« PreviousContinue »
Isotropic Pitch-Based Carbon Fibers
Currently, a variety of carbon fibers are produced from isotropic pitch. However, unless these fibers undergo an expensive and difficult final stretchgraphitization step, their modulus is an order of magnitude less than that of mesophase pitch-based carbon fibers. Although they are useful for applications such as filtration, asbestos replacement, and static dissipation, the poor mechanical properties of isotropic pitch-based carbon fibers limit them to nonstructural applications. Therefore, the manufacturing of isotropic pitch-based carbon fibers has not been included in this chapter, and the reader is referred to Edie (ref. 16) for a detailed discussion of this process.
Manufacture of Vapor-Grown Carbon Fibers
This process is able to produce only short, discontinuous lengths of carbon fiber. Nevertheless, these short fibers may be attractive for applications such as CC brake pads. Since some believe that the vapor-growth process may produce the first low-cost, discontinuous, high-performance, carbon reinforcing fiber, it is not surprising that several companies currently are conducting pilot-scale studies to evaluate its potential.
Although only recently developed as a continuous process for producing highperformance reinforcing fibers, this technique was one of the first used to produce carbon filaments. Hughes and Chambers first detailed the vapor-growth process in an 1889 patent (ref. 18). In this patent they showed that small carbon filaments could be grown in a hot iron crucible under an atmosphere of methane and hydrogen—the vapor-growth process. Vapor growth is a dendritic type of growth from a catalyst particle. As Hughes and Chambers first discovered, metallic particles, normally containing iron as the primary constituent, tend to catalyze the growth of thin, partially graphitic fibers, when exposed to a hydrocarbon atmosphere at temperatures of approximately 1000°C. If the carburizing potential of the gas is low, a fraction of the fibers can be grown to macroscopic length, while still retaining the diameter of the catalyst particle. Because the most effective catalyst particles have a diameter of only 15 nm, these initial fibers are extremely thin. However, if the carbon potential of the gas is raised to a sufficient level, pyrocarbon can be deposited on the surface, permitting the filament diameter to increase to that of conventional carbon fibers (approximately 10 /xm). This step is critical because the small diameter of the initial filaments makes them a potential carcinogen. Because pyrocarbon deposits with the basal planes parallel to the fiber surface, the fiber is highly oriented and has a high modulus. Figure 12 shows a schematic of the various stages of this catalyst-induced growth of carbon fibers (ref. 19).
Tibbetts and his coworkers at General Motors Corporation used this growth technique to produce filaments with lengths up to 30 cm in an atmosphere of
(b) Fiber thickening stage.
Figure 12. Schematic showing fiber growth and thickening stages during vapor-growth process.
methane and hydrogen (ref. 19). Several metals (including nickel, cobalt, ironnickel powder, and Fe(NO3)3) have been used as catalysts (refs. 19-21). Even though the feed stock (methane and hydrogen) was inexpensive and process temperatures of only 1000°C were used, the batch nature of this process used for early studies made it uneconomical for commercialization.
To overcome the low productivity of the batch process, Koyama and Endo (ref. 22) recently patented a continuous method for producing vapor-grown fibers. In their process, the catalyst particles are either incorporated in the feedstock or produced in the reactor by the decomposition of an organometallic. A simple schematic of this vapor-grown carbon fiber process is shown in figure 13. The catalyst and hydrocarbon feed are introduced at the top of the heated reactor, and short fibers are continuously withdrawn from the bottom. Fiber lengthening and thickening can be continuously controlled by adjusting the carbon potential of the gas within the reactor. The technique, which could be considered fluidized catalytic growth, allows carbon filaments with varying length, diameter, and physical properties to be continuously produced. If the thickening step can be accurately controlled and the projected process costs are correct, this continuous process could well replace isotropic and mesophase pitch-based, as well as PAN-based carbon fibers, in composite applications where chopped or short-fiber reinforcement is adequate.
Figure 13. Schematic of continuous process for producing vapor-grown carbon fibers.
Mechanical Properties of Carbon Fibers
As the next chapter will explain, the structure of carbon fibers, to a large extent, controls their tensile strength and modulus. Because of this, the manufacturers of both PAN-based and pitch-based carbon fibers are attempting to develop new methods that can modify this structure during the fiber formation or heat treatment steps. Currently, as figure 14* shows. PAN-based carbon fibers exhibit higher tensile strengths, but lower moduli than mesophase pitch-based carbon fibers. However, the new varieties of mesophase pitch-based fibers' * recently introduced by du Pont and Nippon Steel (denoted in fig. 14 as "Improved mesophase pitch") exhibit significantly improved tensile strengths. The reported moduli for vaporgrown fibers are comparable to carbonized PAN fibers, but their tensile strengths are slightly lower. As expected, isotropic pitch fibers exhibit the lowest strengths and moduli of all carbon fibers. In the manufacture of both PAN-based and mesophase pitch-based carbon fibers, increasing the final heat treatment temperature improves the degree of preferred orientation within the fiber and, thus, the fiber modulus. Because of this, the various grades of fiber available from a particular manufacturer are normally the result of changes in this temperature.
0 100 200 300 400 500 600 700 800 900 1000
Fiber modulus, GPa
Figure 14. Tensile strength and modulus of various types of carbon fibers (ref 19 and footnotes *,f, and \). G means the final heat treatment temperature is above 2000°C. and C indicates it is below 2000°C.
The more perfect graphitic structure of mesophase pitch-based carbon fibers, compared with PAN-based carbon fibers, accounts for its higher thermal
Bacon. R.: Amoco Performance Products, Incorporated, personal communication, 1989.
Sato. K.: Nippon Steel Corporation, Tokyo. Japan, personal communication. 1989.
+ Ross. R.: E. I. du Pont de Nemours and Company, Incorporated, Chattanooga. Tennessee, personal communication. 1989.
conductivity. In fact, mesophase pitch-based fibers recently developed by Amoco Performance Products, Incorporated, exhibit a thermal conductivity that is three times that of copper.
The precursor fibers used to produce current commercial carbon fibers are produced by melt, melt-assisted, and solution spinning. Melt spinning normally is the preferred fiber formation process because it eliminates the problems of solvent recovery and produces a purer precursor fiber. However, conventional melt spinning cannot be used for polymers, such as PAN and cellulose, which degrade below their melt temperatures. Nevertheless, melt-assisted spinning, a new process, permits PAN to be spun as pseudo-melt. The PAN-based and rayon-based precursor fibers are thermoset, carbonized, and graphitized with similar equipment and at similar conditions. However, their low tensile strength makes pitch-based carbon fibers much more difficult to handle before final heat treatment; therefore, special oxidation ovens are often used for this product. Short, vapor-grown carbon fibers represent the latest entry to the high-performance fiber field. If health issues can be adequately addressed, discontinuous filaments could become a low-cost reinforcement for composites.
The authors thank G. P. Daumit of BASF Structural Materials, Incorporated, and Elsevier Science Publishers B. V. for permission to reproduce figure 3. Thanks are also given to Kluwer Academic Publishers for permission to reproduce figures 8, 10, and 12. Finally, the authors thank R. Bacon for generously providing the original drawing of figure 14.
1. Diefendorf, R. J.; and Tokarsky, E.: High-Performance Carbon Fibers. Polymer Eng. & Sci., vol. 15, no. 3, Mar. 1975, pp. 150-159.
2. Ram, Michael J.; and Riggs, John P.: Process for the Production of Acrylic Filaments. U.S. Patent 3,657,409, Apr. 1972.
3. Daumit, Gene P.; and Ko, Yoon S.: A Unique Approach to Carbon Fiber Precursor Development. High Tech—The Way Into the Nineties, Klaus Brunsch, Hans-Dieter Golden, and Claus-Michael Herkert, eds., Elsevier Science Publ. Co., Inc., 1986, pp. 201-213.
4. Delmonte, John: Technology of Carbon and Graphite Fiber Composites. Van Nostrand Reinhold Co., c.1981.
5. Clarke, A. J.; and Bailey, J. E.: Oxidation of Acrylic Fibres for Carbon Fibre Formation. Nature, vol. 243, no. 5402, May 18, 1973, pp. 146-154.