Since their discovery in 1991, carbon nanotubes (CNTs) have been the subject of intense research due to their unique electrical and mechanical properties. Because of these properties, CNTs are desirable for many applications including material reinforcement , catalyst-support media [47-49], hydrogen storage , gas detectors , field emission displays , and many others. For several of these applications, only modest quantities of nanotubes will be needed. In this respect, current production methods such as chemical vapor deposition (CVD) may be satisfactory because nanotubes can be produced with good control and low impurities at a reasonable cost . On the other hand, applications such as composite materials require industrial-scale quantities of carbon nanotubes.
In general, carbon nanotubes are produced in one of two ways: either on a supported catalyst or on a free-floating catalyst. In the supported catalyst method, particles such as iron are attached to a substrate and a carbon source such as a hydrocarbon or CO is passed over the substrate in a high-temperature environment. The nanotubes then grow from the catalyst particles. With the free-floating (aerosolized) catalyst method, particles are produced in the gas phase by flowing a catalyst precursor into the high-temperature environment with a source of carbon. The precursor typically decomposes to form the catalyst particles as an aerosol.
Aerosol methods have perhaps the highest probability of succeeding in the area of mass production of CNTs because they are continuous. Unfortunately, gas phase methods including laser ablation of composite metal targets [54,55], spray pyroly-sis [56-59], the HiPCO process , and flame synthesis [61-66] do not approach kg/hr production rates due to current limitations of each method. Surpassing these limitations is hindered by a slow optimization process, requiring feedback from time consuming TEM studies. Recognizing the importance of production of carbon nanotubes in aerosols, we have, in this section, developed an approach to online size characterization.
To rapidly evaluate the effects of process variables, online characterization of size, number concentration, and purity of carbon nanotubes is needed, and in this section we describe such an approach, wherein a DMA is employed. As noted in Section 9.2.2, the DMA is capable of online classification of fibers [42,44,45]. In addition, Maynard et al. have classified nanotubes using the DMA, although a method for determining CNT size was not explicitly developed .
As mentioned above, the DMA classifies particles by their electrical mobility. Following the mobility theory described in Section 9.3.2, the electrical mobility of a classified nanotube can be determined as a function of CNT dimensions within the uncertainty of the number of charges carried by a nanotube. We show in Section 9.3.3 that a system-specific charge parameter can be established as a function of the diameter, length, and number of charges carried by a nanotube. Using this expression and electrical mobility theory, if one dimension of the CNT is known, the other can be measured. In addition, the number concentration can be obtained for each size using a scanning mobility particle sizer (SMPS), which is a combination of the DMA and a condensation particle counter (CPC). Finally, we show in Section 9.3.5 that carbon nanotubes can be distinguished from excess catalyst particles that do not produce nanotubes. This is possible using a DMA in combination with weak particle charging because under conditions of only a few charges, the mobility of the nanotubes is small relative to spherical particles as has been experimentally demonstrated by Nasibulin et al. . Distinguishing excess catalyst particles from CNTs is important in order to minimize impurities due to excess catalyst particles.
To test this approach to online size characterization of CNTs, multiwall carbon nanotubes (MWNTs) ranging from 10 to 100 nanometers in diameter and 0.5 to 40 |im in length with a purity of 95% were aerosolized and the DMA was used to size-select CNTs by their electrical mobility. These CNTs were subsequently collected directly onto TEM grids using an electrostatic sampler. Transmission electron microscopy was used to determine the dimensions of the carbon nan-otubes. These dimensions were then used to test the methodology, and the results are presented in Section 9.3.5.
Was this article helpful?