R. Terry K. Baker
Department of Chemistry, 10 Hurtig Hall
Northeastern University, Boston, MA 02115
In recent years, a new type of fibrous carbon material has been developed in our laboratory from the metal catalyzed decomposition of certain hydrocarbons at temperatures ranging from 400 - 800°C (Rodriguez 1993; Rodriguez, Chambers, and Baker 1995). The nanofibers consist of graphite platelets perfectly arranged in various orientations with respect to the fiber axis, giving rise to assorted conformations. One of the most outstanding features of these structures is the presence of large number of edges, which in turn constitute sites readily available for chemical or physical interaction, particularly adsorption. Perhaps one of the most unexpected findings is that such ordered crystalline solids can exhibit high surface areas (300 - 700 m2/g), where the totality of the surface area is chemically active. From the physical point of view, carbon nanofibers vary from 5 to 100 microns in length and are between 5 to 100 nm in diameter.
From in-situ electron microscopy studies it has been possible to determine the sequence of events leading to the formation of carbon nanofibers. The key steps in the process are shown schematically in Figure 9.1. When a hydrocarbon is adsorbed on a metal surface (A) and conditions exist that favor the scission of a carbon-carbon bond in the molecules, then the resulting atomic species may dissolve in the particle (B), diffuse to the rear faces, and ultimately precipitate at the interface (C) to form a carbon nanostructure. The degree of crystalline perfection of the deposited fiber (D) is dictated by the chemical nature of the catalyst particle, the composition of the reactant gas, and the temperature. In this particular case, the graphite platelets are oriented in a "herringbone" arrangement. Surface science studies (Goodman et al. 1980; Nakamura et al. 1989) have revealed that certain faces favor precipitation of carbon in the form of graphite, whereas less ordered carbon will be deposited from other faces (Yang and Chen 1989). By judicious choice of the catalyst, the ratio of the hydrocarbon/hydrogen reactant mixture, and reaction conditions, it is possible to tailor the morphological characteristics, the degree of crystallinity, and the orientation of the precipitated graphite crystallites with regard to the fiber axis. In the "as-grown" condition the graphite layers are separated from one another by a distance of 0.34 nm. This spacing can be increased by introducing selected groups between the layers, a process known as intercalation, thereby generating new types of sophisticated molecular sieves. Such unique structural conformations found in carbon nanofibers opens up numerous possibilities in the fabrication of new materials.
Depending on the crystallographic orientation of the faces that exist at the metal-carbon interface, it is possible to generate nanofibers that consist entirely of graphite platelets or contain a certain fraction of amorphous carbon, i.e., exhibit a duplex structure. Over the past few years we have performed a very comprehensive evaluation of the potential of a number of metals and bimetallics as catalysts for the production of carbon nanostructures. From the data compiled from these experiments it appears that certain nickel- and iron-based alloys are among the most effective catalysts for the reaction (Kim, Rodriguez, and Baker 1992; Rodriguez, Kim, and Baker 1993 a and b; Krushnankutty, Rodriguez, and Baker 1996).
While the "herringbone structures" are frequently found when alloy catalysts are used in the nanofiber process, we have learned that it is possible to tailor this arrangement, and two further conformations are shown in Figure 9.2. Examination of the high resolution electron micrographs clearly indicates that the graphite platelets in these two examples are aligned in directions perpendicular (Fig. 9.2a) and parallel (Fig. 9.2b) to the fiber axis. A more detailed appreciation of these structures can be seen from the respective 3-D models, where the darker geometric shapes represent the metal catalyst particles responsible for generating these conformations. The metal catalyst particles (< 0.4%) can be easily removed by acid treatment, thus producing high purity graphite nanofibers.
The laboratory scale production of carbon nanofibers has been optimized, and it is possible to routinely grow 100 gram quantities in a given experiment. It is estimated that if the method can be successfully scaled up to commercial dimensions, then large quantities of the material could be grown at the relatively low cost of about $2.00/lb., approximately one-tenth of the commercial price of graphite.
Fig. 9.2. High resolution electron micrographs and schematic representation of carbon nanofibers with their graphite platelets, (a) "perpendicular" and (b) "parallel" to the fiber axis.
Because of the high mechanical strength exhibited by some types of carbon nanofiber structures, the material can be used in liquid phase reactions where it can not only withstand vigorous agitation, but also provide improved transport properties over more conventional adsorbates. Furthermore, because of their size dimensions, separation of the nanofibers from liquid phase reactants and products is a relatively simple task.
Advanced polymeric composites have several advantages (including high specific strength and energy absorption, light weight, styling flexibility, good noise/vibration/harshness characteristics, and excellent corrosion resistance) that make such materials ideal for fabrication of body parts for heavy vehicles. Furthermore, technological advances in processing and materials appear to make advanced composites suitable for high-volume applications; low-pressure fabrication processes such as resin transfer molding could require very low investment costs and, depending on the choice of resin and type of fibers, offer fast cycle times. The high strength-to-weight ratio combined with superior stiffness have made carbon fibers the material of choice for high performance composite structures. For most current applications, composites are manufactured with continuous fibers produced from the thermal decomposition of organic polymer precursors such as cellulose (rayon), polyacrylonitrile (PAN), or pitch (Bacon 1973; Delmonte 1981; Donnet and Bansal 1984; Watt and Perov 1985; Fitzer 1985; Otani 1965; Edie 1990). Carbon fibers have been used to reinforce several matrices, including polymers, metals, and also brittle materials such as carbon and ceramics. When used in the latter application carbon fibers require some form of protective coating to prevent both chemical interaction with the matrix and attack by oxygen during service at high temperature. In high temperature applications a critical factor that can determine the viability of a given fiber/matrix composite is the difference in the thermal expansion coefficients of the two components, which can result in mechanical failure during thermal cycling.
We have focused on modifications in the chemical, physical, and mechanical properties of carbon fibers resulting from the growth of carbon nanofibers on their surfaces. Emphasis was placed on establishing the conditions for optimum growth of carbon nanofibers without concomitant degradation in the mechanical properties of the parent fibers. Using this "whiskerization" procedure, tests demonstrated that it was possible to obtain an improvement of over 4.5 times in the interfacial shear strength of the fibers following deposition of a critical amount of nanofibers (Downs and Baker 1995). In another investigation we showed that incorporation of carbon nanofibers into various epoxy resins had a dramatic impact on the curing reaction; small amounts of the filler tended to retard the process, whereas large amounts of nanofibers exerted the opposite effect. Modifications in curing characteristics were compared with those produced from incorporating carbon black into the same resins, and major enhancements in both the mechanical and electrical properties of the composites were found when large amounts of carbon nanofibers were present in the systems (Yin et al. 1993).
Recent reports have disclosed a rather serious problem associated with several nuclear production and processing facilities in the United States. Over the years, large quantities of low level radioactive waste, some in aqueous solution that also may contain significant amounts of organics, have been accumulating at these sites in various types of storage vessels. These materials have been found to leak from the containers and contaminate the surrounding soil over large areas. Efforts are being mounted to clean up these regions and procedures developed to remove and concentrate heavy metals and radio nuclides, many of which are present in the form of aqueous solutions.
A number of technical approaches have been proposed to overcome this problem. One method utilizes activated carbon as an adsorbent to remove trace quantities of metals. While this form of carbon is very effective for the adsorption of gaseous molecules, its performance in a liquid medium is far from satisfactory. Some success has been achieved by the use of electrochemical deposition on certain metal electrodes. While such a procedure is extremely effective at lowering the concentration of metallic species on a small scale, extension of the concept to a commercially viable scale has been thwarted due to the relatively low surface area of the metallic electrode. One of the major problems when dealing with metal electrode systems is their vulnerability to attack by acid or alkali media. Certain noble metals are exceptions to this rule; however, in these cases expense of the material becomes a prime concern. Because of its unique physical and chemical properties combined with a relatively low cost, graphite appears to offer distinct advantages over metal electrodes, but demineralization of dilute solutions requires the use of carbon electrodes possessing considerably higher surface areas than those presently available.
We are attempting to overcome the shortcomings of active carbon and graphite by using graphite nanofibers. In this investigation, graphitic nanofibers are being employed as the electrode material, in the fabrication of unique types of electrochemical filtration, and in selective collection systems for the removal of heavy metals from liquid phase streams. The electrode consists of graphite nanofibers either in the form of a free-standing interwoven network or grown on a conductive support. When a sufficiently high negative potential is applied to the nanofibers, the metal ions in solution are deposited in the fully reduced state, and over an extended period of time the uniformly deposited film can grow up to several monolayers in thickness. If the reduction potential of a given metal is more negative than that of water (i.e., more negative than -1.05 V vs. Ag/AgCl) then an alternative pathway is available -- electrosorption of ions. In this process it is possible to adsorb the contaminant in the cationic state at a more positive potential than its reduction potential. Preliminary studies conducted with graphite nanofibers have indicated that the material exhibits a superior performance for the removal of metal ions from low concentration solutions when used as an electrode in an electrochemical cell over that of graphite felts, metal mesh, and other forms of conductive carbon. Furthermore, by application of a reverse potential, the metal species that are bound on the nanofiber surface can be rapidly released into a more concentrated solution for eventual recovery or disposal.
Failure to create a practical storage system has up to now prevented hydrogen from reaching the commercial forefront as a transportation fuel. The ideal hydrogen storage unit should be lightweight, compact, relatively inexpensive, safe, and reusable without the need for complex regeneration treatments. Four methods are currently being considered for hydrogen storage in commercial applications: (1) pressurized gas storage, (2) liquefied hydrogen, (3) selected metal hydrides, and (4) refrigerated superactivated carbon. While the latter two approaches may offer advantages over the other technologies with regard to cost and safety aspects, they do have their own drawbacks. Metal hydrides are heavy, expensive, and release heat during the hydrogen absorption process. Active carbons are very effective adsorption agents for a variety of gases; however, interactions between the adsorbent and the solid are only of a physical nature, and as such, retention of hydrogen can only be achieved at cryogenic temperatures. It is clear that both storage and controlled release of hydrogen at moderate conditions of temperature and pressure are major issues that must be addressed if this technology is to be exploited on a commercial scale. The ideal solution would be an inexpensive, lightweight material capable of not only absorbing and retaining hydrogen at room temperature and moderate pressures, but also possessing the ability to release the gas at moderate temperatures. One material that fulfills this requirement is graphite nanofiber.
Data obtained in our laboratory has demonstrated that when these structures are pretreated under conditions whereby all adsorbed and absorbed gases are eliminated, then on subsequent exposure to hydrogen at moderate pressures, they are capable of absorbing and retaining up to 30 liters of molecular hydrogen per gram of carbon at room temperature. Theoretical calculations made on the hydrogen absorption capacity of single-crystal graphite have indicated that it is possible to store 6.2 liters of molecular hydrogen per gram of graphite as a single flat layer. It is evident from these calculations that the experimentally determined absorption capacity of the material far exceeds that of the theoretical value, which indicates that the process occurs in a multilayer fashion. It is clear from these data that one could transport molecular hydrogen in a liquid-like state without need for refrigeration or the high volume and weight associated with compressed gas. The advantages of graphite nanofibers over other materials for hydrogen storage can be summarized as follows:
It is clear that for the graphite nanofiber technology to go forward to the commercial arena it will be necessary to devote a significant amount of research effort and funding to the development of a large scale production process. In this endeavor the focus should be placed on a growth procedure that is continuous rather than merely a scaling-up of the current batch method. Control of not only the yield of material but also its physical and chemical properties are essential factors in being able to generate large quantities of graphite nanofibers possessing reproducible characteristics.
Bacon, R. 1973. In Chemistry and physics of carbon. Vol. 9, p. 7. Ed. P.L. Walker, Jr. New York: Marcel Dekker.
Delmonte, J. 1981. Technology of carbon and graphite fiber composites. New York: Van Nostrand Reinhold.
Donnet, J.B., and R.C. Bansal. 1984. Carbon fibers. New York: Marcel Dekker.
Downs, W.B., and R.T.K. Baker. 1995. J. Mater. Res. 10: 625.
Edie, D.D. 1990. In Carbon fibers, filaments and composites, p. 43. Ed. J.L. Figueiredo et al. Netherlands: Kluwer Academics Publishers.
Fitzer, E. 1985. Carbon fibers and their composites. New York: Springer-Verlag.
Goodman, D.W., R.D. Kelley, T.E. Madey, and J.T. Yates. 1980. J. Catal. 63: 226.
Kim, M.S., N.M. Rodriguez and R.T.K. Baker. 1992. J. Catal. 134: 253.
Krishnankutty, N., N.M. Rodriguez, and R.T.K. Baker. 1996. J. Catal. 158: 217.
Nakamura, J., H. Hirano, M. Xie, I. Matsuo, T. Yamada, and K. Tanaka. 1989. Surf. Sci. 222: L809.
Otani, S. 1965. Carbon 3: 31.
Rodriguez, N.M. 1993. J. Mater. Res. 8: 3233.
Rodriguez, N.M., A. Chambers and R.T.K. Baker. 1995. Langmuir 11: 3862.
Rodriguez, N.M., M.S. Kim and R.T.K Baker. 1993a. J. Catal. 140: 16.
_____. 1993b. J. Catal. 144: 93.
Watt, W., and B.V. Perov. 1985. Strong fibers. Handbook of composites. Vol. 1. Series eds. A. Kelly and Y.N. Rabotnov. Amsterdam: North Holland.
Yang, R.T., and J.P. Chen. 1989. J. Catal. 115: 52.
Yin, M., J.A. Koutsky, T.L. Barr, N.M. Rodriguez, R.T.K. Baker, and L. Klebanov. 1993 Chem. Mater. 5: 1024.