
On the other hand, polarizability of PET and PCT ([P.sub.ET], [P.sub.CT]) can be evaluated assuming the additivity of the bond polarizability of constituent chemical bonds (11). Polarizability of the copolymers [P.sub.CO] may be calculated from the following equation.
[P.sub.CO] = [[Alpha].sub.ET][P.sub.ET] + [[Alpha].sub.CT][P.sub.CT] (3)
where [[Alpha].sub.ET] and [[Alpha].sub.CT] are the compositions (mole fractions) of ET and CT components. Equation 3 may not hold if the atoms at the each end of ET and CT monomer units are not the same. Polarizabilities of copolyesters calculated in this manner are also listed in Table 2. The agreement between calculated and experimentally obtained polarizabilities confirms that the refractive indices and density for the as-spun fibers of copolymers measured in this work, and the copolymer composition obtained in the previous work by using the 1H NMR (5) are reasonably correlated.
For the comparison of the molecular orientation of the high-speed spun copolyester fibers, evaluation of the intrinsic birefringence for each copolymer may be necessary. The intrinsic birefringence can be estimated theoretically from the anisotropic polarizabilities of a fully extended molecule.
The anisotropic polarizabilities of a molecule parallel and perpendicular to the molecular chain axis ([P.sub.[parallel], [P.sub.[perpendicular]]) can be expressed as follows.
[P.sub.[parallel]]= [summation over i] [P.sub.Li][cos.sup.2][[Theta].sub.i] + [summation over i] [P.sub.Ti][sin.sup.2][[Theta].sub.i] (4)
[P.sub.[perpendicular]] = (3P - [P.sub.[parallel]])/2 (5)
where [[Theta].sub.i] is the angle between i-th chemical bond and chain axis of the polymer, and [P.sub.Li] and [P.sub.Ti], the longitudinal and transverse polarizabilities of i-th bond (11).
For the estimation of the anisotropic polarizabilities of fully extended chain conformation, the angles [[Theta].sub.i] for PET and PCT were obtained using the atomic coordinates in the unit cell (12, 13) because both PET and PCT molecules are known to have almost fully extended conformation in the crystal. Only the conformation of trans-isomer was considered for the cyclohexane ring in PCT.
With the lack of well-defined knowledge for the fully extended conformation of copolymer molecules, the anisotropic polarizabilities of copolyesters ([P.sub.[parallel]-CO], [P.sub.[perpendicular] - CO]) were evaluated from the polarizabilities of PET and PCT as follows. In the case of PET, the angle between molecular axis and C-O bond axis which connects two monomer units is 18.4 [degrees]. This angle for PCT is 30.9 [degrees] . Thus, when the ET unit is connected to the CT unit, molecular axes of ET and CT units may not be directed to the same direction. The inclination angle between these two units [[Theta].sub.ET/CT] may be around 12.5 [degrees] . Considering this inclination angle, [P.sub.[parallel]-CO] and [P.sub.[perpendicular] - CO] were estimated using the following equations.
[P.sub.[parallel]-CO] = [[Alpha].sub.ET]([P.sub.[parallel]-ET][cos.sup.2][[Theta].sub.ET] + [P.sub.[perpendicular] - ET][sin.sup.2][[Theta].sub.ET]) + [[Alpha].sub.CT]([P.sub.[parallel]-CT][cos.sup.2][[Theta].sub.CT] + [P.sub.[perpendicular] - CT][sin.sup.2][[Theta].sub.CT]) (6)
[P.sub.[perpendicular] - CO] = (3[P.sub.CO] - [P.sub.[parallel]-CO])/2 (7)
where [[Theta].sub.ET] and [[Theta].sub.CT] are the angles between copolymer chain axis and molecular axis of each monomer unit. The angles [[Theta].sub.ET] and [[Theta].sub.CT] were estimated from the ET and CT monomer unit lengths, [L.sub.ET] and [L.sub.CT], and [[Theta].sub.ET/CT] as shown in Fig. 5.
On the other hand, anisotropic refractive indices [Mathematical Expression Omitted] were estimated according to the Lorentz-Lorenz equation (10), and the intrinsic birefringence [Delta] [n.sub.*] was calculated as [Mathematical Expression Omitted]. For this calculation, the density of copolyesters were evaluated from the volume fraction of ET and CT components, and the volume fraction was obtained from the composition and crystal density of each component. The result of the calculation is summarized in Table 3. It has to be noted that there remains some ambiguity in the extended chain conformation and the definition of the ideal density for the copolyesters. The estimated intrinsic birefringence, however, may not be affected significantly by these uncertainties.
The birefringences [Delta] n shown in Fig. 2 were normalized by the intrinsic birefringence of each copolyester [Delta] [n.sup.*] and replotted in Fig. 6. The value [Delta] n/[Delta] [n.sup.*] corresponds to the Hermans orientation factor. It can be recognized from the Figure that below 5 km/min, the orientation factor for copolyesters with higher CT composition increased more steeply with the take-up velocity. The maximum orientation factor for all the polyesters showed almost the same value of [approximately] 0.4 to 0.45.
In high-speed melt spinning, when viscosity of the polymer melt is not extremely high and the spun filament is relatively thick - these are the cases in this experiment - the spinline stress at the solidification point is dominated by the inertia stress (14). This fact implies that the solidification stress, which has [TABULAR DATA OMITTED] close relation with the molecular orientation of as-spun fibers, are almost similar for all the polymers. Consequently, the difference in the orientation behavior might be attributable to the difference in the rigidity of the polymer chains. In other words, the CT component has lower flexibility than the ET component, and the rigidity of copolyester chains increase with an increase in the CT component.
The relation between the orientation factor and density is plotted in Fig. 7. It is widely known that this relation for PET follows two straight lines, and the intersect of the two lines corresponds to the onset point of orientation-induced crystallization (7). Except for the result of ET/CT = 66/34 fibers, the data for each polyester showed the intersect of two straight lines. The shaded area in the Figure corresponds to the region where the crystallization occurred in the spinline. Although polymers with higher CT composition tended to orient under low tensile stress, the orientation-induced crystallization of PCT appeared to start at higher molecular orientation than PET. With the addition of CT copolymer component to PET, and ET to PCT, the onset of crystallization shifted to higher orientation. Moreover, the increase in density, which corresponds to the progress of crystallization, also became smaller.
Mechanical properties of as-spun fibers from four polyesters are shown in Fig. 8. The initial Young's modulus and tenacity increased, whereas the elongation at break decreased with the take-up velocity. On the other hand, the modulus, elongation and tenacity decreased with an increase in the CT composition. At lower take-up velocities, the modulus of the fibers with higher CT composition increased more rapidly. This result and the lower elongation for higher CT composition polymers may be attributable to the steeper increase in the orientation factor with the take-up velocity.
Because of the bulkiness of CT component, the cross-sectional area per chain for PCT is larger than that for PET. Moreover, the chain extending bonds from the phenylene ring in PCT are more inclined to the chain axis than that in PET. These facts suggest that the theoretical crystalline modulus and tenacity along the chain axis for PCT is lower than those for PET. This might be one of the reasons for the inferior mechanical properties of fibers with higher CT composition.
Structural Change With Annealing of ET/CT = 66/34 Fibers
For the fibers of ET/CT = 66/34, which did not crystallize in the high-speed melt spinning process, the variation of birefringence with increasing temperature was investigated. As shown in Fig. 9, the birefringence started to decrease at [approximately] 80 [degrees] C which corresponds to the glass transition temperature of the polymer. This phenomena is due to the orientation relaxation. The onset of relaxation appeared to shift to lower temperatures with the development of molecular orientation.
The birefringence of filaments spun at 1 and 2 km/min entirely disappeared above 90 [degrees] C, whereas re-orientation occurred above [approximately] 90 [degrees] C in the cases of 4 and 6 km/min as-spun filaments and drawn filaments. The birefringence of high-speed spun filaments even became higher than that of the original filaments above [approximately] 120 [degrees] C. At 130 [degrees] C, the birefringence started to decrease again and vanished at 190 [degrees] C.
Since the driving force for the spontaneous re-orientation was speculated to originate from the crystallization, that is, the incorporation of molecules into the oriented crystals, the 4 km/min fibers were heat-treated in the same manner as in the birefringence measurement up to 140 [degrees] C, and subjected to the WAXD measurement. The equatorial WAXD intensity curves for the as-spun and heat-treated fibers were compared in Fig. 10. The appearance of three crystalline peaks after the heat-treatment confirmed that the ET/CT = 66/34 copolymer, which has been recognized as an amorphous polymer, can form PET crystals if an adequate thermal history is applied under molecular orientation.
High-speed spinning of ET/CT copolyesters was performed. Molecular orientation of copolyesters with higher CT composition increased more steeply with the take-up velocity, however, saturated orientation values for all the studied copolyesters were almost the same. The orientation-induced crystallization of PCT started at a lower take-up velocity than PET. On the other hand, molecular orientation at the onset of crystallization was lower for PET than that for PCT. In the orientation-induced crystallization during high-speed spinning process of copolyesters, crystals of the major component were generally formed. With the addition of CT component to PET, and ET to PCT, the crystallization became less significant. In the as-spun fiber of ET/CT = 66/34 copolyester, which has been recognized as an amorphous polyester, the crystallization was not observed even at the attained highest take-up velocity. It was confirmed, however, that these fibers can form PET crystals if an adequate thermal history is applied under the molecular orientation. It may be concluded from these results that while the incorporation of copolymer component to the homopolymer suppresses the crystallization behavior, the molecular orientation enhances the crystallization of copolymer systems significantly in a similar manner as in the cases of homopolymers.
1. C. J. Kibler, A. Bell, and J. G. Smith, U.S. Pat. 2,901,466 (1959),
2. P. A. Aspy and E. E. Denison, Modern Plastics, August, 1983, p. 74.
3. R. Benavente and J. M. Perena, Polym. Eng. Sci., 27, 913 (1987).
4. R. Benavente and J. M. Perena, Macromol. Chem., 189, 1207 (1988).
5. N. Yoshie, Y. Inoue, H. Y. Yoo, and N. Okui, Polymer (in press).
6. H. Y. Yoo, S. Umemoto, T. Kikutani, and N. Okui, Polymer (in press).
7. J. Shimizu, N.Okui, and T. Kikutani, Sen'i Gakkaishi, 37, T-135 (1981).
8. R. P. Daubeny, C. W. Bunn, and C. J. Brown, Proc. R. Soc., A226, 531 (1954).
9. C. A. Boye, J. Polym. Sci., 55, 275 (1961).
10. M. F. Vuks, Opt. Spectrosc., 20, 361 (1966).
11. C. W. Bunn and R. D. Daubeny, Trans. Farad. Soc., 50, 1173 (1954).
12. Y. Y. Tomashpol'skii and G. S. Markova, Polym. Sci. U.S.S.R., 6, 316 (1964).
13. B. Remillard and F. Brisse, Polymer, 23, 1960 (1982).
14. J. Shimizu, N. Okui, and T. Kikutani, in High-Speed Fiber Spinning, p. 173, A. Ziabicki and H. Kawai, eds., John Wiley & Sons, New York (1985).
|Printer friendly Cite/link Email Feedback|
|Author:||Kikutani, Takeshi; Morohoshi, Katsumi; Yoo, Hee Yeoul; Umemoto, Susumu; Okui, Norimasa|
|Publication:||Polymer Engineering and Science|
|Date:||Jun 15, 1995|
|Previous Article:||Chaotic mixing in the enhanced mixing simulator.|
|Next Article:||Effect of holding pressure on orientation distribution.|