Single wall carbon nanotubes (SWCNT) can be synthesized using thermal-CVD in the gas phase process by catalytic disproportionation of CO or by polymerization and hydrogenation of acetylene on iron particles. Iron is suplied in the form of iron pentacarbonyl. The yield of material from acetylene (aprox. 400mg/2h) is higher than from CO (approx.50mg). For the acetylene case, the synthesis is performed at 900°C at atmospheric pressure, while for the HIPCO case (CO disproportionation) the temperature is 1100°C.

Figure is a typical TEM image showing that the CO material consists of "ropes" of SWCNT, along with decoration by iron particles which are encapsulated in graphitic carbon

Figure 1: TEM images of SWCNT

Multi-wall carbon nanotubes (MWCNT) can be grown in both systems: thermal and plasma-CVD. The catalyst (Fe) is deposited on a Si substrate via e-beam evaporation. Typical films of MWCNT produced by thermal-CVD can be viewed in the figure 2. With this method it is possible to grow large arrays of aligned and non-aligned MWNT films.

Fig.2. SEM images of aligned MWCNT grown on patterns.

The average diameter of such tubes is around 15nm. These tubes are produced at atmospheric pressure at a temperature of 750°C using 10 time more hydrogen than acetylene. Argon is also introducedas a carrier and diluted gas. The homogeneity of the nanotube film on the substrate is very poor for very short growth time in the thermal CVD set-up. The density and diameter of the MWCNT can be changed as well if the amount of the catalyst film is varied. By this method we can produce very clean and aligned MWCNT film having a carpet-like aspect.

Using plasma enhanced-CVD as a production method, Fe catalyzed nanotubes films are much thinner and theirs length can be better controlled by time. The gaseous mixture and temperature conditions are similar for the production in thermal-CVD. The only difference is the pressure (7Torr) and the electric field used for plasma ignition which seems to play an imporatant role in their excellent alignment. Figure 3 illustrates the CNT films whose their length was varied by time from 10 sec till 45 minutes.

Fig.3. SEM sequence of nanotubes alignment obtained in plasma-CVD set-up for different growth time

In the first 5-8 minutes the nanotubes have a very high growth rate of approx. 9um/min. Unlike the nanotubes produced by thermal-CVD which will not suffer any changes in their structure and morphology if they are grown for longer time (t.ex. 3hous), the nanotubes synthesised by plasma-CVD are affected by the changing in the chemical composition of the gaseous mixture in a plasma atmosphere. In the case of tubes grown for time beyond 200s (tube lengths greater than ca. 30m), was observed drastically modifications in their morphology. These changes are not visible in the low resolution pictures, but can be clearly seen in the TEM micrographs. Figure 4 shows a combination of SEM and TEM images. The higher resolution TEM resolved very well the upper part of the tube. The outer layer consists of graphitic flakes with an interlayer distance of order 0.34nm.

Fig.4. SEM and TEM sequence of the nanotubes grown for longer time in plasma-CVD Upper images are SEM which depicts the graduate modification in the NT film from the bottom till top. TEM micrographs reveals the thorn-like aspect of the nanotubes longer than 200 um.

Individual vertically aligned MWCNT and carbon nanofiber are produced in plasma-CVD when are catalysed by nickel (Ni). These nanostrucutures can be produced when using ammonia and acetylene in plasma atmosphere at 700°C. Figure 5 shows examples of the nanostructures when a 10nm thick Ni films was deposited on silicon substrate with native oxide layer.

Fig.4.SEM images of individual nantubes/nanofibers grown in acetylene: ammonia plasma atmosphere at 700°C.

The nanotubes diameters are in the range 10-100nm. TEM investigation show that the growth is a via tip-growth mechanism and as the nanotubes diameter increases tubes change theirs structures. The small diameter structure are MWCNT, while the large diameter (>100 nm) show a bamaboo-like aspect of the inner-walls. Figure 6 show the TEM image of a substrate cross-section where Ni catalysed nanostructures were grown for 15 minutes.

Fig.6.TEM cross- section of the substrate where vertically aligned CNT/(CNF) were grown

The individuality of such nanostructures gives the possibility of a more controllable growth on patterned substrates where the position of the catalyst is specified. It was prepared two types of substrates: Ni patterned dots and Ni dots situated at the bottom of the 250nm deep holes in Si. The results after 15 minutes growth are shown in figure 7.

Fig.7.Growth of individual CNT/CNF on patterned dots.

One of the aim of these experiments is to test the ability of direct growth of individual nanostructures for future NEMS devices and CNT's based devices. Figure 8 show the first attemp where the NT was directly grown between the electrodes, but couldn't be measured due to the electrodes modification during the plasma growth. In the present, we are in the begining of understanding the fabrication process of these types of devices, but in the nearest future we will be able to produce them with high reproducibility.

Fig.8. First device attempt