Carbon fibers (CFs) were prepared using low-cost, textile-grade polyacrylonitrile fibers, which were 200% to 400% drawn in a hot water bath at 90°C or/and in a tubular furnace at 180°C. X-ray diffractograms confirmed that the drawing process led to higher crystallinity and molecular orientation. These fibers were stabilized in a convection oven at 25–255°C for 390 min. After stabilization, carbonization was performed to obtain carbon fibers. The tensile strength of CFs without drawing was ∼0.8 GPa; however, CFs with 200% and 200% drawing in a hot water bath at 90°C and in a tubular furnace at 180°C, respectively, showed a tensile strength of ∼1.7 GPa. These results suggest that the drawing process of precursor fibers affected the tensile properties of the resulting CFs significantly.
Carbon fibers (CFs) have been used as reinforcements in various matrices such as polymer, metal, cement, and other carbon materials, owing to their high specific strength and modulus, and the resulting composites have been widely used in various fields from sporting goods to aircraft manufacturing (1, 2). Although Rayon was the first precursors for commercialized CFs, the most manufactured CFs are polyacrylonitrile (PAN)-based CFs because PAN provides a greater yield of CFs than other precursors with high tensile strength (3–5).
PAN-based CFs are prepared by stabilization and carbonization of special-grade PAN fibers. The manufacturing process of PAN-based CFs consists of a low-temperature thermal stabilization in air, followed by a high-temperature carbonization in a nitrogen atmosphere. Stabilization is carried out in air at relatively low temperatures (200–300°C), leading to the conversion of the linear chain of PAN precursors to a thermally stable structure for withstanding high-temperature carbonization processing. These structural changes result in a significant shrinkage of fibers, which should be avoided by tension on PAN fibers in stabilization (6). In addition, it is well known that stabilization including cyclization, dehydrogenation, and oxidation is a necessary and important step in achieving mechanically strong CFs (6). If stabilization temperature is too high, PAN fibers can overheat and fuse or even burn. In contrast, too low a temperature results in an incomplete stabilization, causing poor CF properties. Carbonization involves thermal treatment in an inert environment at more than 1200°C, and most of the noncarbon elements within the stabilized PAN fibers are volatilized in the form of H2O, CH4, NH3, CO, HCN, CO2, and N2 (7). Finally, the carbon content of the resulting fibers increases to 85–99% depending on carbonization conditions.
Although CFs have attractive properties for reinforcing composites, the high price of CFs mainly due to the high cost of processing and precursor (special-grade acrylic fibers for CF such as itaconic acid-modified PAN fibers) has been an obstacle to extending their use in various applications. Therefore, lowering the price of CFs is one of the most important issues with CFs and their composite industries. To reduce processing cost, plasma-assisted stabilization was introduced (8). Compared to the conventional stabilization process of using heat treatment only, a significant decrease in processing time was observed in the former because of the atmospheric pressure of plasma-induced oxygen species, which diffused into PAN fibers for efficient stabilization. Hunt et al. (9) reported patterned CFs prepared with sulfonated polyethylene (PE), which produced carbon residue even though pure PE could not be converted into CFs during carbonization. Because PE is cheap and melt spinnable, which is more suitable for mass production, it can be a good candidate for low-cost CFs. Cheap commercial acrylic fibers as a precursor for carbon fibers, e.g., textile-grade acrylic fibers, can also be used to reduce the price of CFs. Since the intrinsic properties of PAN fibers affect the mechanical and morphological properties of the resulting CFs significantly, intensive research studies on PAN precursors have been done actively (10).Textile-grade PAN fibers for blankets, carpets, and clothes have a thick diameter, lower tensile strength, and various types of co-monomers compared to CF-grade PAN fibers. In recent years, there have been a few reports on the use of textile-grade PAN fibers for preparing carbon fibers (11).
Drawing process in hot water, steam, or a tubular furnace can be used to obtain high-strength PAN fibers; however, there is a possibility of drawing in steam that creates voids in the filaments. Therefore, drawing in a tubular furnace can be used to remove the voids. The drawing process is performed at a temperature of 180°C, and it has been observed that higher temperature can break the fibers by fusion. It is also expected that drawing leads to highly oriented polymer molecular chains along the fiber axis with finer diameter compared to as-received, textile-grade PAN precursor fibers. Furthermore, drawing can substantially reduce the concentration of structural defects such as pores in the fiber and minimize structural inhomogeneities, particularly sheath-core structures during stabilization (12). Hence, drawing of textile-grade PAN fibers can reduce the diameter and distribution of structural imperfections to enhance the mechanical strength of the resulting CFs.
In this study, commercially available, low-cost, textile-grade PAN fibers were used as precursors in producing CFs. The fibers were drawn in a hot water bath or/and in a tubular furnace under various conditions, and structural changes were observed with a wide-angle X-ray diffractometer. The thermal and mechanical properties of drawn fibers were determined using differential scanning calorimetry and single-filament tensile testing, respectively. The fibers were stabilized in air and carbonized in nitrogen atmosphere at 1200°C, and the effect of the fiber drawing process on the mechanical properties of the resulting CFs produced by textile-grade PAN fibers was reported.
2 Results and discussion
2.1 Thermal and structural changes of drawn fibers
As-received, textile-grade PAN fibers were drawn in a hot water bath at 90°C or/and in a tubular furnace at 180°C, and the diameters of fibers decreased with increasing drawing ratio, as expected. The thermal behavior of drawn textile-grade PAN fibers was investigated using non-isothermal differential scanning calorimetry (DSC) analysis. The first heating thermograms at a rate of 10°C/min in N2 atmosphere are shown in Figure 1. The cyclization temperatures (Tc) are summarized in Table 1. The fibers before drawing showed one cyclization peak at 296.0°C. The onset temperature was 280.6°C, and the curves completed at 307.4°C (Figure 1). The peaks shifted down to 292.5°C with drawn fibers in a hot water bath at 90°C for 200%, compared to the as-received PAN fibers, and a further drawing up to 400% led to a further decrease in cyclization temperature (290.7°C). Drawing in a tubular furnace at 180°C for 200% also reduced the cyclization temperature (292.0°C). The fibers drawn in a hot water bath and in a tubular furnace for 200% and 200%, respectively, consecutively showed the lowest cyclization temperature (290.0°C). It is inferred that the drawing resulted in higher molecular orientational order along the longitudinal direction, which helped to reduce the diameter of PAN fibers and the dipole-dipole interactions among the nitrile groups (13). Therefore, less energy was required for cyclization, causing a peak to be observed at lower temperature with more drawing fibers.
|Drawing condition||Tc in N2 (°C)||Crystallinity||Orientation index (%)|
A further structural investigation was performed using X-ray diffractometry. The highest scattering intensity in the diffraction intensity (2θ) profile of as-received PAN fiber and its drawn samples was observed around 2θ of 16.9° (Figure 2), resulting from the (100) plane (13, 14). Another distinct peak with a relatively small intensity appeared around 2θ of 29.4° in all fibers, corresponding to the (110) plane. From Bragg’s equation, d-spacing was found to be 5.25 and 3.03 for the (100) and (110) planes, respectively, and their ratio was 1.73 indicating a hexagonal packing of rod-like PAN polymer chains (15). The degree of crystallinity was estimated using the Bell and Dumbleton method, which was the ratio of the area of the (100) plane to total area in the diffraction intensity (2θ) profile (16). As-received PAN fibers showed ∼18% crystallinity. In contrast, higher drawing temperature revealed a higher crystallinity of up to ∼42% because drawing induced a significant improvement in the packing of polymer chains (Table 1) (13). Therefore, it is evident that drawing does not affect the hexagonal structure of PAN chains even though the degree of crystallinity increased. However, the better hexagonal packing with decreasing diameter of fibers by drawing could play a role in reducing the cyclization temperature observed in DSC (Figure 1).
To investigate the change in orientational order during drawing, the azimuthal intensity distribution from the diffraction patterns of wide-angle X-ray diffraction was obtained. In the diffraction patterns of as-received fibers, broad arcs due to the (100) plane appeared (Figure 3). In contrast, narrower arcs were observed with 200–0% drawn fibers, and further drawing revealed intense arcs with a 400–0% sample. Drawn fibers with 200–200% samples showed the most intense arcs. It was noted that the shorter the arc length, the higher the orientational order. Quantified azimuthal intensity profiles from the (100) peak around 2θ of 16.9° are shown in Figure 4, which showed that the intensity profiles of as-received PAN fiber and its drawn fibers were concentrated in the equatorial region. A higher drawing ratio of fibers without regard to the drawing method led to narrower profiles, indicating that PAN molecular chains become more oriented along the longitudinal direction. From these profiles, the index of orientation was calculated using the following equation (17, 18):
where H is the full width at half the maximum intensity of the (100) peak around 2θ of 16.9°. Wang et al. (17) obtained the orientation index of PAN and its stabilized fibers with different drawing conditions using the equation. The index of orientation of commercially available special-grade acrylic PAN fiber was found to be ∼85% in their study. As expected, the values increased with increasing drawing ratio from 66.7% to 93.3% for as-received PAN fibers and their 200–200% drawn fibers, respectively (Table 1).
2.2 Tensile properties of textile-grade PAN and its carbon fibers
Table 2 shows the tensile properties and diameters of textile-grade PAN and its drawn fibers. The diameters of five samples such as 0–0%, 200–0%, 400–0%, 0–200%, and 200–200% were found to be 18.2±1.6, 14.7±1.6, 11.2±1.3, 14.3±1.5, and 10.1±1.3 μm, respectively, obtained using scanning electron microscopy. The tensile strength of fibers without drawing was 288.8±51.5 MPa. Drawn fibers with 200–0%, 400–0%, and 0–200% samples showed a tensile strength of 657.9±142.5, 815.0±103.1, and 650.2±154.7 MPa, respectively, and the highest value (866.6±112.7 MPa) was observed with the 200–200% drawn samples. It suggests that an increase in orientational order due to drawing resulted in a significant increase in tensile strength. Tensile moduli increased with increasing drawing ratio, as expected. The drawn fibers were stabilized in a convection oven at temperatures ranging from 25°C to 255°C for 390 min with a constant tension. Both ends of fibers were tied to avoid significant shrinkage during stabilization. The diameters of stabilized fibers were found to be 15.6±1.0, 12.3±0.7, 8.5±0.7, 12.1±0.9, and 8.2±0.8 μm with 0–0%, 200–0%, 400–0%, 0–200%, and 200–200% drawing, respectively. Compared to the diameters of drawing fibers before stabilization in Table 2, all fibers showed a decrease in diameter. Fitzer et al. (6) reported that shrinkage during the stabilization of PAN fibers was significant; therefore mechanical stress on PAN fibers in oxidative stabilization, which causes an enhancement of intramolecular cyclization leading to more effective ladder structure, affects the tensile properties of the resulting carbon fibers (6). Chen and Harrison (19) stabilized special-grade acrylic fibers under restraint at constant length to apply tension on them, and they observed a very small decrease in fiber diameters, which is consistent with our results. Figure 5 shows the Fourier transform infrared spectroscopy (FTIR) spectra of fibers before and after stabilization. Before stabilization, distinct absorption bands appeared at 2939, 2243, 1454, and 1360 cm-1 assigned to νC-H in CH2, νC≡N in CN, δC-H in CH2, and δC-H in CH, respectively. With all stabilized fibers, a band at 1595 cm-1 appeared and the intensities of bands at 2939, 2243, and 1454 cm-1 decreased owing to the cyclization and dehydrogenation of fibers, indicating that the fibers were stabilized significantly in a given condition. The resulting fibers were carbonized at 1200°C at a rate of 5°C/min. During carbonization, there was no tension on the fibers.
|Drawing condition||Diameter (μm)||Tensile strength (MPa)||Tensile modulus (GPa)||Strain-to-failure (%)|
Table 3 shows the tensile properties of CFs. The diameter of CFs from as-received PAN fibers was found to be 13.0±1.3 μm, which revealed a tensile strength and a modulus of 0.8±0.1 and 79.2±8.4 GPa, respectively, with a strain value of 1.2±0.2%. Drawn fibers with 200–200% samples after carbonization showed the highest values with a tensile strength and a modulus of 1.7±0.3 and 144.2±12.4 GPa, respectively, with a strain value of 1.2±0.2%, indicating that the higher orientation that developed with the drawing led to higher tensile properties of CFs. The diameter of 200–200% drawn fibers after carbonization was only 7.0±0.9 μm, and it is evident that drawing before stabilization was significant to reduce the diameter of the resulting CFs. Mukundan et al. (20) reported melt-processible acrylonitrile fibers with increasing drawing ratio, resulting in a decrease in fiber diameters from ∼40 to ∼14 μm. They observed that thinner acrylonitrile fibers became thinner, stabilized, and carbonized fibers with higher tensile strength, which is similar to our textile fiber results. To improve the tensile strength of PAN-based CFs, carbon nanotubes (CNTs) were incorporated into CFs. Chae et al. (21) prepared CFs with an additional 1 wt% CNTs, which showed a 64% increase in tensile strength compared to pure CFs. As seen in Table 3, drawing of textile-grade PAN fibers led to a significant enhancement of the tensile strength of the resulting CFs, and 200–200% drawn fibers after carbonization showed a more than 100% increase in tensile strength. Therefore, it is worthy to note that only simple drawing can improve the tensile properties of CFs without the addition of nanomaterials. In Figure 6, scanning electron micrographs revealed circular CFs after carbonization. The results with the tensile properties suggest that textile-grade PAN fibers required a longer stabilization for preparing CF compared to special-grade acrylic fibers for CF such as itaconic acid-modified PAN fibers. It is likely that methyl acrylate as a co-monomer in the textile-grade PAN fibers was not involved in cyclization reaction and even delayed stabilization. In addition, noncyclized methyl acrylate could be gasified during carbonization and create defects in CFs. Therefore, CFs from textile-grade PAN fibers could not show high tensile properties. However, it is interesting to note that the drawing process can somehow overcome the drawbacks of textile-grade PAN fibers in CF manufacturing.
|Drawing condition||Diameter (μm)||Tensile strength (GPa)||Tensile modulus (GPa)||Strain-to-failure (%)|
Low-cost, textile-grade PAN fibers were drawn in a hot water bath at 90°C or/and in a tubular furnace at 180°C. The tensile strength and modulus of fibers without drawing were ∼288.8 and ∼7.9 GPa, respectively. Drawn fibers showed significantly increased tensile properties, and drawing of 200% and 200% in a hot water bath and in a tubular furnace resulted in a tensile strength and a modulus of ∼866.6 and ∼18.9 GPa, respectively. X-ray diffractograms revealed that the drawing process led to higher crystallinity and molecular orientation, contributing to a significant increase in tensile properties. In addition, cyclization temperature shifted to a lower temperature in DSC thermograms when fibers were drawn, resulting in reduction in the dipole-dipole interactions among the nitrile groups and inducing a better hexagonal packing of polymer chains. Those fibers were stabilized in a convection oven at 25–255°C for 390 min. After stabilization, carbonization at 1200°C was performed to obtain CFs. The tensile strength of CFs without drawing was ∼0.8 GPa, but CFs with 200–200% drawing showed a tensile strength of ∼1.7 GPa, indicating that the drawing process of textile-grade PAN fibers affected the tensile properties of the resulting CFs significantly.
4 Experimental part
Textile-grade PAN fibers supplied by Taekwang Industrial Co. Ltd. (Taekwang, Seoul, Korea) as a copolymer of acrylonitrile (AN) and methyl acrylate (MA) (AN/MA 90:10) were used for preparing CFs. The molecular weight of as-received Taekwang fibers was found to be ∼96,000 g/mol by gel permeation chromatography (GPC), which consisted of a Hitachi L-6000 pump (Hitachi, Tokyo, Japan) and Shodex GPC LF-804 columns (Showa Denko America Inc., New York, NY, USA) with a dimethylacetamide flow rate of 1.0 ml/min at room temperature. Polystyrene was used as a standard material.
Textile-grade PAN fibers were drawn in a hot water bath at 90°C or/and in a tubular furnace at 180°C under various drawing conditions. The sample name of “A%–B%” indicates that PAN fibers were drawn for A% in a hot water bath and for B% in a tubular furnace, compared to the original fibers. For this study, five samples were prepared: 0–0%, 200–0%, 400–0%, 0–200%, and 200–200%. The drawn textile-grade PAN fibers were stabilized in a convection oven at temperatures ranging from 25°C to 255°C for 390 min. Both ends of fibers were tied to avoid a significant shrinkage during stabilization. First, the temperature increased from 25°C to 200°C with a heating rate of 3°C/min, and the samples were held for 60 min. For a further temperature increase to 215°C, with a heating rate of 0.5°C/min, the samples were held for 60 min. The same step was repeated at 230°C. The temperature increased to 255°C with a heating rate of 0.5°C/min, and the samples were held for 60 min. Carbonization of stabilized fibers under nitrogen atmosphere (99.9999%) was carried out using a tubular furnace (Korea Kiyon Co. Ltd., Bucheon, Korea). After fiber loading without tension, the temperature increased to a set point of 1200°C at a rate of 5°C/min, and the fibers were cooled down naturally.
The exothermic reaction of textile-grade PAN fibers was determined using a differential scanning calorimeter (DSC-7 module, PerkinElmer, Waltham, MA, USA). Samples were heated up to 400°C at a rate of 10°C/min in nitrogen atmosphere. X-ray diffractograms were obtained using an ATX-G X-ray diffractometer (Rigaku, Tokyo, Japan) with a Cu rotating anode X-ray source over the 2θ range of 10° to 50° to observe the structural change due to various drawing processes. The orientation of PAN and its drawn fibers was observed from the azimuthal intensity distribution of wide-angle X-ray diffraction (WAXD) patterns obtained using a D8 DISCOVER X-ray diffractometer (Bruker AXS GmbH, Karlsruhe, Germany) using a Cu sealed-tube X-ray source. To investigate the degree of stabilization, the FTIR spectra of stabilized fibers were obtained using a Nicolet iS10 FTIR spectrometer (Thermo Scientific, Waltham, MA, USA) at a range of 400–4000 cm-1 in attenuated total reflectance mode. For tensile testing of PAN fibers and the resulting carbon fibers, a single fiber was loaded on a paper tab with 25 mm gauge length. An epoxy resin was applied to both ends of fibers, which were then cured for 24 h at 50°C. Fiber diameters were measured by an optical microscope (Olympus BX51, Olympus, Tokyo, Japan). The tensile properties were determined using an Instron 5567 universal tester (Instron, Norwood, MA, USA) at 25°C. The crosshead speed was 5 mm/min for all specimens, and 20 replicates were tested at each condition. The surfaces and cross sections of PAN and its carbon fibers were observed by scanning electron microscopy (S-4100 and FE S-4800, Hitachi, Tokyo, Japan).
This work was supported by a grant from Korea Institute of Science and Technology Institutional program, Republic of Korea.
1. Lee SH, Kim J, Ku B-C, Kim J, Joh H-I. Structural evolution of polyacrylonitrile fibers in stabilization and carbonization. Adv Chem Eng Sci. 2012;2:275–82.10.4236/aces.2012.22032Search in Google Scholar
2. Fitzer E. Pan-based carbon fibers – present state and trend of the technology from the viewpoint of possibilities and limits to influence and to control the fiber properties by the process parameters. Carbon. 1989;27(5):621–45.10.1016/0008-6223(89)90197-8Search in Google Scholar
3. Richard B, Millington LA, Robert C. U.S. Patent 3,322,489, 1967.Search in Google Scholar
4. Yoneshiga I, Teranishi H. Japanese Patent 2775/70, 1970.Search in Google Scholar
6. Fitzer E, Frohs W, Heine M. Optimization of stabilization and carbonization treatment of PAN fibres and structural characterization of the resulting carbon fibres. Carbon. 1986;24(4):387–95.10.1016/0008-6223(86)90257-5Search in Google Scholar
7. Lee SM. International encyclopedia of composites. New York: VCH Publishers; 1990. pp. 158–253.Search in Google Scholar
8. Lee SW, Lee HY, Jang SY, Jo S, Lee HS, Choe WH, Lee S. Efficient preparation of carbon fibers using plasma assisted stabilization. Carbon. 2013;55:361–5.10.1016/j.carbon.2012.10.062Search in Google Scholar
10. Kobashi T, Takao S. U.S. Patent 4,535,027, 1985.Search in Google Scholar
11. Sedghi A, Farsani RE, Shokuhfar A. The effect of commercial polyacrylonitrile fibers characterizations on the produced carbon fibers properties. J Mater Process Technol. 2008;198(1–3):60–7.10.1016/j.jmatprotec.2007.06.052Search in Google Scholar
13. Chen JC, Harrison IR. Modification of polyacrylonitrile (PAN) carbon fiber precursor via post-spinning plasticization and stretching in dimethyl formamide (DMF). Carbon. 2002;40(1):25–45.10.1016/S0008-6223(01)00050-1Search in Google Scholar
16. Tsai JI, Lin CH. Effect of comonomer composition on the properties of polyacrylonitrile precursor and resulting carbon fiber. J Appl Polym Sci. 2003;43(4):679–85.Search in Google Scholar
17. Wang P, Liu J, Yue Z. Thermal oxidative stabilization of polyacrylonitrile precursor fiber – progression of morphological structure and mechanical properties. Carbon. 1992;30(1): 113–20.10.1016/0008-6223(92)90114-CSearch in Google Scholar
18. Alexander AE. X-ray diffraction methods in polymer science. New York: Wiley Interfaces; 1969.Search in Google Scholar
19. Chen JC, Harrison IR. Modification of polyacrylonitrile (PAN) carbon fiber precursor via post-spinning plasticization and stretching in dimethyl formamide (DMF). Carbon. 2002;40(1):25–45.10.1016/S0008-6223(01)00050-1Search in Google Scholar
20. Mukundan T, Bhanu VA, Wiles KB, Johnson H, Bortner M, Baird DG, Naskar AK, Ogale AA, Edie DD, McGrath JE. A photocrosslinkable melt processible acrylonitrile terpolymer as carbon fiber precursor. Polymer. 2006;47(11):4163–71.10.1016/j.polymer.2006.02.066Search in Google Scholar
21. Chae HG, Minus ML, Rasheed A, Kumar S. Stabilization and carbonization of gel spun polyacrylonitrile/single wall carbon nanotube composite fibers. Polymer. 2007;48(13):3781–9.10.1016/j.polymer.2007.04.072Search in Google Scholar
©2014 by Walter de Gruyter Berlin/Boston
This article is distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.