Multidimensional Aspects of Single-Wall Carbon Nanotube Synthesis

Abstract

Keywords

reaction systems

growth mechanisms

catalysts

carbon sources

CVD

mass production

structural controllability

2.1. Various synthesis methods for single-wall carbon nanotubes

In 1993, the first single-wall carbon nanotubes (SWCNTs) were discovered [1] in the soot produced by the arc discharge using graphite-rod electrodes. However, it is well-known that the arc-discharge method in those days could produce only an extremely low yield of SWCNTs with lots of carbonaceous impurities. After the discovery, vigorous researches on a variety of production processes, such as laser vapourization [2] and chemical vapour deposition (CVD) [3–8], have been carried out, which have resulted in the significant increase in the SWCNT productivity. Progresses in the SWCNT production have promoted various researches in the broad field of study on the physical and chemical properties of SWCNTs. In this sense, the development of production processes for high-quality SWCNTs is the crucial key technology for SWCNT researches.

In general, the SWCNT production processes, such as arc-discharge and laser-vapourization methods, generate the high-crystallinity of graphene-like network in SWCNTs because of the high growth temperature. Accordingly, SWCNTs produced by these methods have afforded to show excellent potential in the application research [9]. For example, Figure 2.1, showing the recently reported performances of transparent conducting SWCNT films, supports the superiority of SWCNTs produced using high reaction temperature [9, 10].

On the other hand, the quality of SWCNTs produced by CVD has also been becoming better due to the technology maturity. As shown in Figure 2.1, the transparent conductivity performance of SWCNTs produced by CVD is steadily advancing, and to date it reaches to the level of the high-temperature production processes [11–14].

Arc-discharge and laser vapourization methods need lots of energy consumption for high reaction temperature and large-equipment configurations [15]. Furthermore, in these methods, the development of mass-production technology from lab to commercial scale is difficult in principle. On the other hand, the CVD method is generally known to be easy and well suited for the mass production. CVD method affords to produce SWCNTs from raw materials containing carbon-source molecules and catalysts, i.e. SWCNTs grow by supplying carbon generated by the chemical decomposition of carbon-source molecules to the catalyst nanoparticles. For the efficient carbon supply on the catalyst nanoparticles in the CVD growth of SWCNTs, various techniques decomposing carbon-source molecules, such as plasma and hot filament, have been utilized. Therefore, we frequently address separately these different CVD processes, such as thermal-, plasma- and hot-filament CVDs.

2.2. Catalysts for the SWCNT growth

Regardless of production processes, nanoparticles as the catalyst are indispensable for the growth of SWCNTs as a general rule, and the nanoparticles of iron-group transition metal, such as iron, cobalt and nickel, are frequently used [4]. It is well known that these iron-group transition metals possess the catalytic activity for the decomposing hydrocarbons and forming the graphite on the surface of their bulk crystals [16]. The carbon-source molecules in the gas-phase decompose on the surface of the catalyst nanoparticles and generate carbon atoms that would dissolve in the catalyst metal. As a result, according to the supersaturation of carbon in the metal, the depositing carbons nucleate on the surface of catalyst nanoparticles to form SWCNTs. It is expected that the catalyst deactivation in the above-mentioned growth mechanism can be caused by the carbonization and/or oxidization of nanoparticles due to the excess supply of the carbon source. Because SWCNTs in the solid phase are produced from carbon-source molecules in the gas phase and the catalyst nanoparticles in the liquid phase due to the effects of their size and eutectic reaction, this growth mechanism is called as vapour–liquid–solid (VLS) model [17]. However, it is still controversial whether the catalysts nanoparticles are in the liquid or solid phases, especially in CVD growth of SWCNTs because the reaction temperature is generally lower than that in the arc-discharge and laser vapourization methods [18].

Beside the iron-group metal, it is known that nanoparticles of aluminium [19] and noble metals, such as palladium, platinum, gold, silver and copper [20], can work as the catalyst for CVD growth of SWCNTs. Furthermore, studies on SWCNT growth by using non-metallic nanoparticles, such as silicon, germanium, silicon carbide [21], alumina [22] and diamond [23], have been also reported. Because the melting points of these non-metallic catalysts are relatively high in their bulk states, the growth mechanism of SWCNTs on these catalysts is expected to be different from the above-mentioned VLS model, although the detail is still an open question.

2.3. The way to introduce the catalyst on the CVD growth of SWCNTs

In terms of the way to introduce the catalysts in the reaction system for SWCNT growth, we can broadly classify lots of CVD processes into two general categories, i.e. supported- [3–5,7] and floating-catalyst [6,8,24–29] methods. In the former, catalyst nanoparticles are supported upon inert substrates or porous materials, while in the latter, they exist as aerosol within a reactive environment.

Because the supported-catalyst method has its advantage in the inexpensive equipment and wide process-condition range, a lot of commercially available SWCNTs are produced by this method. However, the following problems are generally known in the supported-catalyst method: firstly, the Ostwald ripening of catalyst nanoparticles due to the high temperature in the reactor expands their size distribution, which causes not only the deactivation of catalysts but also the large variability of the quality in diameter, length and crystallinity [30]. Secondly, the supported-catalyst method is basically batch process and it is difficult to develop the process for continuous production. Furthermore, the supported-catalyst method generally needs larger amounts of substrates or porous materials than the amount of produced SWCNTs, which can result in a significant increase in the cost. Although the supported-catalyst method is becoming improved by vigorous researches on the technological development, most of these problems are inherently difficult to be solved.

On the other hand, the floating-catalyst method has an advantage in its continuity of synthesizing products. That is, the floating-catalyst CVD synthesis, which is represented by HiPco [8], is believed to be an ideal continuous process for the production of SWCNTs at commercial scale. SWCNTs grown from the aerosol of catalyst nanoparticles floating in the furnace within the reactive environment flowed out and collected continuously. Continuous supply of fresh catalysts for the growth of SWCNTs and their quick outflow from the reactor prevent broadening the size distribution of the catalyst nanoparticles, which can lead to the production of SWCNTs with homogeneous quality.

As much as the author knows, the first study on the SWCNT growth by floating-catalyst method was reported by Cheng et al. in 1998 [6]. Similar to that reported for the VGCF production by Endo and Shikata [31], they produced high-quality SWCNTs by using the floating-catalyst CVD reactor claiming the relatively large-scale and low-cost production of SWCNTs. Thereafter, a variety of studies on the floating-catalyst method, including growth processes with different catalysts and carbon sources, have been reported [8,24–29], and some of them have demonstrated the specific controllability of structural features as described in subsequent sections.

2.4. Carbon sources leading to the...