Introduction

For centuries, diamond and graphite have been the only known crystalline forms of carbon. However, in recent years, new carbon allotropes have been revealed. Kroto et al. [1] discovered fullerenes in 1985 while Ijima discovered multi-walled carbon nanotubes in 1991 [2] and single-walled carbon nanotubes in 1993 [3]. Single-walled nanotubes were independently and almost simultaneously observed by Bethune et al. [4]. The carbon bonds in fullerenes and nano-tubes are more similar to those in graphite than to those in diamond. Diamond has a coordination number of four with sp3 hybridization while the same as graphite involves three-coordinate carbons, in which three electrons are in sp2 hybridization and one is delocalized. Fullerenes and nanotubes also have carbon bonds with sp2 hybridization as graphite, but while the graphite structure is made up of flat planar honeycomb, the structures of fullerenes and nano-tubes involve a high degree of curvature. Fullerenes such as C60 and C70 exhibit a closed-cage carbon structure that is spherical or nearly spherical.

Carbon nanotubes can be formally described as a gra-phene honeycomb rolled into a single-walled cylinder or into several concentric graphene cylinders. The former are termed single-walled nanotubes (SWNTs) and the latter multiwalled nanotubes (MWNTs). The diameter of a typical SWNT is in the order of a nanometer but its length can be up to several micrometers, or more. Each nanotube can be considered as a single molecule made up of a honeycomb network of covalently bonded carbon atoms. A schematic representation of a SWNT structure is shown in Figure 1. Each SWNT is fully characterized by two integers (n, m).

These integers specify the number of unit vectors ct1 and a2 in the graphene structure that constitute the roll-up vector (or chiral vector) V = na1 + ma2- The graphene structure is rolled up such that the chiral vector V forms the nanotube circumference. Therefore, the (n, m) indices determine the nanotube diameter, according to the equation d = (n2 + m2 + nm)1'2 0.0783 nm

At the same time, these indices determine the orientation of the carbon hexagons with respect to the nanotube axis. This orientation is the so-called "chirality" of the nanotube. Nanotubes with indices (m, 0) are termed "zigzag" due to the shape of the atomic configuration along the perimeter of such a nanotube. When m = n, the resulting nanotubes are called "armchair" because the position of the C atoms arrange in an "armchair" pattern.

Single-walled carbon nanotubes have unique electronic and mechanical properties combined with a very light weight. Depending on their n, m values, which determine their chirality and diameter, nanotubes can be either electrically metallic or semiconductor. At the same time, they have shown evidence for high stiffness (Young modulus), high resilience, and the ability to reversibly buckle and collapse. Therefore, they are promising candidates as components of strong fibers with light weight and high electrical conductivity. At the same time, other characteristics of SWNTs have generated intense attention. For example, due to their high aspect ratio (i.e., high length/diameter ratio and their chemical and mechanical stability they show promise as excellent electron field emitters for commercial applications as flat panel displays and other cold cathode emitters. Nanotubes can be functionalized with a wide variety of chemical groups, which greatly broadens the scope of their applications in diverse fields that range from conductive coatings to fuel cells or from nanosensors to biotechnology.

Carbon nanotubes can be produced by a variety of methods. Iijima first produced nanotubes by arc discharge [2, 3]. The system differed only slightly from that developed earlier by Kratschmer et al. for the production of fullerenes [5].

Copyright © 2004 by American Scientific Publishers All rights of reproduction in any form reserved.

Encyclopedia of Nanoscience and Nanotechnology Edited by H. S. Nalwa Volume 10: Pages (125-147)

Figure 1. (a) Middle section of a SWNT (http://www.phys.psu.edu/ ~crespi/research/carbon.id/public/). (b) Identification of carbon nanotubes in terms of the indices (n, m) (http://www.photon.tutokyo.ac.jp/ ~maruyana/kataura/chirality.html/). Consider a flat graphene sheet. Vectors n and m placed at the chosen origin (0, 0) determine the specific nanotube. For example, the nanotube (7, 7) can be obtained by drawing a line between the origin (0, 0) and the point indicated as (7, 7) and then cutting the sheet along two lines perpendicular to this line. The cut sheet can then be folded into the (7, 7) nanotube.

Figure 1. (a) Middle section of a SWNT (http://www.phys.psu.edu/ ~crespi/research/carbon.id/public/). (b) Identification of carbon nanotubes in terms of the indices (n, m) (http://www.photon.tutokyo.ac.jp/ ~maruyana/kataura/chirality.html/). Consider a flat graphene sheet. Vectors n and m placed at the chosen origin (0, 0) determine the specific nanotube. For example, the nanotube (7, 7) can be obtained by drawing a line between the origin (0, 0) and the point indicated as (7, 7) and then cutting the sheet along two lines perpendicular to this line. The cut sheet can then be folded into the (7, 7) nanotube.

A similar system had been employed by Bacon more than 30 years earlier for the production of graphite whiskers [6]. In the system used by Ijima [2], the graphite electrodes were kept separated by a small gap, rather than being in contact as in Kratschmer's method. It was observed that about half of the carbon evaporated from the anode deposited as nano-tubes on the negative electrode while the rest condensed in the form of soot. One of the most important observations was that SWNTs were only obtained when metal particles were present in the anode. The metals used in the earlier experiments were cobalt and nickel, packed inside bored graphite rods. Later, other metals and mixtures with lanthanides were found to catalyze the formation of SWNTs as well [7]. In a typical arc discharge apparatus [8], the arc is generated between two graphite rods, which are mounted in a stainless steel vacuum chamber equipped with a vacuum line and a gas inlet. One of the graphite electrodes (cathode) is fixed and it is connected to a negative potential. The other electrode (anode) is moved from outside the chamber through a linear motion feedthrough to adjust the gap between the electrodes. A viewport is mounted on the chamber to allow for a direct observation of the arc discharge.

In a standard operation, a given background pressure is stabilized within the cell by adjusting the incoming flow of an inert gas such as helium and the pumping speed. It has been found that with a continuous flow of He better results are obtained than with a static He pressure. A stabilized voltage of about 20 V is applied while the electrodes are far apart. As the anode is moved in, a gap space is reached (1-3 mm) at which arcing occurs. Depending on the He pressure, the rod diameter, and the gap between the electrodes, the electric current can vary between 40 and 250 A. During the discharge, a plasma is formed generating temperatures of the order of 3700 °C. The temperature is particularly high on the anode surface and this electrode is consumed by vaporization. To keep the proper interelectrode gap constant during operation, the position of the anode must be adjusted manually [9] or by an automated system [10]. The synthesis process lasts only a few minutes [11]. After the synthesis, several types of carbon deposits are obtained: (a) a rubbery soot that condenses on the chamber walls; (b) a weblike structure between the cathode and the chamber walls; (c) a cylindrical deposit on the cathode face; (d) a small collar around the cathode deposits, which contains the highest concentration of SWNTs.

Studies have shown that a number of experimental parameters have important effects on SWNT yield and selectivity. For example, increasing the diameters of both electrodes results in yield and selectivity losses, but decreasing the anode diameter, while keeping the cathode diameter and current density constant, results in an increase in yield [12]. Another critical parameter that greatly affects the SWNT yield is the pressure of the background gas, usually He. While some authors indicate that the yield increases with He pressure [12], other authors have found that beyond a certain pressure, further increase in pressure does not result in an increase in yield [8] and it may even result in yield losses [13]. Not only the yield of SWNTs but also their diameter distribution is affected by the background gas pressure. Saito et al. [14] observed that the distribution of SWNT diameters shifted systematically to small values as the helium pressure decreased. The most frequently occurring diameters were centered at 1.4 nm for the production conducted at 1520 Torr, and it shifted to 0.95 nm for the production at 50 Torr. Among the various experimental parameters, the choice of catalytic additives is of paramount importance. In the first place, the characteristic feature that is common to all SWNT synthesis methods is the participation of a catalyst. So far, no SWNT has been synthesized without the participation of a catalyst.

Smalley et al. showed that single-wall carbon nanotubes SWNTs can also be produced in high yield by laser vaporization of a graphite rod doped with Co and Ni [15]. It was found later [16] that the structure of the carbon species produced in the laser ablation method strongly depends on the background argon pressure that is used. When the pressure is lower than 100 Torr no SWNTs are produced. They are only formed at higher pressures. Apparently, the Ar pressure plays an important role in the heat transfer phenomena and assists in the metal evaporation. Investigation of the carbon/metal target after the ablation showed that at low pressures, the metal evaporation was inhibited. The texture of the target was also found to have an effect on the quantity of

SWNT produced in the laser method. For example, twice as many SWNTs were produced when carbon targets containing Ni and Co nitrates were used as when Ni-Co metals or oxides were employed. This effect was ascribed to the porous structure that results when the Ni or Co nitrates decompose inside the carbon target. At the same time, a better dispersion and smaller particle size of the metals are obtained when using nitrates than when using metals or oxides [17]. Eklund et al. have recently developed a modification of the pulsed laser vaporization technique that, according to the authors, should result in large-scale production of high-quality SWNTs [18]. In this method, ultrafast ablation was achieved by using a high power free-electron laser. The modified setup includes a T-shaped quartz growth chamber placed inside a furnace to keep the chamber at 1000 °C. The laser radiation enters from a sidearm that protrudes out the furnace near the center of the hot zone and strikes the carbon target, which is mounted on a rotating/translating rod. A jet of preheated argon deflects the ablation plume away from the incident laser beam, continuously sweeping the target region. The SWNT soot is then collected from a water-cooled copper coldfinger at the end of the quartz nanotube. A special feature of this method was the use of the free-electron laser operated at a peak laser flux that is about 1000 times greater than the flux used in typical Neodymium(3 + )-yttrium-aluminum-garnet laser (Nd:YAG) based systems, but each FEL pulse is only 1/200,000 as long as the typical 10 ns Nd:YAG pulse.

The catalytic decomposition of carbon-containing compounds on appropriate metal catalysts is another method of producing SWNTs. This method, sometimes referred to as CVD (chemical vapor deposition), has a good potential for the production of large quantities of SWNTs at low cost. They also provide the possibility of producing SWNTs on a surface or inside a solid. This method is similar to those used for several decades in the synthesis of carbon filaments. It has been repeatedly demonstrated [19-21] that the diameter of the carbon nanotube is determined by the size of the metal cluster responsible for its growth. It is therefore important to tailor the precursor so that the catalyst particles retain a small size during the SWNT growth. In most studies, even when the metal particles may have had nano-metric dimensions before the growth started, they quickly sinter at the high temperatures needed for SWNT synthesis. Laurent et al. [22-27] and Resasco et al. [28-32] have independently designed catalysts which have the common characteristic of being produced only under reaction conditions. Solid oxide solutions such as Mg1-xCoxO [22] or silica supported Co-molybdates have been found to be effective catalysts when they are reduced in-situ by the same carbon-containing reactant that produces the SWNT. Only under specific conditions and using the proper catalyst formulation can SWNTs with high selectivity be produced. Successful examples of SWNT growth by the CVD method are the dis-proportionation of CO at 850 °C on Co:Mo/SiO2 (1:2 molar ratio) catalysts [29] and CH4 decomposition at 1000 °C on a Mg0 9Co01O solid solution catalyst [33]. In other cases, the CVD method has resulted in low selective growth of SWNT, with simultaneous formation of other carbon species such as graphite nanofibers and double walled or multiwalled nanotubes [34, 35].

An alternative to the CVD method of catalytic decomposition of carbon-containing molecules has been the so-called "floating catalyst" method. Sen et al. [36] prepared carbon nanotubes by decomposition of ferrocene, cobaltocene, and nickelocene under reductive conditions. In this case, the precursor provides both the carbon and the metal to catalyze the synthesis reaction. Similarly, in other methods benzene or hexane has been added to ferrocene improving the yield [37]. A variation of the floating catalyst method resulted in the commercial process known as HiPCO that produces 10 g/day of high-purity carbon singlewalled nanotubes [38]. In this process, SWNTs are grown at high pressure (30-50 atm) and high temperature (9001100 °C) under flowing CO. The catalyst is iron in the form of small clusters that are generated in-situ as continuously added iron pentacarbonyl decomposes in the reactor.

0 0

Post a comment