Printer Friendly

Effect of heating and stretching polyacrylonitrile precursor fibers in steam on the properties of stabilized fibers and carbon fibers.


The preferred orientation of crystallites along the fiber axis is one of the most important structure parameters for carbon fibers (CFs). Higher preferred orientation usually gives rise to higher tensile strength and modulus (1), (2). It is well known that stabilization is a necessary process for converting polyacrylonitrile (PAN) precursor fibers to CFs (3), (4). During stabilization, the preferred orientation of PAN molecules along the fiber axis decreases clue to the thermal shrinkage of molecular chains and chemical reactions of nitrite groups (5), (6). It has been proved that applying tensions along the fiber axis is an effective method to prevent molecular chains from relaxing and losing their orientation (2), (7), (8). However, the preferred orientation of the stabilized fibers is still much lower than that of the precursor fibers. In our previous work, the PAN precursor fibers was heated and stretched in nitrogen before oxidation to form a rigid ladder structure by intramolecular cyclization of linear PAN molecules, as a result, the misorientation of molecules derived from crosslink and oxidative reactions was restricted during further stabilization in air (9). However, it seemed that the fibers with rigid ladder polymers were difficult to be further stretched, although the intermolecular cyclization was avoided in the inert atmosphere. Thus, a more effective stretching method should be introduced.

The strong dipole-dipole interaction between the nitrile groups causes a poor plasticity for PAN fibers (8), (10), so reducing the interaction is essential to obtain a more effective drawing before or during stabilization by postspinning modification. Several attempts on postspinning modification were carried out to reduce molecular dipole interaction by modifying the precursor fibers with plasticizers, such as succinic acid, cuprous chloride, cobaltous chloride, potassium permanganate, and dimethyl formamide (10-14). The effect of moisture on stabilization for PAN fibers was investigated (15), (16), the results indicated that, with the moisture in air increasing, not only the oxidation on the skin of fibers was effectively depressed but also the elongation rate of the fibers was enhanced to a larger extent when the fibers was subjected to a constant load. This suggested that water molecules acted as a plasticizer for the fibers having cyclized molecules. However, the effect of water steam on the tension in fibers and the molecular orientation along the fiber still needs further investigation. In addition, all the heat treatments related to stabilization, mentioned in referred literatures, were carried out in an oxygen-containing atmosphere, the oxidation and crosslink reactions would make it difficult for the molecules to reorient further.

Based on the previous studies, the pretreatment of heating and stretching the PAN precursor fibers in a steam bath mixed with nitrogen (HSSN) was carried out to improve the preferred orientation of ladder molecules. The stretching ratio was further enhanced under the effect of steam as a plasticizer.



PAN precursor fibers (wet-spun from acrylonitrile/acrylamide/methacrylic acid copolymer, and having 3000 filaments per tow and 11.3 [micro]m average diameter) were supplied by Mitsubishi Co.

Heat Treatment

Stabilization of the PAN fibers was conducted continuously through a 10-zone tube furnace, as illustrated in Fig. 1. The fibers entered from Z1 and went through all the 10 zones successively under different conditions of temperature, stretching ratio and atmosphere. From Z1 to Z4, the fibers were heat treated and stretched in an inert atmosphere. Z1 and Z2 were filled with nitrogen, and Z3 and Z4 were filled with a mixture of water steam and nitrogen. The volume ratios of steam to nitrogen ([epsilon]) were controlled at three different values: [[epsilon].sub.1] = 0 (0-100%), [[epsilon].sub.2] = 1 (50-50%), and [[epsilon].sub.3] = 3 (75-25%). The flow rate for nitrogen at the flow meter was maintained at a constant value of 15 ml/s, and that for steam was adjusted to a corresponding value according to the volume ratio of steam to nitrogen. The steam at the flow meter was heated to 220[degrees]C before entering Z3 and Z4. The fibers coming out of Z4 were partially stabilized and were called PSFs. The following six zones (Z5-Z10) were filled with air. The fibers coming out of ZlO were fully stabilized, called SFs.

The densities of PSFs and SF's were maintained at about 1.205 g/c[m.sup.3] and 1.365 g/c[m.sup.3], respectively. To compensate the difference caused by the changes of atmosphere, three temperature programs were used for the three kinds of atmospheres with different steam ratios. Three stretching programs were used in this study. As illustrated in Fig. 1, the tension between the drive rollers D1 and D2 was measured at PI, and so on. Moreover, the stretching ratio between the drive rollers D1 and D2 was expressed as stretching ratio at P1, and so on. All the processing parameters during thermal stabilization were listed in detail in Table 1.

TABLE 1. Processing parameters of thermal stabilization.

                                      form Z1 to Z10

Ratio of steam to      Temperature        1          2    3
[N.sub.2] in Z3 and      program

0 ([[epsilon].sub.1])  [T.sub.a]               185  210  220

1 ([epsilon].sub.2)    [T.sub.b]               185  210  223

3 ([[epsilon].sub.3])  [T.sub.c]               185  210  226

Stretching program


[[lambda].sub.1]                                      5


[[lambda].sub.3]                                      5

Ratio of steam to            4          5    6    7    8    9
[N.sub.2] in Z3 and

0 ([[epsilon].sub.1])             230  205  225  240  250  260

1 ([epsilon].sub.2)               235  207  227  240  250  260

3 ([[epsilon].sub.3])             240  210  230  240  250  260

Stretching program         Stretching
                       ratios from P1
                            to P4 (%)

                                        P2             P3

[[lambda].sub.1]                         4              0


[[lambda].sub.3]                         8              0

Ratio of steam to      10
[N.sub.2] in Z3 and

0 ([[epsilon].sub.1])  270

1 ([epsilon].sub.2)    270

3 ([[epsilon].sub.3])  270

Stretching program


[[lambda].sub.1]        -2


[[lambda].sub.3]        -2

Carbonization was carried out immediately after stabilization in a temperature range of 350-480-600-800-1300[degrees]C in high-pure nitrogen (99.999%) with an initial stress of 15 MPa. Parts of the resultant CFs were graphitized at 2500[degrees]C in high-pure argon (99.999%) with a heating rate of 20[degrees]/min and a holding time of IS min at final temperature in a batch graphitization furnace, to obtain graphite fibers (GFs) with higher Young's modulus. During graphitization, no tension was used to the fibers.


Densities of fibers were measured by a sink-float method with a PZ-B-5 liquid specific gravity balance according to the Chinese Standard GB3362-82.

A DTMB-1000 tension tester was used to measure the tension in the fibers on line. Four testing points, marked as black triangles, are illustrated in Fig. 1. A SCY-III sound velocimeter was used to measure the sound veloc-ity along the fiber axis to indicate the overall orientation of molecules.

Mechanical properties of CF tow were measured using a WDW3020 electronic universal testing equipment according to Chinese Standard CNS-GB-3362/3366-82. Mechanical properties of GF filaments, which were adhered on a gauge paper frame with a length of 20 mm, were measured using a XQ-1C fiber testing equipment at a crosshead speed of 2 mm/min. For each sample, at least 20 filaments were tested to obtain the average value and standard deviation. The diameter of each filament was measured under a CYG-055C optical microscopy with a closed circuit television camera.

The crystalline-related structural parameters of CFs and GFs, including preferred orientation of graphite planes parallel to the fiber axis R, interlayer spacing between graphite planes [d.sub.002], crystallite stacking height Lc(002) and crystallite width La(100), were investigated by X-ray diffraction (XRD) on a D/Max-2550 PC XRD apparatus. The [d.sub.002] and Lc(002) were determined by (002) reflection from equatorial scan. The La(100) was determined by (100) reflection from meridional scan; and the orientation was measured by performing an Azimuthal scan at the fixed Bragg position of the (002) reflection. Bragg formula and Scherrer equation were used to calculate the values of d~ 0, and crystallite sizes, respectively. The full-width at half maximum (FWHM) of diffraction peak from Azimuthal scan was used to estimate the preferred orientation.


Effect of Steam on Cyclization Rate

The densities and colors of PSFs were listed in Table 2. It was noted that, at the same temperature program in Z3-Z4 (220 and 230[degrees]C), the density gradually decreased and the color changed from orange to light yellow with steam ratio [epsilon] increasing from 0 to 3. The results indicated that the cyclization reaction was inhibited at the presence of steam. The nitrile groups might be shielded by the incorporated water molecules (15), (16). Because density was usually used to evaluate the stabilization extent for the PAN fibers, the temperatures in Z3-Z4 were adjusted appropriately higher for the mixed atmosphere with higher ratio of steam to maintain the similar cyclization degree. As shown in Table 1, when the temperatures in Z3-Z4 were set at 223-235[degrees]C for [[epsilon].sub.2] = 1, and 226-240[degrees]C for [[epsilon].sub.3] = 3, respectively, the densities of PSFs were nearly maintained at the same value.

TABLE 2. Densities and colors of PSFs preheated under various

parameters in
Z3 and Z4

Temperatures       Atmosphere           Density       Color
([degrees]C)                         (g/c[m.sup.3])

220 and 230   [[epsilon].sub.1] = 0           1.205   Orange

220 and 230   [[epsilon].sub.2] = 1           1.196   Yellow

220 and 230   [[epsilon].sub.3] = 3           1.189   Light

220 and 235   [[epsilon].sub.2] = 1           1.206   Orange

220 and 240   [[epsilon].sub.3] = 3           1.204   Orange

Tension Development and Molecular Orientation

The tensions along the stabilization line are related to not only the stretching ratios but also the fiber's stabilization degree depending on the temperature and atmosphere. The tensions at four testing points under different stretching ratios in various atmospheres were shown in Fig. 2. The tensions at testing point P1 for [[lambda].sub.1], [[lambda].sub.2], and [[lambda].sub.3] programs were almost equal (about 500 cN) because of the same processing conditions in Z1-Z2. For a given atmosphere program, whatever for [[epsilon].sub.1], [[epsilon].sub.2], or [[epsilon].sub.3], it was found that the tensions at testing points P2 and P3 increased with stretching ratio increasing. However, at the same stretching ratio but in different atmospheres, it was noted that the tensions at testing points P2, P3, and P4 decreased remarkably with the introduction of steam. Particularly, in the case of [[lambda].sub.3], the fibers could not be stretched successfully without steam existing ([[epsilon].sub.1]), because the high stretching ratio made the filaments broken. However, in the atmosphere containing steam ([[epsilon].sub.2] and [[epsilon].sub.3]), the filament break during stabilization was avoided, and the fibers were able to be stabilized and carbonized successfully with a large stretching rate ([[lambda].sub.3]), attributed to the increased plasticity resulting from the reduced internal stress in the fibers.

Because the dipole-dipole interactions among the nitrite groups prevent the molecular chains from becoming oriented during stretching, decreasing the interactions can enhance the stretching ratio and improve the molecular orientation along the fiber axis effectively during the drawing process (14). The results indicated that water stream serving as a plasticizer reduced Van der Waals force between molecular chains.

Sonic velocity along the fiber axis was used to reflect the overall orientation of molecules along the fiber axis including both crystalline and amorphous part. The sonic velocities along the fiber axis for PSFs and SFs were shown in Fig. 3. It was noted that, at the same stretching ratio program, the sonic velocities of PSFs obtained in the mixed atmosphere of steam and nitrogen ([[epsilon].sub.2] and [[epsilon].sub.3]) and the resulting SFs were larger than those of the corresponding PSFs pretreated in nitrogen without steam ([[epsilon].sub.1]) and their resulting SFs, respectively. However, when the steam fraction was too large ([[epsilon].sub.3]), the sonic velocities of PSFs and the resulting .SFs were smaller than those of the corresponding PSFs and SFs derived from [[epsilon].sub.2], but still larger compared with the corresponding PSFs and SFs derived from [[epsilon].sub.1]. The higher values of sonic velocities in PSFs and SFs derived from [[epsilon].sub.1] and [[epsilon].sub.3] indicated that the orientation for both linear and ladder molecules was enhanced by HSSN pretreatment. The results also demonstrated that a suitable amount of steam ([[epsilon].sub.3]) was able to improve the molecular orientation in fibers more effectively, but too much steam ([[epsilon].sub.3]) would suppress the improvement of orientation. Because the cyclization rate decreased in steam atmosphere, it needed higher temperatures to increase the cyclization rate. The larger amount of steam, the higher temperature was needed. Higher reaction temperatures would make the molecular chain segments move and shrink more easily, as a result, the fiber's orientation decreased more or less.

Microstructures and Mechanical Properties of CFs

The XRD patterns of CFs and GFs were shown in Fig. 4, which displayed the crystallite structures as evidenced by the shape, position, intensity, and FWHM of the diffraction peaks. The equatorial scan patterns (Fig. 4a) displayed wide and diffuse (002) diffraction peaks for CFs, but narrow and sharp (002) peaks for GFs. The (100) diffraction peaks for CFs and GFs were indicated in the meridional scan patterns (Fig. 4b). It was noted that the diffuse (002) and (100) peaks converted to sharp ones after graphitization, which was attributed to the structural transformation from turbostratic to ordered during high-temperature treatment. The smaller FWHMs for (002) and (100) peaks implied the larger crystallite height and width in the measured fibers. The preferred orientation R of crystallite planes parallel to fiber axis evaluated by FWHMs of diffraction peaks from the Azimuthal scan (Fig. 4c) was calculated by Eq. as follows:

R = (1 - [H.sub.1] + [H.sub.2] / 360) x (1)

where [H.sub.1] and [H.sub.2] were FWHMs of the peaks at 0[degrees] and 180[degrees], respectively, in Fig. 4c.

The crystallite parameters derived from XRD patterns (i.e. interlayer spacing, preferred orientation, and crystallite dimensions) and mechanical properties for the CFs were listed in Table 3. Comparing the structural parameters and mechanical properties between the CFs modified by HSSN pretreatment and the CFs without HSSN pretreatment, it was noted that not only the crystallite parameters (including preferred orientation and crystallite sizes) were enhanced, but also the mechanical properties (both tensile strength and Young's modulus) were improved for the resultant CFs modified by HSSN pretreatment. Because larger Lc(002) and La(100) means less chain scission during carbonization and fewer crystallite defects in CFs, it was reasonable to believe that higher preferred orientation and larger crystallite dimensions contributed to the improvements of tensile strength and Young's modulus for the modified CFs.

TABLE 3. Crystallite parameters and mechanical properties of CFs
and GFs.

Sample  Stabilization conditions  [d.sub.002]  Lc(002)  La(100)
                                      (nm)      (nm)      (nm)

CF-1      [[epsilon].sub.1],          0.3598     1.47     3.01

CF-2      [[epsilon].sub.2],          0.3564     1.55     3.17

CF-3      [[epsilon].sub.3],          0.3584     1.58     3.16

CF-4      [[epsilon].sub.1],          0.3579     1.47     3.39

CF-5      [[epsilon].sub.2],          0.3573     1.56     3.52

CF-6      [[epsilon].sub.3],          0.3565     1.56     3.71

CF-7      [[epsilon].sub.2],          0.3580     1.55     3.43

CF-8      [[epsilon].sub.3],          0.3578     1.56     3.62

GF-1      [[epsilon].sub.1],          0.3437     6.47     6.27

GF-6      [[epsilon].sub.3],          0.3440     6.97     7.26

Sample   R   Tensile strength  Young's modulus
        (%)         (GPa)              (GPa)

CF-1   77.4     3.02 [+ or -]   238 [+ or -] 3

CF-2   78.3     3.38 [+ or -]   246 [+ or -] 5

CF-3   78.0     3.27 [+ or -]   249 [+ or -] 6

CF-4   77.8     2.95 [+ or -]   231 [+ or -] 5

CF-5   79.6     3.33 [+ or -]   255 [+ or -] 4

CF-6   79.7     3.42 [+ or -]   252 [+ or -] 4

CF-7   79.6     2.91 [+ or -]   250 [+ or -] 8

CF-8   79.5     3.05 [+ or -]   257 [+ or -] 9

GF-1   90.0     2.06 [+ or -]  439 [+ or -] 11

GF-6   90.3     2.41 [+ or -]  500 [+ or -] 13

After graphitization at 2500[degrees]C, the GFs (GF-1 and OF-6) were obtained from the CFs (CF-1 and CF-6), respectively. The crystallite parameters and mechanical properties for OF-1 and GF-6 were also listed in Table 3. The tensile strength and Young's modulus for GF-1 were 2.06 GPa and 439 GPa, respectively. Compared with GF-1, GF-6 had a higher tensile strength of 2.41 GPa and a higher Young's modulus of 500 GPa. According to the crystallite parameters, it was shown that both preferred orientation and crystallite dimensions of GF-6 were larger than those of GF-1, which was considered as the reason for the mechanical results.


Heating and stretching the PAN precursor fibers in a steam bath mixed with nitrogen before oxidation reduced the intermolecular interactions, enhanced the stretching ratio and improved the molecular orientation along the fiber axis. The shrinkage tension during stabilization was effectively decreased at high-ratio drawing, because water molecules acted as a plasticizer. The PSFs and SFs modified by HSSN showed a higher preferred orientation than the corresponding PSFs and SFs without undergoing this pretreatment. The resultant CFs and GFs had higher preferred orientations and larger crystallite dimensions, which gave rise to considerable improvements for the mechanical properties. On the whole, HSSN was a simple and effective method to reduce the shrinkage tension during thermal stabilization, enhance the molecular orientation for PSFs and SFs, and improve the mechanical properties for the resulting CFs and GFs.

Correspondence to: Yonggrn Lu: e-mail:

Contract grant sponsor: National Basic Research Program of China (973 Program); contract grant numbers: 201 1CB605602 and 2011CB605603; contract grant sponsor: National Special Fund for Forestry Scientific Research in the Public Interest; contract grant numbers: 201004057: contract grant sponsor: Cultivation Fund of the Key Scientific and Technical Innovation Project from Ministry of Education of China.

DOI 10.1002/pen.23328

Published online in Wiley Online Library (

[c] 2012 Society of Plastics Engineers


(1.) W. Ruland, Adv. Mater.,2, 528 (1990).

(2.) D.D. Edie, Carbon, 36, 345 (1998).

(3.) Z. Bashir, Carbon, 29, 1081 (1991).

(4.) S. Dalton, F. Heatley, and P.M. Budd, Polymer, 40, 5531 (1999).

(5.) E. Fitzer, W. Frohs, and M. Heine, Carbon, 24, 387 (1986).

(6.) W.X. Zhang, J. Liu, and G. Wu, Carbon, 41, 2805 (2003).

(7.) A.J. Clarke and J.E. Bailey, Nature, 243, 146 (1973).

(8.) G.P. Wu, C.X. Lu, L.C. Ling, A.M. Hao, and F. He. J. Appl. Polym. Sd., 96, 1029 (2005).

(9.) X.Y. Qin, Y.G. Lu, H. Xiao, Y.C. Hao, and D. Pan, Carbon, 49. 4598 (2011).

(10.) J. Mittal, R.B. Mathur, O.P. Bahl, and M. Inagaki, Carbon, 36, 893 (1998).

(11.) D.P. Bahl, R.B. Mathur, and T.L. Dharni, Mater. Sci. Eng.. 73, 105 (1985).

(12.) T.H. Ko and L.C. Huang, J. Mater. Sei., 27, 2429 (1992).

(13.) R.B. Mathur, J. Mittal, and O.P. Bahl, Carbon, 32, 71 (1994).

(14.) J.C. Chen and 1.R. Harrison, Carbon, 40, 25 (2002).

(15.) G.P. Wu, C.X. Lu, Y.H. Li, Y.G. Lu, F. He, and L.C. Ling, New Carbon Mater., 18, 25 (2003).

(16.) G.P. Wu, C.X. Lu, R. Zhang, X.P. Wu, F.Z. Ren, and K.X. Li, J. Mater. Sci, 39, 2959 (2004).

Xianying Qin, Yonggen Lu, Hao Xiao, Weizhe Zhao

State Key Laboratory for Modification of Chemical Fibers and Polymer Materials, College of Materials Science and Engineering, Donghua University, Songjiang District, Shanghai 201620, People's Republic of China
COPYRIGHT 2013 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Qin, Xianying; Lu, Yonggen; Xiao, Hao; Zhao, Weizhe
Publication:Polymer Engineering and Science
Article Type:Report
Geographic Code:9CHIN
Date:Apr 1, 2013
Previous Article:The effect of recycled polymer addition on the thermorheological behavior of modified lubricating greases.
Next Article:Processing and properties of hydrophilic electrospun polylactic acid/beta-tricalcium phosphate membrane for dental applications.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters