Page 248 - Adsorbents fundamentals and applications

CARBON NANOTUBES 233

(and condensation) of graphite. The former has been the focus for commer-
cial production (and is being used by various manufacturers). Both methods
date back to the 1960’s and 1970’s. Some historical aspects of these two meth-
ods have been reviewed by Dresselhaus and Avouris (2001) and by Dressel-
haus and Endo (2001). These two methods, along with their brief histories, are
described separately.

9.1.1. Catalytic Decomposition
The formation of carbon filaments from decomposition of hydrocarbons was first
reported by Schutzenberger (1890) in 1890, and the first substantive report on
catalytic formation of carbon filaments was made by Radushkevich (as in the
D–R equation) and Luk’Yanovich (1952). A tremendous amount of literature
exists from studies performed by the carbon, catalysis, and metallurgy commu-
nities. A comprehensive review (including history) of the subject prior to 1978
is available in Baker and Harris (1978). Discussion on some of the later lit-
erature is available in Tibbetts (1990) and Yang and Chen (1989). Studies of
carbon filaments paralleled the developments of the transmission electron micro-
scope (TEM), which had become commercially available in the late 1940’s and
early 1950’s, and had a resolution near 1 nm at the time. The formation of car-
bon filaments received considerable attention primarily because of its detrimental
effects on catalyst deactivation and blast furnace operation (due to CO dispro-
portionation). The research objective was to understand the formation process
in order to minimize or alleviate this problem. It has long been known that Fe,
Co, and Ni are the most active catalysts for forming filaments. The filaments
typically had a well-defined hollow core. Figure 9.2 is a typical TEM image of
the hollow graphite filament, by Hilbert and Lange (1958). With the TEM reso-
lution high enough in the early 1970’s to distinguish the graphitic layers, it was
already known that some hollow filaments were formed by concentric layers of
graphite with ∼3.5 ˚ A spacing. Indeed, such TEM images have been published
(Baird et al., 1971; Baker and Harris, 1978; Fryer, 1979). One such example
is shown in Figure 9.3. After 1993, the hollow carbon filaments acquired the
name of MWNT. The term “carbon nanotube” was first used, in 1993, by Iijima
and Ichihashi (1993) and by Bethune et al. (1993). In the 1980’s, research on
catalytic carbon filament growth was driven by a different impetus: to produce
carbon fibers to replace those fibers made from PAN or pitch (Tibbetts et al.,
1986; Dresselhaus et al., 1988; Tibbetts, 1990). This has not occurred because
of the high costs of the catalytically grown carbon fibers.
In principle, carbon nanotubes can be grown from any gaseous hydrocarbons
or CO, onto Fe, Co, or Ni particles dispersed on a substrate under appropriate
reaction conditions. Higher temperatures and slower growth rates favor graphitic
filament formation, while lower temperatures and fast rates lead to nongraphitic
forms (Baker and Harris, 1978). Beside Fe, Co, and Ni, filaments can also be
formed on other metals such as Pt and Cu. Acetylene is among the most reactive
hydrocarbon precursors. Unsaturated hydrocarbons like propylene and butadi-
ene are more reactive than the saturated hydrocarbons such as methane and