Evolution of the microstructure and morphology of polyimide fibers during heat-drawing process

Abstract

The development of crystalline structure and morphology for polyimide (PI) fibers in the heat-drawing process was investigated by simultaneous synchrotron wide-angle X-ray diffraction (WAXD) and small-angle X-ray scattering (SAXS). WAXD results indicated that the drawing process resulted in a high crystal orientation and ordered crystal structure. Especially, as the drawing ratio increases to 2.0, a well-defined crystalline structure forms in the fibers. We propose that the highly oriented molecular chains induce the formation of crystalline regions. Namely, an orientation-induced crystallization occur with stretching in the case of the heat-drawing polyimide fibers. The meridional scattering streaks in the SAXS patterns for the as-spun fibers suggest the presence of periodic lamellar structure in the fibers. These crystalline lamellae may evolve to more complete crystalline regions. The size of microvoids in the cross-section of the PI fibers is analyzed by SAXS. As a result, the drawing process leads to the orientation of microvoids along the fiber, and to reduced diameter of the microvoids in the fiber. Dynamic thermomechanical analysis indicates that the activation energy E_a of α relaxation increases with the increase in the crystallinity and orientation in the fibers.

---

1. Introduction

High performance fibers are the enabling technology for numeruos present and future high-technology products, and significant progress has been made in high performance fibers that are produced from rigid and flexible polymers.^1,2 In the late 1960s, the lyotropic liquid-crystalline behavior of rigid-rod-like aromatic polyamide was discovered at Du Pont, which led to the development of Kevlar fibers.^3,4 A few years later, another high-performance fiber was obtained from thermotropic aromatic polyesters by Eastman Kodak, Du Pont and Celanese.⁵ In the 1980s, aromatic heterocyclic fibers, such as poly(p-phenylenebenzobisthiazole) (PBT) and poly(p-phenylenebenzobisoxazole) (PBO), were developed.^6,7 Among these polymeric fibers, aromatic polyimide fibers have also been successfully prepared.^8–10 PI fibers have been recognized as one of the important members because of their excellent thermal stability, mechanical properties along with their good chemical resistance, low creep and excellent radiation shielding capability.¹¹

The basic technology for the preparation of organic fibers includes spinning the as-spun fiber and heat-drawing treatment. Four major techniques are usually used to prepare the fibers including melt, dry, wet and dry-jet wet spinning process. Among these methods, wet and dry-jet wet spinning process have widely been used for fabricating high-performance aromatic polymeric fibers such as Zylon, Kevlar, M5 and PI fibers.^12,13 In the wet or dry-jet wet spinning process, water or aqueous solutions are often used as coagulating agents. The extrudate is passed into a non-solvent bath where fiber coagulation and solvent extraction occur at the same time. Following the spinning, fibers are usually treated by fixed end annealing or annealing under tension. In most cases, heat or heat-drawing treatment is always applied to improve the mechanical properties of the fibers. In general, various spinning methods and heat-drawing treatment processes affect chain orientation, crystallinity and morphologies in the fibers. These basic parameters affect the final performances of the fibers, especially the mechanical properties. The relationship between the structure and mechanical behavior of polymeric fibers has been a long-standing challenge in academic and industrial communities.

For the aromatic PI materials, the relationships among structure, morphology and properties have been studied extensively.^14–16 For instance, Russell et al.¹⁷ reported that the aggregation structures of pyromellitic dianhydride/4,4′-diaminodiphenyl ether (PMDA/ODA) films ranged from amorphous structures to ordered crystalline structures, depending on the film thickness and preparation conditions. In general, PI fibers do not exhibit definitive crystalline diffraction peaks, indicating the absence of large domains with mesomorphic order between the crystalline and amorphous phases during the spinning process at room temperature, which can be interpreted as liquid crystalline-like (LC-like) ordered domains.¹⁸ The structural changes of the PI fibers during deformation have been investigated in several studies by a variety of methods, including thermomechanical analyzer,¹⁹ scanning electron microscopy,²⁰ transmission electron microscopy and wide angle X-ray diffraction (WAXD) techniques.²¹ However, small angle X-ray scattering (SAXS) as a useful tool has been rarely used to characterize the microstructural evolution during the heat-drawing treatment for the PI fibers due to its penetrability and statistics. It has been known that microvoids in the fiber play an important role for mechanical properties, especially for tensile strength. Therefore, to characterize the evolution of microvoids during the heat-drawing treatment is significant for the preparation of high performance PI fibers. Recently, simultaneous WAXD/SAXS methods have become a unique tool to investigate the structure and morphology of polymers. Synchrotron radiation provides a more powerful technique to carry out online research using simultaneous WAXD and SAXS for the study of fiber deformation.^22,23

In the present work, 2D WAXD and SAXS methods were carried out to investigate the transformation of microvoids and crystallite structure during the heat-drawing treatment process of PI fibers. The information obtained may provide new clues for understanding the structural evolution in the preparation of high-performance fibers.

2. Experimental

2.1. Preparation of the PI fibers

Co-polyimide was synthesized from 3,3′,4,4′-benzophenonetetracarboxylic dianhydride (BTDA), 2,2′-bis(trifluoromethyl)-4,4′-diaminobiphenyl (TFMB) and 2-(4-aminophenyl)-5-aminobenzimidazole (BIA), as described previously in detail.²⁴ In this work, polymerization was carried out with a diamine ratio of TFMB/BIA = 10/90. Equimolar dianhydride and diamine were mixed in N-methyl-2-pyrrolidone (NMP), and the solution was stirred for 3 h at room temperature. Then, isoquinolin (∼0.2 g) was added and further stirred for 3 h at 120 °C. Finally, the mixture was kept at 150 °C for 10 h to ensure that part of the polyamic acid cyclized into PI.

The synthesized solution was filtered and degassed at 60 °C prior to spinning. The PI fibers were prepared by wet-spinning. The PI dopes were extruded through a spinneret (50 holes with a diameter of 80 μm) into a coagulation bath. The solidified filament entered into the second and third washing baths with 60 °C water, and then clustered at the take-up. The fibers were dried at 300 °C for 1 h, and then drawn with various ratios in a furnace over 450 °C.

2.2. Measurements

Morphologies of the fibers were observed on a scanning electron microscope (SEM) (HITACHI SU8010) at an accelerating voltage of 3.0 kV. To observe the cross-section of the fibers, the fibers with various drawing ratios were cooled in liquid nitrogen for half an hour first, and were then fractured. The measurement of thermal dynamic mechanical behavior of the PI fibers was carried out on a thermomechanical analyzer (TA, Q800). Two dimensional wide angle X-ray diffraction/small angle X-ray scattering (2D WAXD/SAXS) profiles were obtained at Beamline (16B1) in Shanghai Synchrotron Radiation Facility (SSRF). The wavelength used was 0.124 nm. A CCD X-ray detector (MAR CCD 165) was employed at distances of 79.54 mm and 5060 mm from the sample for WAXD and SAXS measurements, respectively. In all the patterns, the fiber axes were oriented vertically. All the data analysis (background correction, radial and azimuthal integration) was carried out using Xpolar software (Precision Works Inc., NY, USA).

The X-ray crystallinity determinations of the fibers were carried out by subtracting the background, corresponding to the WAXD pattern of the amorphous glass obtained from the as-spun fibers without any drawing during spinning. The crystal orientations in the fibers were measured based on the Hermans equation:

f_c100% = [3〈cos² ϕ_c〉− 1]/2 (1)

where f_c is the orientation factor along the fiber direction, and ϕ_c represents the angle between the fiber axis and the c axis of the crystal unit cell. Because the (002) crystallographic plane of the fibers along the meridional direction is the most isolated and clear diffraction spot, the numerical values for the mean-square cosine in eqn (1) can be determined from the fully corrected intensity distribution diffracted from the (002) crystallographic planes and I_c(ϕ_c, α), which is averaged over the entire surface of the orientation sphere:

For wet-spun PI fibers, previous studies²⁵ have indicated that the streak on the equator in the SAXS of the fibers is attributed to the scattering of the needle-shaped microvoids along the fiber direction. Therefore, the radius of microvoids with cross-section can be described by Guinier functions,²⁶ as shown in eqn (3):

where R is the radius of microvoids with circular cross-section, q(q = 4π sin θ/λ) is the scattering vector, θ is the scattering angle (−5 to 5°) and λ is the wavelength. And the average microvoids length (L) and misorientation B_Φ, which is parallel to the fiber axis, are determined by the method of Ruland from the following equation:

If the microvoids are perfectly aligned along the fiber direction and have a uniform finite length, L, then the width of the streak in reciprocal space is independent of the scattering vectors (s = 2 sin θ/λ). Both the effects of finite length and orientation can be attributed to the width of the equatorial scattering streak. If we assume that these effects can be described by Lorentzian/Cauchy-type functions, then the angular spread (B_obs) of the experimental data as a function of s can be given by eqn (4).²⁷

3. Results and discussion

3.1. Mechanical properties

In this work, to avoid the effect of the drawing temperature on the structure and mechanical properties of the PI fibers during the manufacturing, the samples were heated drawing at a fixed temperature of 450 °C with various ratios. The mechanical properties of PI fibers with different drawing ratios are shown in Table 1. For the as-spun fibers, the tenacity is 0.57 GPa with a large elongation of up to 13%. With increasing drawing ratios, the tenacity increases remarkably and it reaches 1.43 GPa at a drawing ratio of λ = 1.6 and 2.13 GPa at λ = 2.3, respectively. Correspondingly, the tensile strain at break decreases from 4.0% at λ = 1.6–2.1% at λ = 2.3. The initial modulus calculated from the ratio of the tensile stress to strain at the 1% strain shows a drastic increase as the drawing ratio increases. The as-spun fibers have an initial modulus of 4.3 GPa, which increases to 35.0 GPa at λ = 1.6 and then reaches 109.2 GPa at λ = 2.3. As a result, the mechanical properties of the fibers have been remarkably improved by the heat-drawing treatment, which may be attributed to the fact that during the drawing process, the polymer chains have a better opportunity to rearrange themselves into defect-free positions under the large deformation and oriented along the fibers' axis.

Table 1 Mechanical properties of PI fibers prepared at different heat-drawing ratios

Samples	Diameter (μm)	Modulus (GPa)	Tenacity (GPa)	Elongation (%)
λ = 0	24.5 ± 0.2	4.3 ± 0.2	0.57 ± 0.03	13.0 ± 0.5
λ = 1.5	21.7 ± 0.1	30.5 ± 1.5	1.32 ± 0.07	4.3 ± 0.3
λ = 1.6	20.8 ± 0.1	35.0 ± 3.8	1.43 ± 0.07	4.0 ± 0.2
λ = 1.9	18.9 ± 0.2	50.4 ± 2.5	1.69 ± 0.08	3.4 ± 0.1
λ = 2.0	18.3 ± 0.2	61.3 ± 5.0	1.83 ± 0.09	3.0 ± 0.3
λ = 2.3	17.3 ± 0.1	109.2 ± 5.4	2.13 ± 0.11	2.1 ± 0.1

---

Fig. 1 shows scanning electron microscopy (SEM) images of the surface (top) and cross-section (bottom) of the PI fibers with various drawing ratios. In general, the first orientation of the macromolecular chains occurs in the spinneret. The molecular orientation partially maintains in the coagulation bath because of rapid PI precipitation. The as-spun fiber exhibits a circular cross-section as well as a dense morphology (Fig. 1(B)). Upon increasing drawing ratios, the diameters of the fibers are varied from 24.5 μm to 17.3 μm. Moreover, the cross-section of the fibers becomes increasingly smooth and uniform when the drawing ratios further increase. It is suggested that the internal structure of the fibers becomes denser and the mechanical properties improve if the drawing conditions are optimized further.

Fig. 1 SEM images of the surface (top panels) and cross-section (bottom panels) of the PI fibers with various drawing ratios: (A1), (A2) as-spun fiber; (B1), (B2) λ = 1.5; (C1), (C2) λ = 2.3.

3.2. Crystalline structure and molecular orientation

As two of the basic structure parameters, orientation and crystallinity in the fibers play important roles in the properties of the polymeric fibers. Fig. 2 shows the WAXD patterns of the PI fibers with various drawing ratios. Along the equator, for the as-spun fiber, the equatorial arcs originating from lateral packing are obscure, which are so-called “amorphous halos”. Dramatic changes occur with the increase of drawing ratios, and the reflections become much sharper and two clear diffraction points can be observed, indicating the formation of well-defined lateral packing ordered regions in the fibers. For details, 1D WAXD intensity profiles in the equators are shown in Fig. 3. As can be observed, the as-spun fiber shows an amorphous diffraction pattern. The absence of crystalline diffraction for the as-spun fiber indicates that the fiber consists of a poorly ordered lateral packing structure of PI chains. Two diffraction peaks observed at 2θ = 10.2° and 16.4° appear and gradually become sharper and stronger with increasing drawing ratios, indicating the formation of lateral packing order regions in the fibers. Interestingly, when the drawing ratio increases over 1.9 (Fig. 2(E) and (F)), several diffraction arcs can be seen on the quadrants, which illustrates that the fibers show a well-defined crystalline structure due to the heat-drawing treatment. In addition, the diffraction at 2θ = 10.13° for the fiber with a draw ratio of 1.5 gradually shifts to 10.21° during the drawing process. The crystal interplane d-spacing along the equator is calculated with Bragg's law as follows:

nλ = 2d sin θ (5)

Fig. 2 WAXD patterns of the PI fibers with various heat-drawing ratios: (A) as-spun fiber, (B) λ = 1.5, (C) λ = 1.6, (D) λ = 1.9, (E) λ = 2.0 and (F) λ = 2.3.

Fig. 3 WAXD intensity profiles on the equator direction of the prepared fibers with different drawing ratios.

As shown in Fig. 5(A), with increasing the drawing ratios, the crystal d-spacing for the fiber at λ = 2.3 decreases to 0.68 nm, which is 0.025 nm lower than the fiber at λ = 1.5.

Fig. 4 WAXD intensity profiles on the meridian of the prepared fibers with different drawing ratios.

Fig. 5 Changes in the d-spacing along equator (A) and meridian (B) of the fibers with various drawing ratios.

In the meridian of the WAXD patterns as shown in Fig. 2(A)–(F), there are also substantial changes in diffraction patterns upon the drawing ratio. The 1D WAXD intensity profiles on the meridional direction of the PI fibers with different drawing ratios are shown in Fig. 4. For the as-spun fiber, six diffraction streaks at 2θ = 4.5°, 9.1°, 15.4°, 20.6° and 28.6° with relatively weak intensity can be observed. When the fibers are stretched to higher drawing ratios, the intensity of the diffraction streaks gradually becomes stronger and three additional streaks at 2θ = 6.3°, 11.8° and 33.2° appear, indicating that highly ordered structure along the meridian forms and that the aggregation structure of the chains in the fibers is very close to a highly ordered crystal structure. That is, an orientation-induced crystallization does occur with the stretching in the case of the heat-drawing PI fibers, which is different from other aromatic polymer fibers such as Kevlar, PBO and M5 fibers.²⁸ On the basis of our previous work, we believe that these regularly stacking regions originate in BTDA-BIA segments, in which the biphenyl or phenyl-benzimidazole groups prefer to take the coplanar conformation in the crystalline phase. Whereas, the TFMB units are preferentially excluded from the ordered domains because of their three-dimensional asymmetry. The crystal d-spacing along the meridian at 2θ = 9.1° shows an opposite trend compared with the equatorial direction as shown in Fig. 5(B).

For polymeric fibers, the molecular orientation plays an important role in affecting the mechanical properties. To investigate the molecular orientation in crystalline and amorphous regions, an azimuthal scan was performed on WAXD patterns. The (002) (2θ = 8.9°) diffraction streak is used for calculating crystalline orientation because it is very clean and isolated along the meridian direction, whereas for the amorphous orientation at 2θ = 13–14°, no crystalline peak is present. In the case of crystalline orientation (Fig. 6(A)), as the drawing ratio increases, the peak intensity at 90°, which corresponds to the orientation along the fiber axis, increases. Therefore, a preferential chain orientation occurs in the crystalline region upon the drawing process. Based on the Hermans equation as shown in eqn (1) and (2), we can calculate the Hermans orientation factor, which is used to quantitatively characterize the orientation coefficient. The values of preferred orientation factor for the heat-drawing fibers are 0.66, 0.71, 0.73, 0.82, 0.85 and 0.93. The degree of orientation increases with the continual stretching of PI fibers. On the other hand, an azimuthal scan of the amorphous regions is performed at 2θ = 13–14° on the WAXD patterns, and the amorphous orientation of PI fibers is examined. Fig. 6(B) shows that the molecular orientation in the amorphous region is different from that in the crystalline region. For the as-spun fiber, no specific peak is observed, indicating that amorphous orientation hardly occurred. However, peaks corresponding to amorphous orientation along the direction of fiber axis appear at λ = 1.5. According to the Herman's equation, the values of preferred orientation factor in the amorphous regions are 0.51, 0.60, 0.61, 0.64, 0.63 and 0.64, which indicates that the amorphous orientation is almost independent of the drawing ratios. According to the abovementioned discussion, we can conclude that macromolecules in the crystalline regions are more parallel to the fiber axis and are sensitive to the external force than the amorphous regions.

Fig. 6 Azimuthal scan of PI fibers with various drawing ratios in the crystalline region (2θ = 8.9°) (A) and amorphous regions (2θ = 13–14°).

3.3. Micromorphology

In the following section, we illustrate the corresponding morphological changes extracted by the SAXS technique. During wet spinning, the coagulation of fibers is a dual diffusion process, namely, solvent diffusing into the bath and the coagulant diffusing into the fibers. However, the inner force and external force cannot reach equilibrium because of different concentration grad for the solvent and coagulation, leading to defects such as microvoids in the fibers. Huang et al.²⁵ proposed that the diffuse SAXS feature in wet-spun PI fibers could mainly be attributed to microvoids, where these voids were elongated parallel to the fiber axis, as indicated by the greatest extension of scattering in the equatorial direction. During the heat-drawing process, parameters such as the length L, radius and the orientation of the microvoids will change with the change in the aggregate state in the fibers. Fig. 7 shows the SAXS patterns of PI fibers with various drawing ratios. Apparently, all the scattering patterns show sharp and elongated streaks, attributed to the presence of microvoids in the fibers. The elongated shape of the equatorial streak indicates that microvoids are needle-shaped and align parallel to the fiber axis direction. Upon increasing the drawing ratios, the elongated shapes become more and more sharp, illustrating the presence of morphological changes in the fibers.

Fig. 7 SAXS patterns of the PI fibers prepared with various heat-drawing ratios: (A) as-spun fiber, (B) λ = 1.5, (C) λ = 1.6, (D) λ = 1.9, (E) λ = 2.0 and (F) λ = 2.3.

Interestingly, in the meridian direction of the SAXS patterns, the appearance of the scattering peaks in Fig. 7(A)–(C) from the PI fibers suggests the presence of periodic lamellar structure between the crystalline and amorphous regions, which has never been found in the PI fibers in the past. However, with the increase of the drawing ratios, these meridional streaks become weaker gradually and disappear with a drawing ratio λ = 1.9. As shown in Fig. 2, a well-defined crystalline structure appears when the drawing ratio increases to 1.9. Here, we assume that these crystalline lamellae are stretched into more complete crystalline regions under external force. Then, the long period L_p may be estimated from the position of a meridional scattering maximum,²⁹ which is around 85 nm calculated by the Bragg equation.

As reported by Jiang et al.^30,31 the radius of microvoids with cross-section can be described by Guinier functions, as shown in eqn (4). The Guinier plots of the scattered intensities along the equatorial streak in the horizontal direction were obtained, and subsequently the tangent curve of the Guinier plot was also determined. The Guinier plots were subtracted by the tangent curve and formed new values, the abovementioned procedure was repeated. As shown in Fig. 8, by resolving the curve into successive tangents, a typical polydispersed system can be obtained. The radius of the microvoids in the fibers can be calculated according to eqn (3), and the results are listed in Table 2. Obviously, microvoids in the PI fibers show multi-order cross-section characteristics, which may be mainly dependent on the coagulation conditions. In addition, the radius of the microvoids decreases with the increase in drawing ratios.

Fig. 8 Guinier plots of the scattered intensities along the layer line (q = 0 Å⁻¹) in the horizontal direction: (A) as-spun fiber, (B) λ = 2.3. R₁, measured values; R₂, values after subtracting the resolved curve R₁; and R₃, values after subtracting the resolved curve R₁ and R₂.

Table 2 Microvoids parameters of the PI fibers with various drawing ratios

Drawing ratio	R1 (Å)	R2 (Å)	R3 (Å)	Length (Å)	BΦ (°)
As-spun	21.6	43.6	86.1	4098	9.57
1.5	18.3	42.8	80.5	5413	7.33
1.6	17.9	41.9	79.9	5541	7.30
1.9	17.0	35.2	73.1	6078	7.33
2.0	16.3	34.7	72.5	6147	6.9
2.3	14.3	30.8	65.6	3154	6.5

---

Ran et al.²⁷ reported that if the scattered objects (microfibrillar or microvoids) were perfectly aligned in the fiber direction and had a finite length L, then the angular width of the streak should be constant and the width should not be a function of the scattering angle. However, in our case, this is not true, implying that both the length L and the misorientation of the microvoids have contributed to the streak profile. We used the following method proposed by Ruland³² to analyze the intensity distribution to obtain the information regarding the average length L and misorientation of the microvoids from the fiber axis. As shown in Fig. 9, the peak profiles from azimuthal scans of the equatorial streak of the as-spun fibers are better fitted with Lorentzian function, as expressed in eqn (4).

Fig. 9 Representative azimuthal scans of the equatorial streak from the as-spun fiber. The dots represent the experimental data, the red and blue lines correspond to the Lorentzian and Gaussian fit, respectively.

Fig. 10(A) and (B) show the azimuthal scans extracted at different scattering vectors and the Ruland plot for the as-spun fiber, respectively. Thus, the length L is obtained from the intercept of the B_obs vs. s⁻¹ plot and the slope (B_Φ) yields the microvoids misorientation width. The results are listed in Table 2. It can be assumed that the length L and the misorientation angle Φ in the fibers have a distribution and the measured value represents a mean quality. The data in Table 2 demonstrate that the length L increases at λ < 2.0 and then decreases at λ = 2.3. It is conceivable that some voids may be stretched with a longer length upon the drawing process. Further high strain may result in the breakage of the microvoids, thus the length L decreases at λ = 2.3. The misorientation is found to decrease with increasing the drawing ratios, indicating an increase in the orientation of the microvoids in the fibers.

Fig. 10 (A) Azimuthal scans at different scattering vectors of as-spun PI fibers; (B) Ruland plot of the as-spun PI fiber.

Fig. 11 shows a schematic diagram of the structural deformation of the PI fiber during the drawing process on the basis of WAXD/SAXS analysis. The as-spun fiber is highly defective and contains mainly amorphous regions, microvoids and fractional lamellae, as shown in Fig. 11(A). Under external stretching, the molecular chains in the amorphous regions orient along the fiber axis and the regular lateral packing gradually forms. At a higher drawing ratio, the chain folded lamellae are stretched into crystalline regions under the external force. As the drawing ratios increase to 2.3, polymer chains highly orient along the fiber axis and crystal regions enlarge. Meanwhile, microvoids are stretched to a smaller size, and exhibit a needle-like shape and align parallel to the fiber direction. By this subsequent processing, the fibers possess highly ordered crystal regions and a more uniform and dense structure.

Fig. 11 Schematic diagram of structural evolution for the PI fibers with various drawing ratios.

3.4. Dynamic mechanical properties

Fig. 12(A) shows the DMA spectra for the PI fibers with different drawing ratios. Obviously, PI fibers exhibit gradually enhanced glass transition temperatures (T_gs) with increasing drawing ratios. Moreover, the intensities of α relaxation peaks gradually decrease to a very low level. It should be noted that α relaxation process (the glass transition) corresponds to segmental motion, which is attributed to a mobile amorphous region of polymer materials. Thus, the intensity variations of the α relaxations reflect the energy level of segmental motion of the non-crystalline regions. Moreover, both the crystallinity and the crystalline orientation have significant influence on the intensity of α relaxation peaks. The low α relaxation intensities of the fibers with higher drawing ratios are suggested to be caused by the highly ordered crystalline regions and regular macromolecular packing, as reflected in the 2D WAXD patterns.

Fig. 12 (A) DMA curves of the prepared PI fibers with different drawing ratios; (B) changes in tan δ of the fibers with λ = 2.3.

Fig. 12(B) shows the dynamic mechanical spectra in a frequency region of 0.5 to 100 Hz for the PI fiber at λ = 2.3. The applied frequency has an obvious influence on the T_gs. When frequency increases from 0.5 to 100 Hz, the T_gs increase from 363 °C to 386 °C. The relationship between logarithmic frequency and the reciprocal of the peak temperature is plotted, as shown in Fig. 13. The active energy E_a of the α relaxation can be calculated from its slope using the Arrhenius equation.³³ We have obtained the activation energy of 501 kJ mol⁻¹ for the as-spun fiber and 887 kJ mol⁻¹ for the fiber at λ = 2.3. The increase in the activation energy indicates that the cooperative motion is enhanced possibly due to the increase in crystallinity and orientation in the fibers, which hampers the noncooperative motion.

Fig. 13 Relationships between logarithmic frequency and reciprocal temperatures for two fibers.

4. Conclusions

In this study, synchrotron X-ray diffraction and scattering techniques were employed to investigate the changes in crystal structure and morphology during drawing process. A well-defined crystalline structure was formed in the fibers by the heat-drawing treatment. The preferred orientation factor for the heat-drawing fibers was up to 0.93 at a drawing ratio of λ = 2.3, implying that the polymer chains highly oriented along the fiber axis. SAXS results illustrated the presence of lamellae structure and dominant microvoids and the sizes of the microvoids decreased as the drawing ratios were increased, resulting in a more uniform structure. Consequently, with the drawing process, the microstructure and morphology of the fibers were optimized, which is an essential step for fabricating high performance fibers.

Acknowledgements

Financial support of this work is provided by NSFC (51233001, 51173024), 973 Plan (2014CB643603), and Chinese Universities Scientific Fund (14D310605). Thanks to SSRF for the WAXD/SAXS measurement.

Notes and references

1. H. G. Chae and S. Kumar, Science, 2008, 319, 908–909.
2. M. Afshari, D. J. Sikkema, K. Lee and M. Bogle, Polym. Rev., 2008, 48, 230–274.
3. J. Preston, Polym. Eng. Sci., 1975, 15, 199–206.
4. D. Tanner, J. A. Fitzgerald and B. R. Phillips, Angew. Chem., Int. Ed. Engl., 1989, 28, 649–654.
5. D. Acierno, F. P. La Mantia, G. Polizzotti, A. Ciferri and B. Valenti, Macromolecules, 1982, 15, 1455–1460.
6. X. D. Hu, S. E. Jenkins, B. G. Min, M. B. Polk and S. Kumar, Macromol. Mater. Eng., 2003, 288, 823–843.
7. S. Ran, C. Burger, D. Fang, X. Zong, S. Cruz, B. Chu, B. S. Hsiao, R. A. Bubeck, K. Yabuki and Y. Teramoto, Macromolecules, 2002, 35, 433–439.
8. S. K. Park and R. J. Farris, Polymer, 2001, 42, 10087–10093.
9. M. Eashoo, D. Shen, Z. Wu, C. J. Lee, F. W. Harris and S. Z. D. Cheng, Polymer, 1993, 34, 3209–3215.
10. M. Eashoo, Z. Wu, A. Zhang, D. Shen, C. Tse, F. W. Harris, S. Z. D. Cheng, K. H. Gardner and B. S. Hsiao, Macromol. Chem. Phys., 1994, 195, 2207–2225.
11. Q. H. Zhang, M. Dai, M. X. Ding, D. J. Chen and L. X. Gao, Eur. Polym. J., 2004, 40, 2487–2493.
12. M. Lammers, E. Klop, M. Northolt and D. Sikkema, Polymer, 1998, 39, 5999–6005.
13. E. W. Choe and S. N. Kim, Macromolecules, 1981, 14, 920–924.
14. S. Z. D. Cheng, Z. Wu, M. Eashoo, S. L. C. Hu and F. W. Harris, Polymer, 1991, 32, 1803–1810.
15. J. Wakita, S. Jin, T. J. Shin, M. Ree and S. Ando, Macromolecules, 2010, 43, 1930–1941.
16. K. Takizawa, J. Wakita, S. Azami and S. Ando, Macromolecules, 2011, 44, 349–359.
17. T. Russell, H. Gugger and J. Swalen, J. Polym. Sci., Polym. Phys. Ed., 1983, 21, 1745–1756.
18. Y. Zhuang, X. Liu and Y. Gu, Polym. Chem., 2012, 3, 1517–1525.
19. A. C. Comer, D. S. Kalika, B. W. Rowe, B. D. Freeman and D. R. Paul, Polymer, 2009, 50, 891–897.
20. T. Sukhanova, Y. G. Baklagina, V. Kudryavtsev, T. Maricheva and F. Lednický, Polymer, 1999, 40, 6265–6276.
21. S. Z. D. Cheng, F. E. Arnold, A. Zhang, S. L. C. Hu and F. W. Harris, Macromolecules, 1991, 24, 5856–5862.
22. Y. Liu, K. Cui, N. Tian, W. Zhou, L. Meng, L. Li, Z. Ma and X. Wang, Macromolecules, 2012, 45, 2764–2772.
23. M. O. Oh and S. H. Kim, Polym. Int., 2014, 63, 1247–1253.
24. J. Dong, C. Yin, W. Luo and Q. Zhang, J. Mater. Sci., 2013, 48, 7594–7602.
25. S. B. Huang, Z. M. Gao, X. Y. Ma, H. Q. Guo, X. P. Qiu and L. X. Gao, Chem. Res. Chin. Univ., 2012, 28, 752–756.
26. A. Guinier and G. Fournet, Small-angle scattering of X-rays, Wiley, 1955.
27. S. Ran, D. Fang, X. Zong, B. Hsiao, B. Chu and P. Cunniff, Polymer, 2001, 42, 1601–1612.
28. H. G. Chae and S. Kumar, J. Appl. Polym. Sci., 2006, 100, 791–802.
29. X. Chen, C. Burger, F. Wan, J. Zhang, L. Rong, B. S. Hsiao, B. Chu, J. Cai and L. Zhang, Biomacromolecules, 2007, 8, 1918–1926.
30. G. Jiang, W. Huang, L. Li, X. Wang, F. Pang, Y. Zhang and H. Wang, Carbohydr. Polym., 2012, 87, 2012–2018.
31. G. Jiang, Y. Yuan, B. Wang, X. Yin, K. S. Mukuze, W. Huang, Y. Zhang and H. Wang, Cellulose, 2012, 19, 1075–1083.
32. W. Ruland, J. Polym. Sci., Part C: Polym. Symp., 1969, 1, 143–151.
33. Q. H. Zhang, W. Q. Luo, L. X. Gao, D. J. Chen and M. X. Ding, J. Appl. Polym. Sci., 2004, 92, 1653–1657.

---